pith. sign in

arxiv: 2606.10889 · v1 · pith:3U3O35MQnew · submitted 2026-06-09 · 🧬 q-bio.NC · cs.LG

Sleep EEG Signal Criticality as a Non-Invasive Predictor of Cognitive Decline in Dementia

Stanis{\l}aw Nar\k{e}bski , Tomasz Komendzi\'nski , Tomasz M. Rutkowski This is my paper

Pith reviewed 2026-06-27 10:51 UTC · model grok-4.3

classification 🧬 q-bio.NC cs.LG
keywords sleep EEGMFDFAbrain criticalitydementiaHurst exponentcognitive declinenon-REM sleepbiomarker
0
0 comments X

The pith

Women who later develop dementia exhibit sleep EEG signals shifted away from an optimally critical state.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests whether sleep EEG signal criticality, measured by multifractal detrended fluctuation analysis, can predict future cognitive decline. It compares baseline sleep EEG from older women who remained healthy against those who later met criteria for dementia-related impairment. Healthy individuals showed brain dynamics closer to an optimally critical state at all electrode sites during non-REM sleep. Those progressing to dementia displayed a shift in exponents toward 1.0, pointing to an early reconfiguration of scale-free dynamics. This suggests sleep EEG could provide a non-invasive way to detect prodromal changes before symptoms emerge.

Core claim

Cognitively healthy individuals exhibited signal dynamics significantly closer to an optimally critical state across all electrode locations (p ≤ 0.001), while the dementia group demonstrated a shift in DFA exponents toward 1.0, indicating that a reconfiguration of scale-free neural dynamics during sleep precedes clinical symptoms.

What carries the argument

Multifractal Detrended Fluctuation Analysis (MFDFA) on sleep EEG to compute Hurst exponents H(q) that quantify proximity to a critical state in brain signals.

If this is right

  • MFDFA measures could be integrated into automated sleep-based screening tools for earlier detection.
  • Preventative interventions could be targeted during the prodromal window of dementia.
  • The findings support applying the Brain Criticality Hypothesis to clinical prediction of neurodegeneration.
  • Group separation is visible in UMAP projections throughout the sleep stages.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar criticality shifts might appear in other sleep-related or neurodegenerative conditions beyond this cohort.
  • Future work could test if restoring criticality through interventions alters the trajectory of cognitive decline.
  • The method may extend to daytime EEG or other physiological signals for broader monitoring.

Load-bearing premise

That the observed differences in MFDFA exponents are caused by impending dementia rather than other factors such as medications, comorbidities, or variations in sleep patterns.

What would settle it

Finding no significant difference in Hurst exponents between future dementia cases and controls in a new prospective cohort, or showing that the difference fails to predict decline after accounting for confounders.

Figures

Figures reproduced from arXiv: 2606.10889 by Stanis{\l}aw Nar\k{e}bski, Tomasz Komendzi\'nski, Tomasz M. Rutkowski.

Figure 1
Figure 1. Figure 1: Distributions of Hurst exponents obtained from MFDFA ( [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Supervised UMAP visualization of multifractal EEG dynamics. Two-dimensional embeddings of MFDFA features ( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
read the original abstract

Early detection of neurodegeneration remains a critical clinical challenge. This study investigates whether sleep EEG signal criticality, quantified via Multifractal Detrended Fluctuation Analysis (MFDFA), serves as a non-invasive biomarker for future cognitive decline. We analyzed longitudinal data from the National Sleep Research Resource (NSRR) Study of Osteoporotic Fractures (SOF) cohort, comparing baseline sleep EEG dynamics between women who remained cognitively normal and those who later progressed to dementia-related impairment ($3MS < 78$).Our results reveal significant group-level differences in Hurst exponent $H(q)$ distributions, particularly during non-REM stages N2 and N3. Cognitively healthy individuals exhibited signal dynamics significantly closer to an optimally critical state across all electrode locations ($p \leqslant 0.001$), supporting the Brain Criticality Hypothesis. Supervised UMAP projections confirmed clear spatial separation between groups throughout the overnight sleep architecture.The dementia group demonstrated a shift in DFA exponents toward $1.0$, suggesting that a reconfiguration of scale-free neural dynamics during sleep precedes clinical symptoms. These findings highlight the potential for MFDFA-derived measures to be integrated into automated, sleep-based screening tools, enabling earlier preventative interventions during the prodromal window of dementia.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript analyzes baseline sleep EEG recordings from the SOF cohort using MFDFA to extract multifractal Hurst exponents H(q). It reports statistically significant group differences (p ≤ 0.001) during N2/N3 stages between participants who remained cognitively normal and those who later met a 3MS < 78 threshold for dementia-related impairment, with the healthy group closer to an 'optimally critical' state and the dementia group shifted toward H = 1.0. UMAP embeddings are said to separate the groups, and the authors interpret the baseline differences as evidence that scale-free reconfiguration precedes clinical symptoms, proposing MFDFA measures as a non-invasive prodromal biomarker.

