CoChat

Adolescent development of cortical oscillations: Power, phase, and support of cognitive maturation

Open access
Marek, Scott, Tervo-Clemmens, Brenden, Klein, Natalie, Foran, William, Ghuman, Avniel Singh, Luna, Beatriz · 2018 · PLOS Biology

The DOI matches its Crossref record; 92 of the 92 references it cites were found in a registry or an index.

Abstract

During adolescence, the integration of specialized functional brain networks related to cognitive control continues to increase. Slow frequency oscillations (4-10 Hz) have been shown to support cognitive control processes, especially within prefrontal regions. However, it is unclear how neural oscillations contribute to functional brain network development and improvements in cognitive control during adolescence. To bridge this gap, we employed magnetoencephalography (MEG) to explore changes in oscillatory power and phase coupling across cortical networks in a sample of 68 adolescents and young adults. We found a redistribution of power from lower to higher frequencies throughout adolescence, such that delta band (1-3 Hz) power decreased, whereas beta band power (14-16 and 22-26 Hz) increased. Delta band power decreased with age most strongly in association networks within the frontal lobe and operculum. Conversely, beta band power increased throughout development, most strongly in processing networks and the posterior cingulate cortex, a hub of the default mode (DM) network. In terms of phase, theta band (5-9 Hz) phase-locking robustly decreased with development, following an anterior-to-posterior gradient, with the greatest decoupling occurring between association networks. Additionally, decreased slow frequency phase-locking between frontolimbic regions was related to decreased impulsivity with age. Thus, greater decoupling of slow frequency oscillations may afford functional networks greater flexibility during the resting state to instantiate control when required.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pbio.2004188

