Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 22.
Published in final edited form as: Clin Neurophysiol. 2025 Dec 7;183:2111475. doi: 10.1016/j.clinph.2025.2111475

Auditory-evoked changes in slow oscillations and spindles correlate with memory consolidation in children with epilepsy and controls

Hunki Kwon 1,2, Dhinakaran M Chinappen 1,2, Anirudh Wodeyar 1,2,7, Elizabeth A Kinard 1, Skyler K Goodman 1, Wen Shi 1,2, Bryan S Baxter 2,3, Dara S Manoach 2,3,4, Mark A Kramer 5,6, Catherine J Chu 1,2
PMCID: PMC12820947  NIHMSID: NIHMS2129584  PMID: 41435615

Abstract

Objective:

We investigated the effects of auditory stimulation during sleep on slow oscillations (SOs), SO-spindle complexes, and sleep-dependent memory consolidation in children with Rolandic Epilepsy (RE) and controls.

Methods:

Participants completed two nap visits with auditory or sham stimulation. SOs and SO-spindle complexes rates were measured offline using validated detectors. Sleep-dependent memory consolidation was assessed using the motor sequence typing task.

Results:

Auditory stimulation evoked SOs and SO-spindle complexes with maximal effect over frontal electrodes. Compared to sham, stimulation of background activity increased SOs (29.8%, p<0.001) and SO-spindle complexes (16.8%, p<0.001); stimulation of an ongoing SO upstate maximally evoked SOs (51.3% increase, p<0.001) and SO-spindle complexes (32.3% increase, p<0.001). Changes in frontal SO (1.9% improvement per increase in SO/min; p<0.001) and SO-spindle complexes (9.5% improvement per increase in SO-spindle/min; p=0.007) event rates due to auditory stimulation positively predicted changes in sleep-dependent memory consolidation.

Conclusion:

Auditory stimulation reliably modulates sleep oscillations when delivered on background activity and during the upstate of SOs. As increased event rates improve memory consolidation, stimulation paradigms to increase SO and SO-spindle complex rates are required to enhance memory.

Significance:

Auditory stimulation paradigms that increase SO and SO-spindle complex rates may enhance sleep-dependent memory consolidation.

Keywords: Pediatric, Rolandic epilepsy, open-loop, cognitive function

1. Introduction

Sleep is critical for memory consolidation -- the integration of newly acquired information into long-term storage (Born and Wilhelm, 2012). Slow oscillations (SOs), large low frequency brain rhythms (0.5–2 Hz) during deep non-rapid eye movement (NREM) sleep synchronize thalamocortical spindles (bursts of 10–14 Hz oscillations(Kwon et al., 2023)) and hippocampal sharp wave ripples (80–120 Hz), facilitating communication between the hippocampus and neocortex (Born and Wilhelm, 2012). The depolarizing SO upstate facilitates the initiation of sleep spindles by GABAergic neurons in the thalamic reticular nucleus (McCormick and Bal, 1997, Steriade et al., 1987), leading to an increased occurrence of spindles observed in the cortex and thalamus compared to baseline periods (Schreiner et al., 2021, Steriade et al., 1993, Wodeyar et al., 2024). Sleep spindles may facilitate synaptic plasticity by regulating dendritic calcium shifts (Seibt et al., 2017) and coordinating hippocampal sharp wave ripples that drive neuronal replay and contribute to memory consolidation (Buzsaki, 2015, Staresina et al., 2015). The functional coupling of SOs, spindles, and hippocampal sharp wave ripples during NREM sleep supports synaptic potentiation and interregional communication, essential for memory consolidation (Born and Wilhelm, 2012, Dudai et al., 2015, Geva-Sagiv et al., 2023).

Rolandic epilepsy (RE), also known as self-limited epilepsy with centrotemporal spikes (SeLECTS), is the most common focal developmental epilepsy in childhood, accounting for 8–23% of childhood epilepsy (Ross et al., 2020). RE is characterized by sleep-activated spikes in the inferior Rolandic cortex and varying degrees of cognitive deficits during school-age years epilepsy (Ross et al., 2020). The most prominent cognitive symptom in RE is memory impairment (Wickens et al., 2017), and includes deficits in sleep-dependent memory consolidation (Kwon et al., 2025). In RE, cortical regions with epileptiform spikes also have a paucity of sleep spindles (Kramer et al., 2021) and both sleep-dependent memory and IQ correlate with spindle rate (Kramer et al., 2021, Kwon et al., 2025).

Work in the last decade has demonstrated that auditory stimulation phase-targeted to the upstate of endogenous SOs during NREM sleep can induce a subsequent SO and associated spindle activity (Baxter et al., 2023, Besedovsky et al., 2017, Leminen et al., 2017, Ngo et al., 2013, Ngo et al., 2019, Ong et al., 2016, Prehn-Kristensen et al., 2020). This increase in SOs and spindles was initially reported to improve declarative memory consolidation in healthy adults (Bellesi et al., 2014, Ngo et al., 2013, Ong et al., 2016) and in typically developing children (Prehn-Kristensen et al., 2020). However, subsequent studies found that auditory stimulation reliably evokes SOs and SO-sleep spindle complexes, but does not reliably improve declarative or procedural memory consolidation in healthy adults (Baxter et al., 2023, Harrington et al., 2021, Leminen et al., 2017) or children (Prehn-Kristensen et al., 2020). Importantly, although prior work suggests that the rates of SOs, sleep spindles, and SO-spindle complexes best predict memory consolidation (Astill et al., 2014, Hahn et al., 2019, Hahn et al., 2020, Kwon et al., 2025, Mölle and Born, 2011), most studies evaluating the impact of auditory stimulation on memory consolidation did not evaluate the overall rate of these oscillations. When applied in epilepsy, one prior study found that auditory stimulation during NREM sleep suppressed spike activity in children with RE (Klinzing et al., 2021), however, the impact on sleep oscillations supporting memory and consequent sleep-dependent memory consolidation was not evaluated (Fattinger et al., 2019, Klinzing et al., 2021).