Significance. If the reported group differences survive confounder adjustment, multiple-comparison correction, and out-of-sample validation, the work would provide empirical support for the brain criticality hypothesis in a longitudinal human cohort and could motivate development of sleep-EEG-based screening tools. The use of an existing longitudinal dataset (NSRR/SOF) is a positive feature.

major comments (3)
  1. [Methods] Methods section: sample sizes for the two outcome groups, exact statistical tests, multiple-testing corrections, and EEG preprocessing pipeline (artifact rejection, epoch selection, electrode montage) are not reported, rendering the p ≤ 0.001 claims across electrodes unverifiable and load-bearing for the central group-difference result.
  2. [Results] Results and Discussion: the interpretation that the Hurst-exponent shift constitutes a 'predictor' or 'prodromal' signal is not supported by any longitudinal survival model, covariate-adjusted regression, or medication/comorbidity matching; the 3MS < 78 cutoff is applied without reported validation against clinical dementia diagnosis in this elderly-female cohort.
  3. [Results] Results: the supervised UMAP separation is presented as confirmatory evidence, yet no description of the supervision procedure, feature set, or cross-validation is given, so it is impossible to determine whether the separation reflects genuine predictive structure or retrospective group labeling.
minor comments (2)
  1. [Abstract] Abstract: the phrase 'optimally critical state' is used without stating the reference Hurst value expected for criticality or citing the relevant theoretical literature.
  2. Figure legends and text: electrode locations are referred to collectively as 'all electrode locations' without listing the montage or reporting per-electrode effect sizes.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. We have revised the manuscript to address the methodological omissions, moderate interpretive claims, and clarify the UMAP analysis. Our responses to each major comment follow.

read point-by-point responses

  1. Referee: [Methods] Methods section: sample sizes for the two outcome groups, exact statistical tests, multiple-testing corrections, and EEG preprocessing pipeline (artifact rejection, epoch selection, electrode montage) are not reported, rendering the p ≤ 0.001 claims across electrodes unverifiable and load-bearing for the central group-difference result.

    Authors: We agree these details were insufficiently reported. The revised Methods section now specifies the final sample sizes for each outcome group after inclusion criteria, the exact statistical tests (including any non-parametric alternatives), the multiple-comparison correction procedure applied across electrodes and q-values, and a complete EEG preprocessing description covering artifact rejection, epoch selection from N2/N3 stages, and the electrode montage/referencing scheme. These additions render the reported p-values verifiable. revision: yes

  2. Referee: [Results] Results and Discussion: the interpretation that the Hurst-exponent shift constitutes a 'predictor' or 'prodromal' signal is not supported by any longitudinal survival model, covariate-adjusted regression, or medication/comorbidity matching; the 3MS < 78 cutoff is applied without reported validation against clinical dementia diagnosis in this elderly-female cohort.

    Authors: We accept that the original language overstated the evidence. The baseline group differences demonstrate an association with later impairment but do not constitute a validated predictor without survival modeling or full confounder adjustment. We have revised the Results and Discussion to use more cautious phrasing (e.g., 'associated with future decline' rather than 'predictor' or 'prodromal biomarker'), added an explicit limitations paragraph noting the lack of survival analysis and medication matching, and clarified that the 3MS threshold, while standard in the SOF literature, lacks direct validation against clinical dementia diagnoses in this specific analysis. revision: partial

  3. Referee: [Results] Results: the supervised UMAP separation is presented as confirmatory evidence, yet no description of the supervision procedure, feature set, or cross-validation is given, so it is impossible to determine whether the separation reflects genuine predictive structure or retrospective group labeling.

