Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 59 of the 59 references it cites were found in a registry or an index.

Widespread theta synchrony and high-frequency desynchronization underlies enhanced cognition

Open access

Abstract

The idea that synchronous neural activity underlies cognition has driven an extensive body of research in human and animal neuroscience. Yet, insufficient data on intracranial electrical connectivity has precluded a direct test of this hypothesis in a whole-brain setting. Through the lens of memory encoding and retrieval processes, we construct whole-brain connectivity maps of fast gamma (30-100 Hz) and slow theta (3-8 Hz) spectral neural activity, based on data from 294 neurosurgical patients fitted with indwelling electrodes. Here we report that gamma networks desynchronize and theta networks synchronize during encoding and retrieval. Furthermore, for nearly all brain regions we studied, gamma power rises as that region desynchronizes with gamma activity elsewhere in the brain, establishing gamma as a largely asynchronous phenomenon. The abundant phenomenon of theta synchrony is positively correlated with a brain region's gamma power, suggesting a predominant low-frequency mechanism for inter-regional communication.

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 · 10.1038/s41467-017-01763-2

Cites

59 of 59 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Network construction and basic analysis. a 3D visualization of all surface electrodes included in our data set, colored by the Talairach atlas labels used in this article’s analysis. b Schematic of spectral phase approach that compares the distributions of phase differences between electrodes across all trials of the verbal free-recall task. Significantly tighter distributions indicate greater synchronization. c Connectivity maps were extracted for each of 294 neurosurgical patients, reflecting the connectivity change associated with successful item recall. Effects were pooled across subjects and ROIs to construct the final network. Blue indicates decreased phase synchrony associated with successful encoding, red indicates increased synchrony. d 74 × 74 ROI adjacency matrices representing the z-scored time-averaged and frequency-averaged connection weights during the item presentation interval (0–1600 ms). The high gamma network is constructed from frequencies between 45–100 Hz, and theta from 3–8 Hz. Node indices are organized by lobe per the indicators on the axes. Gray areas represent connections between ROIs with fewer than seven subjects’ worth of data. e 3D visualizations of the whole-brain HG and theta networks. f Summed positive and negative connection weights in each frequency band. In a remembered vs. not-remembered contrast, the total level of synchronous theta connections and asynchronous HG connections were significantly greater than chance (P < 0.01), and there was a significant frequency-synchrony interaction (P < 0.01, χ 2 test). Dotted lines indicate mean chance level, shaded area ±1 STD
Fig. 1 Network construction and basic analysis. a 3D visualization of all surface electrodes included in our data set, colored by the Talairach atlas labels used in this article’s analysis. b Schematic of spectral phase approach that compares the distributions of phase differences between electrodes across all trials of the verbal free-recall task. Significantly tighter distributions indicate greater synchronization. c Connectivity maps were extracted for each of 294 neurosurgical patients, reflecting the connectivity change associated with successful item recall. Effects were pooled across subjects and ROIs to construct the final network. Blue indicates decreased phase synchrony associated with successful encoding, red indicates increased synchrony. d 74 × 74 ROI adjacency matrices representing the z-scored time-averaged and frequency-averaged connection weights during the item presentation interval (0–1600 ms). The high gamma network is constructed from frequencies between 45–100 Hz, and theta from 3–8 Hz. Node indices are organized by lobe per the indicators on the axes. Gray areas represent connections between ROIs with fewer than seven subjects’ worth of data. e 3D visualizations of the whole-brain HG and theta networks. f Summed positive and negative connection weights in each frequency band. In a remembered vs. not-remembered contrast, the total level of synchronous theta connections and asynchronous HG connections were significantly greater than chance (P < 0.01), and there was a significant frequency-synchrony interaction (P < 0.01, χ 2 test). Dotted lines indicate mean chance level, shaded area ±1 STDCC BY · PubMed Central
Fig. 2 Synchrony effects from 3 to 120 Hz. a Overall level of SME synchrony/asynchrony in six frequency bands spanning 3 Hz to 120 Hz, measured as in Fig. 1f (edge weight sum). Shaded gray areas represent chance mean ±1 STD. b Z-scored brain-wide phase synchronization subsequent memory effect (SME) in the memory encoding interval, measured by summing all connection weights in the network, compared to the sum expected by chance. This analysis is performed in successive 200 ms windows spanning the encoding interval. Red reflects increased synchrony associated with successful memory encoding, blue reflects decreased synchrony (see Methods section for details). Vertical black lines indicate word onset and offset