In this study, we hypothesized that auditory stimulation during stages 2 and 3 NREM sleep would evoke SO and SO-spindle complexes in children with RE and controls, compared to sham stimulation (i.e., the same detection procedure but with no sound played) delivered at the same phase. We also hypothesized that changes in SO and SO-spindle complex rates due to auditory stimulation would predict changes in sleep-dependent memory consolidation. To test these hypotheses, we performed a prospective cross-over study and evaluated the rates of SOs and SO-spindle complexes and sleep-dependent memory consolidation during a nap with randomly time auditory stimulation compared to a nap with sham stimulation. Understanding how auditory stimulation impacts sleep oscillations and memory consolidation will enable translation of this non-invasive, scalable, neuromodulatory approach to enhance cognitive function in children impacted by epilepsy and other conditions that impact memory.

2. Materials and Methods

2.1. Participants

We prospectively recruited children with Rolandic epilepsy (RE) and control children to participate. All participants that enrolled in the study reported in (Kwon et al., 2025) were invited to participate in this study. Diagnosis of RE was confirmed by a pediatric epileptologist (C.J.C.) based on International League Against Epilepsy criteria, requiring a history of at least one focal motor or generalized seizure and EEG evidence of sleep-activated centrotemporal spikes (1989, Fisher et al., 2014). Children with attention disorders or mild learning difficulties were included as these are common symptoms in RE (Ross et al., 2020, Wickens et al., 2017). Informed consent was obtained from all participants, and the study was approved by the Massachusetts General Hospital Institutional Review Board.

2.2. Study design

The study consisted of two visits to the Athinoula A. Martinos Center for Biomedical Imaging. For each visit, participants arrived at approximately 10:00 AM and began with motor sequence task training (Figure 1A). This was followed by a nap opportunity lasting approximately 90 minutes, scheduled between 1:00 PM and 3:00 PM. After the nap, participants completed the motor sequence task testing. Each visit was assigned to either no stimulation or auditory stimulation (see Auditory Stimulation). In each case, the EEG was monitored continuously during the nap opportunity and stimulation was delivered throughout stage 2 and stage 3 NREM sleep.

Figure 1. Experimental overview.

Figure 1.

Open in a new tab

A) One visit included auditory stimulation with randomly timed 50 ms bursts of pink noise (STIM) during Stages 2 and 3 NREM sleep, and one visit included no auditory stimulation (SHAM). B) Participants performed the finger tapping motor sequence task with their left and right hand separated by a 10-minute break. C) Sleep-dependent memory consolidation is the percentage difference between the number of correct sequences during the last three trials of training and the first three trials of testing.

2.3. Motor sequence task

To measure sleep-dependent memory consolidation, participants completed the finger tapping motor sequence task (MST) as in (Kwon et al., 2025). This involved typing with the left hand a 5-element sequence on a labeled number pad (e.g., 2-4-1-3-2, Figure 1B) as quickly and accurately as possible for twelve 30 second trials, each separated by a 30-second rest interval (Manoach et al., 2016, Wamsley et al., 2012). The sequence was displayed during trials to minimize working memory load. After a 10-minute break, participants performed the task with their right hand using a different sequence. To exclude performance outliers across trials, for each participant and each hand, we fit an exponential model to the learning curve during training, and included a constant offset to the exponential model during the post-sleep testing, as in (Kwon et al., 2025, Manoach et al., 2004). If performance on a trial was less than expected (defined as more than two standard deviations below the model fit), it was excluded as an outlier. Outliers could be due to misplaced fingers or task interruption, for example. Sleep-dependent memory consolidation was calculated as the percentage difference between the mean performance of the last three trials before sleep and the first three trials after awakening and measured separately for the left- and right-hand MST (Figure 1C), as done previously (Kwon et al., 2025).

2.4. EEG recordings

EEG recordings were acquired using 70-channel high density EEG caps (Easycap, Vectorview, Elekta-Neuromag, Helsinki, Finland) at 2035 Hz or 2000 Hz sampling rates. Data were downsampled to 407 Hz or 400 Hz for analysis. Electrode impedances were maintained below 10 kΩ. Electrode positions were digitized using a 3D digitizer (Fastrak; Polhemus Inc., Colchester, VA). A board-certified neurophysiologist visually inspected the data in both average and nasion referenced montages to identify channels with poor recording quality and to remove epochs containing large muscle, motion, or electrode artifacts that would contaminate interpretation or analysis of the signals of interest. Data were then sleep staged following standard procedures (Grigg-Damberger et al., 2007).

2.5. Auditory stimulation

Auditory stimulation was delivered using Psychtoolbox (Kleiner et al., 2007) in MATLAB (The MathWorks, R2021b, Natick, MA), while EEG data were continuously monitored in real-time by a board-certified neurophysiologist (C.J.C.). When participants entered stages 2 and 3 NREM sleep, as determined by visual inspection, 50 ms bursts of pink noise were randomly presented through in-ear headphones, with an interstimulus interval ranging from 2 to 30 seconds. The volume of auditory stimulation was set based on participant feedback prior to the nap opportunity, adapted for comfort without disturbing sleep, within a range of 32 to 35 dB. The ambient noise level in the room was ~30dB. To prevent sleep disruptions, stimulations were paused manually if arousals or awakenings occurred, and distractions in the recording environment (e.g., noise in the recording room) were minimized.

2.6. Offline slow oscillation detector

SOs were identified offline after referencing sleep-staged artifact-free EEG data to the nasion. A midline non-cephalic reference was used due to the global spatial distribution of SOs (Tononi and Cirelli, 2014, Wodeyar et al., 2024). Stages 2 and 3 NREM EEG data were bandpass filtered (0.5–4 Hz) and SOs detected if two consecutive positive-to-negative zero crossings occurred within 0.5–2 seconds and the negative peak during this interval was less than or equal to −40 μV, adapted from (Mölle et al., 2002).

2.7. Offline sleep spindle detector

For spindle detection, stages 2 and 3 NREM EEG data were re-referenced to an average reference. An average reference was used to improve topographic localization of focal spindles (Kramer et al., 2021, Kwon et al., 2023). To detect sleep spindles, we used an automated spindle detector, developed to perform well in the setting of sharp events in EEG, such as pediatric vertex waves or epileptiform spikes (Kramer et al., 2021, Kwon et al., 2023). Spindles were required to last at least 0.5 seconds (Iber and Medicine, 2007, Purcell et al., 2017) and spindles detected within 1 second of each other were concatenated (Kramer et al., 2021, Purcell et al., 2017).