    Authors: We agree the UMAP description was incomplete. The revised manuscript now details the supervision procedure (group labels as the target variable), the exact feature set (MFDFA-derived H(q) values across q-range, electrodes, and sleep stages), and the cross-validation approach used to assess whether the observed separation generalizes beyond the training labels. These additions clarify that the visualization is exploratory and supported by the cross-validation results. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper computes MFDFA Hurst exponents independently on baseline sleep EEG, then compares the resulting distributions between groups whose membership is fixed by a later clinical threshold (3MS < 78) that is external to the EEG processing pipeline. No parameter of the MFDFA procedure is fitted or tuned to the dementia label, no self-citation supplies a uniqueness theorem that forces the observed separation, and the supervised UMAP step is presented only as visualization rather than as a predictive claim whose success is tautological. The reported p-values therefore reflect an ordinary between-group contrast rather than a quantity that reduces to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review performed on abstract only; full text may contain additional free parameters in MFDFA implementation or further domain assumptions.

axioms (1)
  • domain assumption Brain Criticality Hypothesis
    The claim that healthy individuals are closer to an optimally critical state and that deviation indicates impending decline rests on this hypothesis being a valid interpretive framework.

pith-pipeline@v0.9.1-grok · 5768 in / 1366 out tokens · 37542 ms · 2026-06-27T10:51:06.613037+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

21 extracted references · 1 canonical work pages · 1 internal anchor

  1. [1]

    Dementia prevention, intervention, and care: 2024 report of the lancet standing commission,

    G. Livingston, J. Huntley, K. Y . Liu, S. G. Costafreda, G. Selbæk, S. Alladi, D. Ames, S. Banerjee, A. Burns, C. Brayne,et al., “Dementia prevention, intervention, and care: 2024 report of the lancet standing commission,”The Lancet, vol. 404, no. 10452, pp. 572–628, 2024

    2024

  2. [2]

    Mild cognitive impairment: an overview,

    R. C. Petersen and S. Negash, “Mild cognitive impairment: an overview,” CNS Spectrums, vol. 13, no. 1, pp. 45–53, 2008

    2008

  3. [3]

    Five-year follow- up of cognitive impairment with no dementia,

    H. Tuokko, R. Frerichs, J. Graham, K. Rockwood, B. Kristjansson, J. Fisk, H. Bergman, A. Kozma, and I. McDowell, “Five-year follow- up of cognitive impairment with no dementia,”Archives of Neurology, vol. 60, no. 4, pp. 577–582, 2003

    2003

  4. [4]

    Sleep problems and risk of all-cause cognitive decline or dementia: an updated system- atic review and meta-analysis,

    W. Xu, C.-C. Tan, J.-J. Zou, X.-P. Cao, and L. Tan, “Sleep problems and risk of all-cause cognitive decline or dementia: an updated system- atic review and meta-analysis,”Journal of Neurology, Neurosurgery & Psychiatry, vol. 91, no. 3, pp. 236–244, 2020

    2020

  5. [5]

    Association of sleep duration in middle and old age with incidence of dementia,

    S. Sabia, A. Fayosse, J. Dumurgier, V . T. van Hees, C. Paquet, A. Som- merlad, M. Kivimäki, A. Dugravot, and A. Singh-Manoux, “Association of sleep duration in middle and old age with incidence of dementia,” Nature Communications, vol. 12, no. 1, p. 2289, 2021

    2021

  6. [6]

    Sleep drives metabolite clearance from the adult brain,

    L. Xie, H. Kang, Q. Xu, M. J. Chen, Y . Liao, M. Thiyagarajan, J. O’Donnell, D. J. Christensen, C. Nicholson, J. J. Iliff,et al., “Sleep drives metabolite clearance from the adult brain,”Science, vol. 342, no. 6156, pp. 373–377, 2013

    2013

  7. [7]

    Predicting future risk of developing cognitive impairment using ambulatory sleep EEG: Integrating univariate analysis and multivariate information theory approach,

    S. Haghayegh, R. Herzog, D. A. Bennett, S. Redline, K. Yaffe, K. L. Stone, A. Ibáñez, and K. Hu, “Predicting future risk of developing cognitive impairment using ambulatory sleep EEG: Integrating univariate analysis and multivariate information theory approach,”Journal of Alzheimer’s Disease, p. 13872877251319742, 2025

    2025

  8. [8]

    25 years of criticality in neuro- science—established results, open controversies, novel concepts,

    J. Wilting and V . Priesemann, “25 years of criticality in neuro- science—established results, open controversies, novel concepts,”Cur- rent Opinion in Neurobiology, vol. 58, pp. 105–111, 2019