Cites

92 of 92 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Workflow diagram. Preprocessing: Raw MEG and structural MRI data were preprocessed and coregistered. After surface ROI time series were extracted, a PLV was calculated for each frequency in the interval from 1–49 Hz, resulting in an ROI × ROI PLV matrix at each frequency interval for each subject. Global PLV: For each frequency, the mean PLV between all ROI pairs was calculated for each subject. Subject age was then regressed onto this global mean at each frequency to test for significant age effects, controlling for power. Regional PLV: Slow frequency (5–9 Hz) PLV matrices were averaged for each subject. Age was then regressed onto PLV for each edge of the matrix. The beta weight associated with age for every edge was extracted from each regression model. To summarize regional changes, we summed down the columns of the matrix, resulting in a composite linear age effect for each ROI. Global/regional power: Similar to the PLV pipeline, we calculated relative power for each ROI at each frequency interval. To obtain a global measure of power, we averaged power across all ROIs within a frequency band. For each region, we regressed power at a given frequency interval onto age and extracted the beta weight from the age regressor for additional analyses. Regional power estimates were examined for age effects and also included as nuisance regressors in all PLV × Age models. ICA, independent components analysis; MEG, magnetoencephalography; MNE, minimum-norm estimate; PLV, phase-locking value; ROI, region of interest; tSSS, temporal signal space separation.
Fig 1 Workflow diagram. Preprocessing: Raw MEG and structural MRI data were preprocessed and coregistered. After surface ROI time series were extracted, a PLV was calculated for each frequency in the interval from 1–49 Hz, resulting in an ROI × ROI PLV matrix at each frequency interval for each subject. Global PLV: For each frequency, the mean PLV between all ROI pairs was calculated for each subject. Subject age was then regressed onto this global mean at each frequency to test for significant age effects, controlling for power. Regional PLV: Slow frequency (5–9 Hz) PLV matrices were averaged for each subject. Age was then regressed onto PLV for each edge of the matrix. The beta weight associated with age for every edge was extracted from each regression model. To summarize regional changes, we summed down the columns of the matrix, resulting in a composite linear age effect for each ROI. Global/regional power: Similar to the PLV pipeline, we calculated relative power for each ROI at each frequency interval. To obtain a global measure of power, we averaged power across all ROIs within a frequency band. For each region, we regressed power at a given frequency interval onto age and extracted the beta weight from the age regressor for additional analyses. Regional power estimates were examined for age effects and also included as nuisance regressors in all PLV × Age models. ICA, independent components analysis; MEG, magnetoencephalography; MNE, minimum-norm estimate; PLV, phase-locking value; ROI, region of interest; tSSS, temporal signal space separation.CC BY · PubMed Central
Fig 2 Developmental differences in global cortical phase-locking and power. (A) Across most frequency bands, adolescents displayed similar resting-state phase-locking to adults. However, in the 5–9 Hz frequency band, there was a significant linear decrease in phase-locking throughout development (gray shaded region; p < 0.05, FDR corrected). Top color bar represents the magnitude of the t statistic from the PLV × Age regression model. Data displayed categorically after segregation into 2 groups via a median split. (B) Power as a function of frequency. Delta band power significantly decreased with age, whereas beta band power significantly increased with age. Top color bar represents the magnitude of the t statistic from the Power × Age regression model. Data displayed categorically after segregation into 2 groups via a median split. In both (A) and (B), shaded gray patches represent frequency intervals demonstrating a significant linear relationship with age. Red and blue lines and shaded bars in the line plots represent the mean (solid line) and standard error of the mean (shaded patch around mean), respectively, in adolescents (red) and adults (blue). See S1 Data for individual data points. FDR, false discovery rate; PLV, phase-locking value.
Fig 2 Developmental differences in global cortical phase-locking and power. (A) Across most frequency bands, adolescents displayed similar resting-state phase-locking to adults. However, in the 5–9 Hz frequency band, there was a significant linear decrease in phase-locking throughout development (gray shaded region; p < 0.05, FDR corrected). Top color bar represents the magnitude of the t statistic from the PLV × Age regression model. Data displayed categorically after segregation into 2 groups via a median split. (B) Power as a function of frequency. Delta band power significantly decreased with age, whereas beta band power significantly increased with age. Top color bar represents the magnitude of the t statistic from the Power × Age regression model. Data displayed categorically after segregation into 2 groups via a median split. In both (A) and (B), shaded gray patches represent frequency intervals demonstrating a significant linear relationship with age. Red and blue lines and shaded bars in the line plots represent the mean (solid line) and standard error of the mean (shaded patch around mean), respectively, in adolescents (red) and adults (blue). See S1 Data for individual data points. FDR, false discovery rate; PLV, phase-locking value.CC BY · PubMed Central
Fig 3 Regional age-related differences in phase-locking. (A) Regional age-related decreases in theta band phase-locking. (B) Scatter plot containing summed regional age effect (beta weight from theta PLV × Age model) as a function of the region’s anatomical y-coordinate center of mass. (C) PLV × Age anatomical gradient as a function of frequency. We found that the greatest anterior-to-posterior gradient developmental effect was in the 6–15 Hz regime. Gray error bars represent standard error of the model fit. Red shaded bar denotes theta/alpha regime. y-Axis represents the beta weight (slope) of the relationship between PLV and age with the anatomical y-coordinate of the region’s center of mass. See S1 Data for individual data points. PLV, phase-locking value.
Fig 3 Regional age-related differences in phase-locking. (A) Regional age-related decreases in theta band phase-locking. (B) Scatter plot containing summed regional age effect (beta weight from theta PLV × Age model) as a function of the region’s anatomical y-coordinate center of mass. (C) PLV × Age anatomical gradient as a function of frequency. We found that the greatest anterior-to-posterior gradient developmental effect was in the 6–15 Hz regime. Gray error bars represent standard error of the model fit. Red shaded bar denotes theta/alpha regime. y-Axis represents the beta weight (slope) of the relationship between PLV and age with the anatomical y-coordinate of the region’s center of mass. See S1 Data for individual data points. PLV, phase-locking value.CC BY · PubMed Central
Fig 4 Regional age-related differences in power. (A) Regional age-related decreases in delta band power. (B) Regional age-related increases in beta band power. Scatter plots contain summed regional age effect (beta weight from Power × Age model) as a function of the region’s anatomical y-coordinate center of mass. (C) Power × Age anatomical gradient as a function of frequency. Error bars represent standard error of the model fit. Red shaded bar denotes delta and beta band regimes from panels A and B. y-Axis represents the beta weight (slope) of the relationship between power and age with the anatomical y-coordinate of the region’s center of mass. See S1 Data for individual data points.
Fig 4 Regional age-related differences in power. (A) Regional age-related decreases in delta band power. (B) Regional age-related increases in beta band power. Scatter plots contain summed regional age effect (beta weight from Power × Age model) as a function of the region’s anatomical y-coordinate center of mass. (C) Power × Age anatomical gradient as a function of frequency. Error bars represent standard error of the model fit. Red shaded bar denotes delta and beta band regimes from panels A and B. y-Axis represents the beta weight (slope) of the relationship between power and age with the anatomical y-coordinate of the region’s center of mass. See S1 Data for individual data points.CC BY · PubMed Central
Fig 5 Network changes in phase-locking. (A) Age-related decreases in phase-locking tended to be within and between association networks (e.g., DM, FP, and SAL), while within- and between-network oscillations involving processing networks remained relatively stable. (B) Age-related increases in slow frequency decoupling were greater in association networks than in processing networks (p = 10−9). Oscillations in the SAL network became significantly more decoupled compared to any other association or processing network, with the exception of the CP network (all p < 0.05, corrected). See S1 Data for individual data points. AUD, Auditory; CO, Cinguloopercular; CP, Cinguloparietal; DA, Dorsal Attention; DM, Default Mode; FP, Frontoparietal; NONE, Unknown; RT, Retrospenial Temporal; SAL, Salience; SMH, Somatomotor Hand; SMM, Somatomotor Mouth; VA, Ventral Attention; VIS, Visual.
Fig 5 Network changes in phase-locking. (A) Age-related decreases in phase-locking tended to be within and between association networks (e.g., DM, FP, and SAL), while within- and between-network oscillations involving processing networks remained relatively stable. (B) Age-related increases in slow frequency decoupling were greater in association networks than in processing networks (p = 10−9). Oscillations in the SAL network became significantly more decoupled compared to any other association or processing network, with the exception of the CP network (all p < 0.05, corrected). See S1 Data for individual data points. AUD, Auditory; CO, Cinguloopercular; CP, Cinguloparietal; DA, Dorsal Attention; DM, Default Mode; FP, Frontoparietal; NONE, Unknown; RT, Retrospenial Temporal; SAL, Salience; SMH, Somatomotor Hand; SMM, Somatomotor Mouth; VA, Ventral Attention; VIS, Visual.CC BY · PubMed Central
Fig 6 Pairwise age-related decreases in resting-state phase-locking. Pairwise increases in decoupling between the top 5% of brain regions that showed age-related increases in decoupling (developmental hubs) and their respective top 5% pairwise interactions. Regions (little circles) are colored by the network to which they are affiliated. Link color represents the network affiliation to which the developmental hub belonged. The most significant pairwise increases occurred between regions of the DM, and FP networks to other association networks. DM, Default Mode; FP, Frontoparietal.
Fig 6 Pairwise age-related decreases in resting-state phase-locking. Pairwise increases in decoupling between the top 5% of brain regions that showed age-related increases in decoupling (developmental hubs) and their respective top 5% pairwise interactions. Regions (little circles) are colored by the network to which they are affiliated. Link color represents the network affiliation to which the developmental hub belonged. The most significant pairwise increases occurred between regions of the DM, and FP networks to other association networks. DM, Default Mode; FP, Frontoparietal.CC BY · PubMed Central
Fig 7 Frontolimbic 5–9 Hz phase-locking is related to decreased impulsivity during adolescence. (A) Anatomical location PLV × Age and PLV × Impulsivity relationships. Red links denote the significant PLV × Age NBS cluster. Orange links denote the significant PLV × Impulsivity NBS cluster. Yellow links denote overlap between the 2 clusters. These overlapping links were tested for mediation. (B) Mediation model including statistics for specific paths. Note PLV of these 3 interactions fully mediated the relationship between age and impulsivity (difference in p-values between path C and paths in C’), confirming overlap of clusters. NBS, network-based statistic; PLV, phase-locking value.
Fig 7 Frontolimbic 5–9 Hz phase-locking is related to decreased impulsivity during adolescence. (A) Anatomical location PLV × Age and PLV × Impulsivity relationships. Red links denote the significant PLV × Age NBS cluster. Orange links denote the significant PLV × Impulsivity NBS cluster. Yellow links denote overlap between the 2 clusters. These overlapping links were tested for mediation. (B) Mediation model including statistics for specific paths. Note PLV of these 3 interactions fully mediated the relationship between age and impulsivity (difference in p-values between path C and paths in C’), confirming overlap of clusters. NBS, network-based statistic; PLV, phase-locking value.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Research works printed but not linked: 2