2.8. Evoked SO and SO-spindle complex percentage calculation

Randomly timed auditory stimulations were delivered during NREM stages 2 and 3 sleep. The percentage of stimulations that evoked SOs was calculated as the proportion of SOs occurring within 1 second following auditory stimulation divided by the total number of stimulations multiplied by 100. The percentage of stimulations that evoked SO-spindle complexes was determined as the number of spindles occurring within 1 second after the negative peak of an evoked SOs following auditory stimulation divided by the total number of stimulations multiplied by 100. The percentage of stimulations that evoked spindles was calculated as the number of spindles occurring within 2.7 s following stimulation (chosen to match the duration of the longest SO duration across participants) divided by the total number of stimulations, multiplied by 100. These percentages of evoked SO and SO-spindle complexes were calculated separately for stimuli delivered during baseline activity and for stimuli delivered during an endogenous SO (i.e., a spontaneous SO not evoked by auditory stimulation). For stimulations delivered during endogenous SOs, percentages were calculated across 12 phase bins, each representing 30° increments of the endogenous SO phase. To estimate the endogenous SO phase, the EEG data were bandpass filtered (0.5–2 Hz) and the smoothed phase was estimated after correcting for waveform distortions caused by narrowband filtering (Davis et al., 2020). This approach enabled us to evaluate the proportion of auditory stimulations that evoked a SO and/or SO-spindle complex relative to the endogenous SO phase at the time of auditory stimulation. Percentages were then compared to sham stimulations delivered during baseline activity or at the same phase of an endogenous SO.

2.9. Time frequency analysis

To visualize spectral dynamics around auditory stimulation, sleep-staged artifact-free stages 2 and 3 NREM EEG data were analyzed using the short-time Fourier transform. For each participant, spectrograms were computed with a 250 ms Hamming window and 90% overlap, converted to decibels (dB), and baseline-corrected using the interval −1000 to −500 ms before auditory stimulation onset. The baseline-corrected spectrograms were then averaged across trials. Separate analyses were performed for endogenous SOs (upstate or downstate stimulations) and for periods without detected SOs, with the stimulation and sham conditions analyzed separately.

2.10. Statistical analyses

To identify significant intervals of times where the EEG signals, spindle detections, evoked SO, or evoked SO-spindle complex percentages differed between auditory stimulation and sham stimulation, temporally contiguous cluster-based permutation tests were performed (Kwon et al., 2021, Kwon et al., 2023). A two-sided paired t-test was first used to identify time points (or phase bins) that exhibited significant differences between stimulation conditions across participants. Time points (or phase bins) with a p-value below the critical alpha levels (p < 0.05 and p< 0.01) were clustered if the p-values were temporally adjacent within 50 ms (or adjacent based on phase bins). Temporally contiguous cluster-level statistics were determined by summing the absolute t-values within each cluster. To determine which temporally contiguous cluster-level statistics were unlikely to occur by chance, condition order was randomly shuffled within each participant, and the largest temporally contiguous cluster-level statistic was measured from the resampled data using the same method as for the unpermuted data. Temporally contiguous clusters from the unpermuted data were considered significant if their statistics exceeded the top 5% and 1% of statistics computed from 5000 resamplings of the data. Identical procedures were used to detect differences in phases of the endogenous SOs that evoked spindle detections, evoked SO detections, and evoked SO-spindle complex events. Identical procedures were also used to identify spatial clusters of electrodes where the evoked SO or SO-spindle complex percentage differed between auditory stimulation and sham stimulation, except that a spatially contiguous cluster-based permutation test was performed instead of a temporally contiguous cluster-based permutation test (Kwon et al., 2021, Kwon et al., 2023). For this approach, channels with a p-value below the critical alpha level were clustered if they were spatially adjacent. For visualization, topographic maps of spatially contiguous significant clusters were interpolated using the Fieldtrip toolbox (http://www.ru.nl/neuroimaging/fieldtrip) (Oostenveld et al., 2011). The same cluster-based permutation framework was also applied to the spectrograms to identify significant differences in spectral power between auditory stimulation and sham conditions. In this analysis, two-sided paired t-tests were performed at each time-frequency bin, and bins with p-values below the critical alpha level were clustered if they were temporally adjacent and contiguous across neighboring frequency bins. Cluster-level statistics were calculated as the sum of absolute t-values within each cluster, and significance was determined using 5000 permutations, with clusters with a p-value below the critical alpha level considered significant.

To test for a relationship between event rates (SO or SO-spindle complex rates) and sleep-dependent memory improvement in the combined sham and auditory stimulation groups, we estimated a linear mixed-effects model with sleep-dependent memory consolidation as the dependent variable, event rate, age, sex, and visit order as predictors, and a participant-specific intercept to account for four observations per participant (left and right hemisphere; sham and auditory stimulation). We assumed that event rates were linked to memory improvement with the contralateral hand (Nishida and Walker, 2007).

To examine how changes in event rates (SO or SO-spindle complex rates) induced by auditory stimulation impacted sleep-dependent memory consolidation compared to sham stimulation, we estimated a linear mixed-effects model with the change in sleep dependent memory consolidation with auditory stimulation (compared to sham) as the dependent variable and the change in event rate with auditory stimulation (compared to sham), age, sex, and visit order as predictors, and a participant-specific intercept to account for multiple observations per participant.

In post-hoc analysis during the review process, we repeated the above tests evaluating spindle rate.

2.11. Data availability

Raw data were generated at Massachusetts General Hospital and the Athinoula A. Martinos Center for Biomedical Imaging. Derived data supporting the findings of this study are available from the corresponding author on request. The spindle detection method is available at https://github.com/Mark-Kramer/Spindle-Detector-Method.