Fig. 2 Synchrony effects from 3 to 120 Hz. a Overall level of SME synchrony/asynchrony in six frequency bands spanning 3 Hz to 120 Hz, measured as in Fig. 1f (edge weight sum). Shaded gray areas represent chance mean ±1 STD. b Z-scored brain-wide phase synchronization subsequent memory effect (SME) in the memory encoding interval, measured by summing all connection weights in the network, compared to the sum expected by chance. This analysis is performed in successive 200 ms windows spanning the encoding interval. Red reflects increased synchrony associated with successful memory encoding, blue reflects decreased synchrony (see Methods section for details). Vertical black lines indicate word onset and offsetCC BY · PubMed Central
Fig. 3 Network hubs. Depiction of hub ROIs identified in the brain-wide theta and high gamma memory encoding networks. The analysis was performed separately for all positive connection weights (red) and all negative connection weights (blue), yielding “synchronous hubs” and “asynchronous hubs,” which respectively increase or decrease their connectivity with the network during successful memory encoding. Significant synchronous and asynchronous hubs for the item presentation interval are displayed according to their approximate localization on an average brain surface, with red circles indicating synchronous hubs and blue indicating asynchronous hubs (larger circles, FDR-corrected P < 0.01; smaller circles, P < 0.05). For each hub, the top five connections between that hub and any other part of the brain is plotted, if the connection weight z-score was greater than 2.5. Line thickness indicates absolute z-score value, according to the figure legend. Dashed lines indicate cross-hemispheric connections. Some labels are excluded from certain views to maintain readability. To aid visualization, hemispheres are reflected from their true position in the skull
Fig. 3 Network hubs. Depiction of hub ROIs identified in the brain-wide theta and high gamma memory encoding networks. The analysis was performed separately for all positive connection weights (red) and all negative connection weights (blue), yielding “synchronous hubs” and “asynchronous hubs,” which respectively increase or decrease their connectivity with the network during successful memory encoding. Significant synchronous and asynchronous hubs for the item presentation interval are displayed according to their approximate localization on an average brain surface, with red circles indicating synchronous hubs and blue indicating asynchronous hubs (larger circles, FDR-corrected P < 0.01; smaller circles, P < 0.05). For each hub, the top five connections between that hub and any other part of the brain is plotted, if the connection weight z-score was greater than 2.5. Line thickness indicates absolute z-score value, according to the figure legend. Dashed lines indicate cross-hemispheric connections. Some labels are excluded from certain views to maintain readability. To aid visualization, hemispheres are reflected from their true position in the skullCC BY · PubMed Central
Fig. 4 Timecourse of ROI participation in memory networks. Node strength as a function of time for 6 key regions that contain hubs in the theta or high gamma networks: right and left MTL, frontal lobe, and temporal lobe. Blue-shaded lines indicate asynchronous hub strength over time, while red indicates synchronous hub strength. Vertical lines indicate word onset and offset at 0 ms and 1600 ms. Above the z-scored timecourses are plotted the total count of specific ROIs within each broader region that reach significance at a given timepoint (P < 0.05). For visualization only, timecourses were smoothed with a 2-point moving average and radial basis filter
Fig. 4 Timecourse of ROI participation in memory networks. Node strength as a function of time for 6 key regions that contain hubs in the theta or high gamma networks: right and left MTL, frontal lobe, and temporal lobe. Blue-shaded lines indicate asynchronous hub strength over time, while red indicates synchronous hub strength. Vertical lines indicate word onset and offset at 0 ms and 1600 ms. Above the z-scored timecourses are plotted the total count of specific ROIs within each broader region that reach significance at a given timepoint (P < 0.05). For visualization only, timecourses were smoothed with a 2-point moving average and radial basis filterCC BY · PubMed Central