    2019

  9. [9]

    Tomen, J

    N. Tomen, J. M. Herrmann, and U. Ernst,The functional role of critical dynamics in neural systems, vol. 11. Springer, 2019

    2019

  10. [10]

    The effect of linear mixing in the EEG on hurst exponent estimation,

    D. A. Blythe, S. Haufe, K.-R. Müller, and V . V . Nikulin, “The effect of linear mixing in the EEG on hurst exponent estimation,”NeuroImage, vol. 99, pp. 377–387, 2014

    2014

  11. [11]

    The national sleep research resource: towards a sleep data commons,

    G.-Q. Zhang, L. Cui, R. Mueller, S. Tao, M. Kim, M. Rueschman, S. Mariani, J. L. Goodwin, and S. Redline, “The national sleep research resource: towards a sleep data commons,”Journal of the American Medical Informatics Association, vol. 25, no. 10, pp. 1351–1358, 2018

    2018

  12. [12]

    Sleep-disordered breathing and cognition in older women,

    A. P. Spira, T. Blackwell, K. L. Stone, S. Redline, J. A. Cauley, S. Ancoli-Israel, and K. Yaffe, “Sleep-disordered breathing and cognition in older women,”Journal of the American Geriatrics Society, vol. 56, no. 1, pp. 45–50, 2008

    2008

  13. [13]

    Appendicular bone density and age predict hip fracture in women,

    S. R. Cummings, D. M. Black, M. C. Nevitt, W. S. Browner, J. A. Cauley, H. K. Genant, D. Mascioli, J. C. Scott, and J. K. Seeley, “Appendicular bone density and age predict hip fracture in women,” JAMA, vol. 263, no. 5, pp. 665–668, 1990

    1990

  14. [14]

    ‘MINI-MENTAL STATE’: A practical method for grading the cognitive state of patients for the clinician,

    M. F. Folstein, S. E. Folstein, and P. R. McHugh, “‘MINI-MENTAL STATE’: A practical method for grading the cognitive state of patients for the clinician,”Journal of Psychiatric Research, vol. 12, no. 3, pp. 189–198, 1975

    1975

  15. [15]

    The modified mini-mental state (3MS) examina- tion,

    E. Teng and H. Chui, “The modified mini-mental state (3MS) examina- tion,”The Journal of Clinical Psychiatry, vol. 48, p. 314—318, August 1987

    1987

  16. [16]

    The AASM manual for the scoring of sleep and associ- ated events: rules, terminology, and technical specification,

    C. Iber, “The AASM manual for the scoring of sleep and associ- ated events: rules, terminology, and technical specification,”American Academy of Sleep Medicine, 2007

    2007

  17. [17]

    Multifractal detrended fluctuation analysis of nonstationary time series,

    J. W. Kantelhardt, S. A. Zschiegner, E. Koscielny-Bunde, S. Havlin, A. Bunde, and H. E. Stanley, “Multifractal detrended fluctuation analysis of nonstationary time series,”Physica A: Statistical Mechanics and its Applications, vol. 316, no. 1-4, pp. 87–114, 2002

    2002

  18. [18]

    Quantifi- cation of scaling exponents and crossover phenomena in nonstationary heartbeat time series,

    C.-K. Peng, S. Havlin, H. E. Stanley, and A. L. Goldberger, “Quantifi- cation of scaling exponents and crossover phenomena in nonstationary heartbeat time series,”Chaos: An Interdisciplinary Journal of Nonlinear Science, vol. 5, no. 1, pp. 82–87, 1995

    1995

  19. [19]

    Machine learning approach for early onset dementia neurobiomarker using EEG network topology features,

    T. M. Rutkowski, M. S. Abe, T. Komendzinski, H. Sugimoto, S. Nareb- ski, and M. Otake-Matsuura, “Machine learning approach for early onset dementia neurobiomarker using EEG network topology features,” Frontiers in Human Neuroscience, vol. 17, 2023

    2023

  20. [20]

    UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction

    L. McInnes, J. Healy, and J. Melville, “UMAP: uniform manifold approximation and projection for dimension reduction,”arXiv preprint arXiv:1802.03426, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  21. [21]

    Neighborhood preservation in nonlinear pro- jection methods,

    J. Venna and S. Kaski, “Neighborhood preservation in nonlinear pro- jection methods,” inProceedings of the Symposium on Self-Organizing Maps, pp. 273–280, 2001

    2001