3. Results

3.1. Participant characteristics

11 children with RE and 9 controls were enrolled. 3 children with RE and 2 controls did not sleep during the nap opportunity and were excluded. In total, 8 children with RE (9.5–17.7 years, 6F) and 7 control children (9.3–16.2 years, 1F) were included. There were no significant difference in age (p=0.55, two-sample t-test) or duration between visits (p=0.12, two-sample t-test) between groups, however more RE participants were female compared to controls (p=0.02, χ2=5.5, Chi-square test). All participants were right-handed. Sleep opportunity, stages 2 and 3 NREM sleep duration, the total number of stimulations, the number of stimulations during SOs, the number of sleep features (spindles, SOs, and SO-spindle complexes), amplitude of SOs, and memory consolidation were not significantly different between sham and stimulation conditions in either group (all p>0.17, paired t-tests). More RE participants had auditory stimulation during their first visit (5/8) and more control participants had sham stimulation during their first visit (6/7). Participant characteristics are provided in Table 1.

Table 1.

Participant characteristics

RE Control p value
Sample size 8 7 -
Age (years) 13.3 (9.5–17.7) 14.2 (9.3–16.2) 0.55a
Female sex (%) 75 14.3 0.02b
Right handedness (%) 100 100 -
Duration between visits (days) 11.1 (1–21) 24.9 (8–71) 0.12a
Total number of stimulations (n) 538.6 (197–934) 565.6 (183–872) 0.91a
RE p value
Condition SHAM STIM
Visit 1 3 5
Sleep opportunity (min) 97.4 (90.0–145.0) 94.4 (82.0–114.0) 0.70c
Sleep duration (N2+N3, min) 26.5 (3.4–68.9) 40.9 (9.4–59.9) 0.17c
Total number of stimulations (n) 224.8 (26–453) 318.9 (64–701) 0.32c
Number of stimulations during SOs (n) 54.1 (9–118) 79.9 (8–184) 0.30c
Number of sleep features (n)
Spindles 192.2 (12–560) 204.4 (29.0–484) 0.89c
SOs 545.2 (30–2191) 1031.6 (113.0–1944) 0.18c
SO-spindle complexes 70.2 (2.0–295) 95.8 (8.0–265) 0.58c
Amplitude of SOs (μV) −72.6 (−107.2 to −50.0) −79.6 (−113.9 to −56.5) 0.45c
Memory consolidation (%)
Left hand 1.4 (−44.2 to 25.0) 14.5 (−9.4 to 64.5) 0.30c
Right hand 8.7 (−6.9 to 30.8) 1.8 (−22.2 to 20.8) 0.32c
Control p value
Condition SHAM STIM
Visit 1 6 1
Sleep opportunity (min) 91.6 (90.0–97.0) 97.7 (90.0–125.0) 0.24c
Sleep duration (N2+N3, min) 37.0 (8.4–65.3) 39.3 (17.4–58.5) 0.83c
Total number of stimulations (n) 297.4 (61–474) 272.4 (122–398) 0.75c
Number of stimulations during SOs (n) 61.9 (16–128) 66.4 (11–122) 0.85c
Number of sleep features (n)
Spindles 244.1 (3–520) 230.1 (101–557) 0.88c
SOs 956.4 (25–2197) 881.4 (203–1760) 0.85c
SO-spindle complexes 89.9 (1–240) 84.7 (39–237) 0.90c
Amplitude of SOs (μV) −71.5 (−84.4 to −52.0) −72.7 (−88.0 to −62.3) 0.85c
Memory consolidation (%)
Left hand 23.3 (−5.5 to 55.7) 14.1 (0.3 to 44.9) 0.43c
Right hand 18.9 (−1.1 to 60.7) 15.3 (−4.7 to 34.4) 0.70c
Open in a new tab
a

Two-sample t-test

b

Chi-square test

c

Paired t-test

3.2. Auditory stimulation during sleep evokes SOs and sleep spindles.

Considering the population results from all participants, regardless of group, auditory stimulation evoked a slow oscillation compared to sham stimulation, indicated by a large triphasic wave with an initial positive deflection near 150 ms, followed by a negative deflection between 300–500 ms, and a second positive deflection near 900 ms (p<0.05, cluster-based statistic, Figure 2A). Auditory stimulation significantly increased SO activity compared to sham (p < 0.05, cluster-based statistic, Figure 2B). Auditory stimulation tended to reduce spindle activity at the time of stimulation (near 0 ms) followed by a significant increase in sleep spindle detections beginning near 800 ms post-stimulation compared to sham stimulation (p<0.05, cluster-based statistic, Figure 2C), temporally coinciding with the second upstate of the induced slow oscillation. The average morphology of the evoked events are provided in Supplementary Figure 1.

Figure 2. Auditory stimulation evokes SOs and spindles.

Figure 2.

Open in a new tab

A) Auditory stimulations (blue) versus sham (black) averaged across all participants ± standard area of the mean. Positive is plotted upward, opposite to EEG convention. B) Auditory stimulations evoke SOs (blue curve) compared to sham stimulations (black curve). C) Following an initial trend toward reduction in spindles at the time of stimulation, auditory stimulations evoke sleep spindles concurrent with the evoked SO compared to s stimulations.

3.3. Topological distribution of evoked SOs and SO-spindle complexes

To examine the spatial distribution of evoked SOs, we evaluated the percentages of evoked SO averaged separately for each electrode and compared between stimulation and sham conditions. Stimulations evoked SOs broadly across nearly all scalp electrodes compared to sham stimulation, with maximal effects observed over the frontal, central, and parietal regions near the vertex (p<0.01, cluster-based statistic; Figure 3A).

Figure 3. Topological distributions of evoked SO and SO-spindle complexes from auditory stimulation.

Figure 3.

Open in a new tab

The distribution of A) evoked SO and B) evoked SO-spindle complex when auditory stimulation is delivered compared to sham stimulation. Circles indicate electrodes with significantly different event proportions between stimulation and sham conditions (grey: p < 0.05; green: p<0.01, cluster-based statistic). Black-outlined circles indicate symmetric electrodes used to evaluate the impact of auditory stimulation on memory consolidation.

Following a similar analysis approach, the percentages of evoked SO-spindle complexes were averaged separately for each electrode and compared between stimulation and sham conditions. Stimulations evoked SO-spindle complexes broadly (p<0.05, gray Figure 3B) but were most prominent over the frontal electrodes near the vertex compared to sham stimulation (p<0.01, cluster-based statistic; green, Figure 3B). In post-hoc analysis, we found that when evaluating all evoked spindles, they follow a similar topographical distribution as evoked SO-spindle complexes, with the largest increases observed over frontal regions (Supplementary Figure 2A).