Fig. 5 Power-synchrony correlations across the whole brain. a Pearson correlation of the modulation of high gamma node strength and power across time and frequency for each ROI, during the word encoding interval. Bar plots show the power-synchrony correlation for each ROI, with blue indicating negative and red indicating positive correlations. Faded bars are not significant after FDR correction for multiple comparisons (α = 0.05). b For four example ROIs we depict time-frequency heatmaps of that ROI’s z-scored spectral power (top) and z-scored node strength (bottom). Red colors indicate a relative increase of power/synchrony when an item is successfully encoded, while blue indicates a relative decrease. For visualization only, absolute z-scores < 1.5 are faded, and vertical bars indicate word onset and offset. c Pearson correlation of z-scored power and z-scored node strength (synchrony) against each other for all timepoints and all ROIs, after averaging within frequency band. HG power and HG synchrony are significantly inversely related (P < 0.001, permutation test), HG synchrony is positively correlated with theta power (P = 0.005), while other tested relationships do not meet significance
Fig. 5 Power-synchrony correlations across the whole brain. a Pearson correlation of the modulation of high gamma node strength and power across time and frequency for each ROI, during the word encoding interval. Bar plots show the power-synchrony correlation for each ROI, with blue indicating negative and red indicating positive correlations. Faded bars are not significant after FDR correction for multiple comparisons (α = 0.05). b For four example ROIs we depict time-frequency heatmaps of that ROI’s z-scored spectral power (top) and z-scored node strength (bottom). Red colors indicate a relative increase of power/synchrony when an item is successfully encoded, while blue indicates a relative decrease. For visualization only, absolute z-scores < 1.5 are faded, and vertical bars indicate word onset and offset. c Pearson correlation of z-scored power and z-scored node strength (synchrony) against each other for all timepoints and all ROIs, after averaging within frequency band. HG power and HG synchrony are significantly inversely related (P < 0.001, permutation test), HG synchrony is positively correlated with theta power (P = 0.005), while other tested relationships do not meet significanceCC BY · PubMed Central
Fig. 6 Correlations in the core memory network. a Power-synchrony correlations in the core memory network: the 37 ROIs with significant HG power subsequent memory effect (SME). b Power-synchrony correlations across the 37 ROIs with the no significant HG power effects. Among the core memory network—consisting mostly of left frontal, temporal, and MTL cortex—z-scored gamma power and z-scored synchrony were significantly anticorrelated, while theta synchrony and gamma power were significantly positively correlated (top row; P < 0.001 and P < 0.01 via permutation test, respectively). Among regions that did not exhibit strongly modulated HG activity in successful memory encoding, HG power and synchrony were still inversely correlated (Pearson correlation, r = −0.158, P < 0.05), but theta synchrony was not significantly predictive of HG power (r = 0.131, P = 0.12)
Fig. 6 Correlations in the core memory network. a Power-synchrony correlations in the core memory network: the 37 ROIs with significant HG power subsequent memory effect (SME). b Power-synchrony correlations across the 37 ROIs with the no significant HG power effects. Among the core memory network—consisting mostly of left frontal, temporal, and MTL cortex—z-scored gamma power and z-scored synchrony were significantly anticorrelated, while theta synchrony and gamma power were significantly positively correlated (top row; P < 0.001 and P < 0.01 via permutation test, respectively). Among regions that did not exhibit strongly modulated HG activity in successful memory encoding, HG power and synchrony were still inversely correlated (Pearson correlation, r = −0.158, P < 0.05), but theta synchrony was not significantly predictive of HG power (r = 0.131, P = 0.12)CC BY · PubMed Central
Fig. 7 Generalization to memory retrieval processes. a Adjacency matrices, reflecting relative recall vs. baseline synchronization, organized as in Fig. 1d. b Summed positive and negative connection weights in each network, showing a strong desynchronization effect in gamma-band and a synchronization in theta (P < 0.01 for both). There was a significant frequency-synchrony interaction (P < 0.01, χ 2 test). c 3D representation of gamma (top) and theta (bottom) retrieval networks, organized as in Fig. 1e. d Correlation of spectral power and phase synchronization across all regions (74) and timepoints spanning a retrieval trial (2). HG power and synchrony were significantly inversely correlated (Pearson correlation, r = −0.698, P < 0.01 via permutation test), while an ROI’s theta synchrony was positively but not significantly predictive of HG power (r = 0.11, P = 0.36)