  1. RW Thatcher, 2008, Development of cortical connections as measured by EEG coherence and phase delays, Human Brain Mapping, 29, 1400, 10.1002/hbm.20474 · DOI · Paper page
  2. G Gonzalez-Burgos, 2014, Functional Maturation of GABA Synapses During Postnatal Development of the Monkey Dorsolateral Prefrontal Cortex, Cereb Cortex

Cited by

See all 48

Metadata

Authors
Marek, S., Tervo-Clemmens, B., Klein, N., Foran, W., Ghuman, A. S., & Luna, B.
Affiliations
  • Washington University in St. Louis (US)
  • Carnegie Mellon University (US)
  • University of Pittsburgh (US)
Corresponding author
Marek, Scott
Venue
PLOS Biology
Published
2018-11-30
Volume
16
Issue
11
Pages
e2004188
ISSN
  • 1544-9173 (ISSN-L)
  • 1545-7885
OpenAlex
W2902082599
Cited by
48
References
91
Citations by year
2026: 8 · 2025: 14 · 2024: 4 · 2023: 6 · 2022: 5 · 2021: 5 · 2020: 4 · 2019: 2
FWCI
1.64
Type
journal article
Language
en
Open access
Locations
  • PLoS Biology (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Institute of Mental Health (MH100019, R21MH103592, R01MH107797, R01MH067924, T32MH100019)
  • National Institute on Drug Abuse (T32DA031111)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Functional Brain Connectivity Studies
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
Keywords
adolescent brain development, cortical oscillations, cognitive control, functional brain networks, delta power, beta power, phase coupling, default mode network, posterior cingulate cortex, impulsivity, prefrontal cortex, association networks, resting-state functional connectivity
MeSH terms
Adolescent, Adult, Brain, Brain Mapping, Cerebral Cortex, Cognition, Female, Humans, Impulsive Behavior, Male, Nerve Net, Magnetoencephalography, Adolescent Development, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.