These findings indicate that this non-invasive approach can be used to evoke SOs and SO-spindle complexes broadly over the cortex but that the effects are maximal over frontal regions near the midline.

3.4. Auditory stimulation delivered during an endogenous SO upstate preferentially evokes SOs and SO-spindle complexes

To examine the impact of the timing of auditory stimulation relative to endogenous brain activity to evoke SO and SO-spindle complexes, we evaluated the percentage of auditory that evoked a SO delivered in the presence or absence of an endogenous SO compared to sham stimulations delivered on a comparable background (Figure 4A). For this analysis, we focused on FZ given the strong effect size observed in topological analysis (Figure 3). When auditory stimulations were delivered in the absence of an endogenous SO, an evoked SO occurred 51.0% of the time (Control: 45.4%; RE: 55.9%) and an evoked SO-spindle complex occurred 26.5% of the time (Control: 23.6%; RE: 29.0%). Compared to sham stimulation, the proportion of evoked SOs with auditory stimulation increased by 29.8% (95% CI [20.0 39.5], p<0.001); an increase was observed in each group (Control: 21.1% increase, 95% CI [7.4 34.8], p=0.009; RE: 37.4% increase, p<0.001, two-sided paired t-test; Figure 4B). Compared to sham stimulation, the proportion of evoked SO-spindle complexes increased by 16.8% (95% CI [10.1 23.5], p<0.001); this was also observed in each group (Control: 13.0%, 95% CI [3.6 22.5], increase, p=0.02; RE: 20.1% increase, 95% CI [9.0 31.3], p=0.004, two-sided paired t-test; Figure 4C).

Figure 4. Impact of endogenous activity on stimulation efficacy.

Figure 4.

Open in a new tab

A) When an endogenous SO is detected (green triangle), the phase is estimated at the time of stimulation (blue vertical line). Stimulation may occur in the absence of an endogenous SO (top row) or during an endogenous SO upstate (middle row) or SO downstate (bottom row). In all cases, an evoked SO (pink triangle) and evoked spindle (blue curve) may occur after auditory stimulation. Evoked responses are compared to sham stimulation delivered on the same background. B) Auditory stimulation evokes a higher percentage of SOs and C) a higher percentage of SO-spindle complexes when delivered in the upstate of an endogenous SO compared to SHAM. The blue horizontal line indicates the null hypothesis of no difference in evoked SOs and spindles between STIM and SHAM.

When auditory stimulations were delivered at the time of endogenous SO, we evaluated the percentage of auditory stimulations that evoked a SO or SO-spindle complex compared to sham stimulations delivered at the same phase of an endogenous SO. Overall, when stimulations were delivered during the upstate of an endogenous SO, an evoked SO occurred 58.3% of the time (Control: 57.7%; RE: 59.0%) and an evoked SO-spindle complex occurred 32.1% of the time (Control: 32.0%; RE: 32.2%). Compared to sham stimulation, when auditory stimulations were delivered during the upstate of an endogenous SO, the proportion of evoked SOs was increased by 35.7% (95% CI [26.1–45.3], p<0.001); this was observed in each group (Control: 30.6% increase, 95% CI [16.6 44.5], p=0.002; RE: 40.2% increase, 95% CI [24.5 56.0], p<0.001, two-sided paired t-test; Figure 4B). Similarly, when auditory stimulation was delivered during the upstate of an endogenous SO, the proportion of evoked SO-spindle complexes was increased by 23.5% (95% CI [10.0–36.9], p=0.002) compared to sham stimulation; this was observed in each group (Control: 21.2% increase, 95% CI [4.3 38.1], p=0.02; RE: 25.4% increase, 95% CI [0.5 50.4], p=0.05, two-sided paired t-test; Figure 4C). SOs were maximally increased (51.3% increase, 95% CI [38.4–64.2], p<0.001 compared to sham) when auditory stimulation was delivered at the peak of an endogenous SO upstate (±30 degrees; Control: 45.1% increase, 95% CI [21.1–69.1], p=0.004; RE: 58.6% increase, 95% CI [45.1–72.0], p<0.001, two-sided paired t-test). Similarly, SO-spindle complexes were maximally increased (32.3% increase, 95% CI [17.1–47.5], p<0.001 compared to sham) when auditory stimulation was delivered near the peak of an endogenous SO upstate (±30 degrees; Control: 25.1% increase, 95% CI [3.8–46.5], p=0.03; RE: 40.7% increase, 95% CI [12.4–68.9], p=0.01, two-sided paired t-test). In contrast, there was no difference detected in the percentage of SOs and SO-spindle complexes evoked when auditory stimulations were delivered during the downstate of an endogenous SO compared to similarly timed sham stimulation (all p>0.49; Figure 4B–C). Time-frequency analysis further revealed that stimulation delivered during periods without detected SOs or during the SO upstate elicited significant increases in slow wave (~1–2 Hz) and spindle band (~12–15 Hz) power, which were not evident after stimulation during the SO downstate (Supplementary Figure 3). These findings indicate that auditory stimulation delivered during an upstate of a SO maximally evokes SOs and SO-spindle complexes, but also evokes SO and SO-spindle complexes at all other times except when delivered during a SO downstate.

3.5. Stimulation induced changes in frontal SO and SO-spindle complexes correlate positively with changes in memory consolidation

To evaluate the impact of auditory stimulation on memory consolidation, we evaluated homotopic frontal electrode pairs where the effect size of auditory stimulation was maximal (black outlined circles, Figure 3B). Across participants and visits, both SO rate (p<0.001, mean MST improvement of 0.8%, for each unit increase in SO per minute, 95% CI [0.4–1.3]) and SO-spindle complex rate (p=0.005, mean MST improvement of 5.1%, for each unit increase in SO-spindle complex per minute, 95% CI [1.6–8.5]) positively correlated with memory consolidation (Figure 5B). In post-hoc analysis, we found a positive trend between spindle rate and memory consolidation (p=0.10, mean MST improvement of 1.5% for each unit increase; Supplementary Figure 2B). We note that auditory stimulation was not expected to increase event rates since it was delivered randomly. Thus, to evaluate whether changes to neurophysiology during auditory stimulation impact memory, we compared the change in SO rate or SO-spindle complex rate between auditory and sham stimulation visits to the change in sleep-dependent memory consolidation between visits for each participant. We found a positive correlation between the change in SO rate and sleep-dependent memory consolidation (p<0.001, mean MST improvement of 1.9% for each unit increase in SO per minute, 95% CI [1.0–2.8]; Figure 5C). We also found a positive correlation between the change in SO-spindle complex rate and change in sleep-dependent memory consolidation (p=0.007, mean MST improvement of 9.5% for each unit increase in SO-spindle complex per minute, 95% CI [2.8–16.2]). Similarly, on post-hoc analysis, we found that changes in spindle rate during auditory stimulation showed a positive trend with changes in memory consolidation (p=0.11, mean MST improvement of 3.1% for each unit increase; Supplementary Figure 2C).