Fig. 7 Generalization to memory retrieval processes. a Adjacency matrices, reflecting relative recall vs. baseline synchronization, organized as in Fig. 1d. b Summed positive and negative connection weights in each network, showing a strong desynchronization effect in gamma-band and a synchronization in theta (P < 0.01 for both). There was a significant frequency-synchrony interaction (P < 0.01, χ 2 test). c 3D representation of gamma (top) and theta (bottom) retrieval networks, organized as in Fig. 1e. d Correlation of spectral power and phase synchronization across all regions (74) and timepoints spanning a retrieval trial (2). HG power and synchrony were significantly inversely correlated (Pearson correlation, r = −0.698, P < 0.01 via permutation test), while an ROI’s theta synchrony was positively but not significantly predictive of HG power (r = 0.11, P = 0.36)CC BY · PubMed Central
Fig. 8 Synchronization of low gamma (30–60 Hz) oscillatory activity. a Example of an electrode exhibiting a gamma oscillation in the middle occipital gyrus, as detected by BOSC (see Methods section for details). Red line reflects average log power across all remembered events, blue line reflects average log power across not-remembered events. An isolated peak in the power spectrum is indicated, between approximately 36 and 74 Hz. b For the electrode in a, heatmap of the t-statistic reflecting the relative frequency of oscillations detected in remembered vs. not-remembered trials. Red colors indicate more oscillations detected at given frequency and timepoint in trials that were later remembered correctly. Vertical lines indicate word onset and offset (0 and 1600 ms). T-statistics < 2 are faded, for visualization only. Increased oscillatory power from 43 Hz to 58 Hz, coincident with the oscillatory peak in a, is indicated. c Count of all electrodes in the 294-subject data set that exhibit an oscillatory subsequent memory effect (SME) between 30 and 58 Hz, at each 200 ms epoch spanning the word presentation interval (Methods section). The most electrodes exhibit oscillatory SMEs at 52 Hz, between 400 and 600 ms after onset of a word (black box). d Count of electrodes in the data set that exhibit different kinds of SMEs between 400 and 600 ms: 52 Hz oscillation, 30–60 Hz average power, or 65–100 Hz average power. e Count of significant synchronous or asynchronous network connections using only the subset of electrodes exhibiting 52 Hz oscillatory SME at 400–600 ms. Synchronization effects were deemed significant at P < 0.05, with the chance mean and standard deviation at this significance level indicated in the gray shaded area. Left: counts observed at 50–55 Hz, near the frequency of maximal oscillatory SMEs (52 Hz). Right: counts observed in the 65–85 Hz range among the same electrode subset. The frequency/synchrony interaction is not significant (P = 0.11). f Average network synchrony (z-score) for the subnetwork of regions sampled in the 52 Hz oscillatory electrode subset, measured by summing the subnetwork connection weights at each frequency in the 400–600 ms window, and comparing to the sum expected by chance
Fig. 8 Synchronization of low gamma (30–60 Hz) oscillatory activity. a Example of an electrode exhibiting a gamma oscillation in the middle occipital gyrus, as detected by BOSC (see Methods section for details). Red line reflects average log power across all remembered events, blue line reflects average log power across not-remembered events. An isolated peak in the power spectrum is indicated, between approximately 36 and 74 Hz. b For the electrode in a, heatmap of the t-statistic reflecting the relative frequency of oscillations detected in remembered vs. not-remembered trials. Red colors indicate more oscillations detected at given frequency and timepoint in trials that were later remembered correctly. Vertical lines indicate word onset and offset (0 and 1600 ms). T-statistics < 2 are faded, for visualization only. Increased oscillatory power from 43 Hz to 58 Hz, coincident with the oscillatory peak in a, is indicated. c Count of all electrodes in the 294-subject data set that exhibit an oscillatory subsequent memory effect (SME) between 30 and 58 Hz, at each 200 ms epoch spanning the word presentation interval (Methods section). The most electrodes exhibit oscillatory SMEs at 52 Hz, between 400 and 600 ms after onset of a word (black box). d Count of electrodes in the data set that exhibit different kinds of SMEs between 400 and 600 ms: 52 Hz oscillation, 30–60 Hz average power, or 65–100 Hz average power. e Count of significant synchronous or asynchronous network connections using only the subset of electrodes exhibiting 52 Hz oscillatory SME at 400–600 ms. Synchronization effects were deemed significant at P < 0.05, with the chance mean and standard deviation at this significance level indicated in the gray shaded area. Left: counts observed at 50–55 Hz, near the frequency of maximal oscillatory SMEs (52 Hz). Right: counts observed in the 65–85 Hz range among the same electrode subset. The frequency/synchrony interaction is not significant (P = 0.11). f Average network synchrony (z-score) for the subnetwork of regions sampled in the 52 Hz oscillatory electrode subset, measured by summing the subnetwork connection weights at each frequency in the 400–600 ms window, and comparing to the sum expected by chanceCC 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

