Aperiodic dynamics track cortical state shifts during sleep K-complexes

Authors

Kaplan, E. S., Schindler, J., Voytek, B.

Abstract

Sleep microarchitecture is defined by brief electrophysiological events such as K-complexes(KCs), which are traditionally interpreted as discrete oscillatory phenomena involved in sleep maintenance, sensory gating, and cortical down-states. However, electroencephalography(EEG) also contains aperiodic, broadband activity that reflects non-oscillatory population dynamics and varies systematically across sleep stages. Because KCs are large, low-frequency transient waveforms, they also manifest across a broader frequency range, which creates a bidirectional measurement problem in which discrete events and aperiodic activity are difficult to disentangle. This is because aperiodic dynamics alone can generate fluctuations that resemble KCs, while conversely, the KC waveform itself can inflate the low-frequency end of the spectrum, steepening the fitted exponent even when the underlying aperiodic state is unchanged. These confounds blur what constitutes a real KC and a real aperiodic change in the presence of a KC. A KC-locked change in the aperiodic exponent could therefore reflect a genuine cortical state shift, a measurement artifact from the waveform biasing the spectral fit, or both. Here, we use time-resolved spectral parameterization of overnight sleep EEG to characterize aperiodic dynamics around expert-annotated KCs, with a series of controls designed to separate genuine aperiodic state changes from waveform artifacts. A component of the KC-locked aperiodic change survives these controls, suggesting that KCs are accompanied by a genuine shift toward a higher-exponent, inhibition-dominated cortical state. These findings establish aperiodic activity as a meaningful index of state dynamics during KCs and highlight measurement confounds that any event-locked spectral analysis of large transients must consider.

Preprint server: bioRxiv
The authors list and abstract were imported from bioRxiv on 30 Sep 2026.

Stats

1-terrible, 9-excellent. How would you rate this preprint? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.