Figure 5.

Figure 5.

Open in a new tab

A) Changes in MST improvements are compared to changes in frontal sleep event rates between SHAM and STIM visits. Sleep features in the contralateral hemisphere are used to predict motor improvements. B) Across all participants and stimulation conditions, SO rate and SO-spindle coupling rates are positively correlated with sleep-dependent memory improvement C) Across all participants, changes in SO rate and SO-spindle complex rate during auditory stimulation positively correlated with changes in sleep-dependent memory improvement. Black (dashed) curves indicate estimated model fit (95% confidence interval).

Given the small size of our cohorts, we lacked power to evaluate each group separately. However, the change in SO rate was significantly associated with the change in sleep-dependent memory consolidation in RE (p=0.009, mean MST improvement of 2.0% for each unit increase in SO per minute, 95% CI [0.6–3.4]) and the effect size was positive but did not achieve significance in controls (effect size 2.4; p=0.1). Similarly, the change in SO-spindle complex rate was significantly associated with the change in sleep-dependent memory consolidation in controls (p<0.001, mean MST improvement of 24.3% for each unit increase in SO-spindle complex per minute, 95% CI [17.1–31.6]) but was not significant in the RE alone (effect size 0.7, p=0.9). Taken together, these results demonstrate that changes in SO and SO-spindle complex event rates caused by auditory stimulation predict changes in memory consolidation. We note that consistent results were observed using a larger amplitude (−60 μV) threshold to detect SO (−60 uV, Supplementary Figure 4).

4. Discussion

These findings demonstrate that auditory stimulation delivered during baseline activity or during the upstate of an endogenous SO during stages 2 and 3 NREM sleep reliably evokes SO and SO-spindle complexes activity in children with RE and control children. Evoked SOs and SO-spindle complexes are broadly distributed over the cortex, but the effect is maximal over the frontal electrodes near the midline. Frontal SO and SO-spindle complex rates positively predict sleep-dependent memory consolidation and changes in frontal SO and SO-spindle complex rates due to auditory stimulation positively predict changes in sleep-dependent memory consolidation. These results support the potential for closed-loop stimulation protocols that increase SO- and SO-spindle complex rates to enhance sleep-dependent memory consolidation.

Prior studies have demonstrated that auditory stimulation during stages 2 and 3 NREM sleep induces SOs and spindles (Baxter et al., 2023, Besedovsky et al., 2017, Leminen et al., 2017, Ngo et al., 2013, Ngo et al., 2019, Ong et al., 2016, Prehn-Kristensen et al., 2020). Prior work has also suggested that auditory stimulation timed to the upstate of an endogenous SO is more effective at evoking SOs and spindles (Batterink et al., 2016, Baxter et al., 2023, Cox et al., 2014, Henin et al., 2019, Leminen et al., 2017, Ngo et al., 2013, Ngo et al., 2019, Ong et al., 2016). Here we strengthened these findings with improved resolution of the proportional impact of stimulation at different phases of the SO and also provide comparison to sham stimulation at the same phases to control for background activities. Taken together, these findings confirm that phase-controlled auditory stimulation during sleep can evoke SOs and SO-spindle complex events. We also show that auditory stimulation evokes SO and SO-spindle complexes compared to sham when delivered on most background activity, except for during a SO downstate. Prior work has focused on the impact of auditory stimulation at the FZ or CZ electrodes (Baxter et al., 2023, Besedovsky et al., 2017, Henin et al., 2019, Leminen et al., 2017, Ngo et al., 2013, Ngo et al., 2019, Ong et al., 2016, Prehn-Kristensen et al., 2020). Here, we extended this work to identify the spatial extent of the evoked signals. These findings suggest that SOs can be evoked over widespread cortical regions, but SO-spindle complexes are indeed maximally evoked over frontal electrodes near the midline. As some studies suggest that spindle activity in different cortical regions support domain specific memory consolidation (Fernandez and Lüthi, 2020, Gais et al., 2002, Morin et al., 2008, Schmidt et al., 2006), this indicates that auditory stimulation may be more successful to enhance memory consolidation related to frontal lobe processing.

Although SO and SO-spindle complex event rates have been found to positively predict sleep dependent memory consolidation across multiple domains (Astill et al., 2014, Baxter et al., 2023, Hahn et al., 2019, Hahn et al., 2020, Muehlroth et al., 2019, Ng et al., 2024), most prior work evaluating the impact of auditory stimulation did not compare memory performance to SO or SO-spindle complex rates (Baxter et al., 2023, Harrington et al., 2021, Leminen et al., 2017, Prehn-Kristensen et al., 2020). When doing so, we found that the change in SO rate and SO-spindle complex event rate positively predicted the change in memory consolidation between the two visits. This is consistent with prior work indicating that sleep spindles in the frontal cortex predict memory consolidation of the motor sequence typing task (Jacobacci et al., 2020, Kwon et al., 2025, Spencer et al., 2022). Importantly, we found SO- and SO-spindle complexes are reliably evoked when stimulation is delivered during background activity in the absence of an endogenous slow oscillation or timed to the upstate of an endogenous slow oscillation. Across multiple studies, closed-loop auditory stimulation (CLAS) timed to the SO upstate, reliably induces SOs and spindles (Esfahani et al., 2023, Henao et al., 2020, Moreira et al., 2021, Navarrete et al., 2022, Ngo et al., 2013, Ngo et al., 2019, Ong et al., 2016, Papalambros et al., 2017, Weigenand et al., 2016). Some studies using CLAS have reported improved memory performance (Leminen et al., 2017, Ngo et al., 2013, Ngo et al., 2015, Ong et al., 2016, Papalambros et al., 2017). However, other experiments (Cox et al., 2014, Ngo et al., 2019, Weigenand et al., 2016), a validation study (Henin et al., 2019), and a recent meta-analysis (Wunderlin et al., 2021) indicate weak or no improvement in memory performance using traditional CLAS (Ngo et al., 2013). Importantly, current CLAS protocols are not designed to increase event rate, the metric known to drive memory consolidation (Gais et al., 2002, Hahn et al., 2019, Hahn et al., 2020, Kramer et al., 2021, Kurdziel et al., 2013, Kwon et al., 2025, McLaren et al., 2023, Schabus et al., 2004, Schabus et al., 2006, Tamminen et al., 2010). Here, using open loop auditory stimulation, we show that changes in SO and SO-spindle event rates between stimulation and sham (i.e., no sound) reliably predict changes in memory. Taken together, these findings indicate that stimulations delivered frequently, both on background activity and also timed to the upstate of endogenous SO may be required to maximally increase SO- and SO-spindle complexes event rates.