Show all 59 references

Cited by

See all 244

Metadata

Authors
Solomon, E. A., Kragel, J. E., Sperling, M. R., Sharan, A., Worrell, G., Kucewicz, M., Inman, C. S., Lega, B., Davis, K. A., Stein, J. M., Jobst, B. C., Zaghloul, K. A., Sheth, S. A., Rizzuto, D. S., & Kahana, M. J.
Affiliations
  • University of Pennsylvania (US)
  • Thomas Jefferson University Hospital (US)
  • Mayo Clinic (US)
  • Emory University (US)
  • The University of Texas Southwestern Medical Center (US)
  • Hospital of the University of Pennsylvania (US)
  • Dartmouth–Hitchcock Medical Center (US)
  • National Institutes of Health (US)
  • Columbia University Irving Medical Center (US)
Corresponding author
  • Solomon, E. A.
  • Kahana, M. J.
Venue
Nature Communications
Published
2017-11-22
Volume
8
Issue
1
Pages
1704-1704
ISSN
2041-1723 (ISSN-L)
OpenAlex
W2770613913
Cited by
221
References
59
Citations by year
2026: 15 · 2025: 31 · 2024: 36 · 2023: 30 · 2022: 21 · 2021: 21 · 2020: 26 · 2019: 23 · 2018: 18
FWCI
10.52
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • The Medicine Forum (submittedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • U.S. Department of Defense (N660011424032, T32NS091006)
  • National Institutes of Health (5r01ns099348-04, MH55687)
  • Defense Advanced Research Projects Agency (N66001, Cooperative Agreement N66001-14-2-4032, N66001-14-2-4032)
  • National Institute of Mental Health (R01MH055687)
  • National Institute of Neurological Disorders and Stroke (T32NS091006, (R01 NS099348)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory and Neural Mechanisms
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Functional Brain Connectivity Studies
Keywords
whole-brain connectivity, memory encoding, memory retrieval, interregional communication, neurosurgical patients, cognitive enhancement
MeSH terms
Animals, Brain, Cognition, Humans, Memory, Mental Recall, Theta Rhythm, Electroencephalography Phase Synchronization, Connectome, Gamma Rhythm

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Widespread theta synchrony and high-frequency desynchronization underlies enhanced cognition. https://app.cochat.ai/papers/10.1038/s41467-017-01763-2
The paper
Solomon, E. A., Kragel, J. E., Sperling, M. R., Sharan, A., Worrell, G., Kucewicz, M., Inman, C. S., Lega, B., Davis, K. A., Stein, J. M., Jobst, B. C., Zaghloul, K. A., Sheth, S. A., Rizzuto, D. S., & Kahana, M. J. (2017). Widespread theta synchrony and high-frequency desynchronization underlies enhanced cognition. Nature Communications. https://doi.org/10.1038/s41467-017-01763-2

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.