Children with RE have deficits in memory consolidation relative to control children. We were underpowered to analyze our two groups separately, however we saw similar qualitative results in response to stimulation between children with RE and controls. Although statistically significant associations with memory were not observed within each group alone, these findings establish proof of concept that auditory stimulation can modulate oscillations relevant to memory in both children with RE and controls, and the groundwork for future studies with larger samples. Additionally, the unbalanced visit order in our cohort could have introduced order effects. To address this, we included visit order as a covariate in our statistical models, ensuring that the observed results were not confounded by visit order. The selection of the threshold for SO detection is inherently arbitrary and dependent on the referencing approach. Children often exhibit larger SO amplitudes than adults. Previous studies have reported thresholds ranging from −40 to −80 μV across both children and adults (Baxter et al., 2023, Castelnovo et al., 2023, Ngo et al., 2013, Piantoni et al., 2013), with approximately −60 μV being typical for pediatric populations (Castelnovo et al., 2023). We evaluated two thresholds (−40 vs −60 μV) and found qualitatively consistent results.

Together, these findings contribute evidence supporting the role of sleep oscillations in sleep-dependent memory consolidation, and insights on the impact of a non-invasive method to optimally control these oscillations in children with RE and typically developing controls. These results raise the possibility that tailored auditory stimulation approaches could serve as scalable, non-invasive therapeutic interventions to support memory consolidation and mitigate the memory impairments associated with epilepsy (Kwon et al., 2025).

Supplementary Material

Supplementary Material

Highlights.

  • Auditory stimulation reliably evokes slow oscillations (SOs) and SO-spindle complexes but with maximal yield during SO upstate.

  • Auditory stimulation evokes SOs broadly over the cortex and SO–spindle complexes with maximal effect near the frontal vertex.

  • Changes in frontal SO and SO-spindle rates by stimulation positively predicted changes in sleep-dependent memory consolidation.

Acknowledgments

This work was supported by NINDS R01NS115868.

Footnotes

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This is the submitted version that was not yet certified by peer review.

References

  1. Proposal for revised classification of epilepsies and epileptic syndromes. Commission on Classification and Terminology of the International League Against Epilepsy. Epilepsia 1989;30(4):389–99. [DOI] [PubMed] [Google Scholar]
  2. Astill RG, Piantoni G, Raymann RJ, Vis JC, Coppens JE, Walker MP, et al. Sleep spindle and slow wave frequency reflect motor skill performance in primary school-age children. Frontiers in human neuroscience 2014;8:910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Batterink LJ, Creery JD, Paller KA. Phase of Spontaneous Slow Oscillations during Sleep Influences Memory-Related Processing of Auditory Cues. The Journal of neuroscience : the official journal of the Society for Neuroscience 2016;36(4):1401–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baxter BS, Mylonas D, Kwok KS, Talbot CE, Patel R, Zhu L, et al. The effects of closed-loop auditory stimulation on sleep oscillatory dynamics in relation to motor procedural memory consolidation. Sleep 2023;46(10). [Google Scholar]
  5. Bellesi M, Riedner BA, Garcia-Molina GN, Cirelli C, Tononi G. Enhancement of sleep slow waves: underlying mechanisms and practical consequences. Frontiers in systems neuroscience 2014;8:208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Besedovsky L, Ngo H-VV, Dimitrov S, Gassenmaier C, Lehmann R, Born J. Auditory closed-loop stimulation of EEG slow oscillations strengthens sleep and signs of its immune-supportive function. Nature Communications 2017;8(1):1984. [Google Scholar]
  7. Born J, Wilhelm I. System consolidation of memory during sleep. Psychol Res 2012;76(2):192–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Buzsaki G Hippocampal sharp wave-ripple: A cognitive biomarker for episodic memory and planning. Hippocampus 2015;25(10):1073–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Castelnovo A, Lividini A, Riedner BA, Avvenuti G, Jones SG, Miano S, et al. Origin, synchronization, and propagation of sleep slow waves in children. Neuroimage 2023;274:120133. [DOI] [PubMed] [Google Scholar]
  10. Cox R, Korjoukov I, de Boer M, Talamini LM. Sound asleep: processing and retention of slow oscillation phase-targeted stimuli. PloS one 2014;9(7):e101567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Davis ZW, Muller L, Martinez-Trujillo J, Sejnowski T, Reynolds JH. Spontaneous travelling cortical waves gate perception in behaving primates. Nature 2020;587(7834):432–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dudai Y, Karni A, Born J. The Consolidation and Transformation of Memory. Neuron 2015;88(1):20–32. [DOI] [PubMed] [Google Scholar]
  13. Esfahani MJ, Farboud S, Ngo HV, Schneider J, Weber FD, Talamini LM, et al. Closed-loop auditory stimulation of sleep slow oscillations: Basic principles and best practices. Neurosci Biobehav Rev 2023;153:105379. [DOI] [PubMed] [Google Scholar]
  14. Fattinger S, Heinzle BB, Ramantani G, Abela L, Schmitt B, Huber R. Closed-Loop Acoustic Stimulation During Sleep in Children With Epilepsy: A Hypothesis-Driven Novel Approach to Interact With Spike-Wave Activity and Pilot Data Assessing Feasibility. Frontiers in human neuroscience 2019;13. [Google Scholar]
  15. Fernandez LMJ, Lüthi A. Sleep Spindles: Mechanisms and Functions. Physiological Reviews 2020;100(2):805–68. [DOI] [PubMed] [Google Scholar]
  16. Fisher RS, Acevedo C, Arzimanoglou A, Bogacz A, Cross JH, Elger CE, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 2014;55(4):475–82. [DOI] [PubMed] [Google Scholar]
  17. Gais S, Molle M, Helms K, Born J. Learning-dependent increases in sleep spindle density. The Journal of neuroscience : the official journal of the Society for Neuroscience 2002;22(15):6830–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Geva-Sagiv M, Mankin EA, Eliashiv D, Epstein S, Cherry N, Kalender G, et al. Augmenting hippocampal–prefrontal neuronal synchrony during sleep enhances memory consolidation in humans. Nature Neuroscience 2023;26(6):1100–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Grigg-Damberger M, Gozal D, Marcus CL, Quan SF, Rosen CL, Chervin RD, et al. The visual scoring of sleep and arousal in infants and children. J Clin Sleep Med 2007;3(2):201–40. [PubMed] [Google Scholar]
  20. Hahn M, Joechner AK, Roell J, Schabus M, Heib DP, Gruber G, et al. Developmental changes of sleep spindles and their impact on sleep-dependent memory consolidation and general cognitive abilities: A longitudinal approach. Dev Sci 2019;22(1):e12706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hahn MA, Heib D, Schabus M, Hoedlmoser K, Helfrich RF. Slow oscillation-spindle coupling predicts enhanced memory formation from childhood to adolescence. Elife 2020;9. [Google Scholar]
  22. Harrington MO, Ngo HV, Cairney SA. No benefit of auditory closed-loop stimulation on memory for semantically-incongruent associations. Neurobiol Learn Mem 2021;183:107482. [DOI] [PubMed] [Google Scholar]
  23. Henao D, Navarrete M, Valderrama M, Le Van Quyen M. Entrainment and synchronization of brain oscillations to auditory stimulations. Neurosci Res 2020;156:271–8. [DOI] [PubMed] [Google Scholar]
  24. Henin S, Borges H, Shankar A, Sarac C, Melloni L, Friedman D, et al. Closed-Loop Acoustic Stimulation Enhances Sleep Oscillations But Not Memory Performance. eneuro 2019;6(6):ENEURO.0306–19.2019. [Google Scholar]
  25. Iber C, Medicine AAoS. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications: American Academy of Sleep Medicine, 2007. [Google Scholar]
  26. Jacobacci F, Armony JL, Yeffal A, Lerner G, Amaro E, Jovicich J, et al. Rapid hippocampal plasticity supports motor sequence learning. Proceedings of the National Academy of Sciences 2020;117(38):23898–903. [Google Scholar]
  27. Kleiner M, Brainard DH, Pelli D, Ingling A, Murray R, Broussard C. What’s new in Psychtoolbox-3. Perception 2007;36:1–16. [Google Scholar]
  28. Klinzing JG, Tashiro L, Ruf S, Wolff M, Born J, Ngo HV. Auditory stimulation during sleep suppresses spike activity in benign epilepsy with centrotemporal spikes. Cell Rep Med 2021;2(11):100432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kramer MA, Stoyell SM, Chinappen D, Ostrowski LM, Spencer ER, Morgan AK, et al. Focal Sleep Spindle Deficits Reveal Focal Thalamocortical Dysfunction and Predict Cognitive Deficits in Sleep Activated Developmental Epilepsy. The Journal of neuroscience : the official journal of the Society for Neuroscience 2021;41(8):1816–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kurdziel L, Duclos K, Spencer RMC. Sleep spindles in midday naps enhance learning in preschool children. Proceedings of the National Academy of Sciences 2013;110(43):17267–72. [Google Scholar]
  31. Kwon H, Chinappen DM, Kinard EA, Goodman SK, Huang JF, Berja ED, et al. Association of Sleep Spindle Rate With Memory Consolidation in Children With Rolandic Epilepsy. Neurology 2025;104(2):e210232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kwon H, Kronemer SI, Christison-Lagay KL, Khalaf A, Li J, Ding JZ, et al. Early cortical signals in visual stimulus detection. Neuroimage 2021;244:118608. [DOI] [PubMed] [Google Scholar]
  33. Kwon H, Walsh KG, Berja ED, Manoach DS, Eden UT, Kramer MA, et al. Sleep spindles in the healthy brain from birth through 18 years. Sleep 2023;46(4). [Google Scholar]
  34. Leminen MM, Virkkala J, Saure E, Paajanen T, Zee PC, Santostasi G, et al. Enhanced Memory Consolidation Via Automatic Sound Stimulation During Non-REM Sleep. Sleep 2017;40(3). [Google Scholar]
  35. Manoach DS, Cain MS, Vangel MG, Khurana A, Goff DC, Stickgold R. A failure of sleep-dependent procedural learning in chronic, medicated schizophrenia. Biological psychiatry 2004;56(12):951–6. [DOI] [PubMed] [Google Scholar]
  36. Manoach DS, Pan JQ, Purcell SM, Stickgold R. Reduced Sleep Spindles in Schizophrenia: A Treatable Endophenotype That Links Risk Genes to Impaired Cognition? Biological psychiatry 2016;80(8):599–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. McCormick DA, Bal T. Sleep and arousal: thalamocortical mechanisms. Annu Rev Neurosci 1997;20:185–215. [DOI] [PubMed] [Google Scholar]
  38. McLaren JR, Luo Y, Kwon H, Shi W, Kramer MA, Chu CJ. Preliminary evidence of a relationship between sleep spindles and treatment response in epileptic encephalopathy. Ann Clin Transl Neurol 2023;10(9):1513–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mölle M, Born J. Slow oscillations orchestrating fast oscillations and memory consolidation. Prog Brain Res 2011;193:93–110. [DOI] [PubMed] [Google Scholar]
  40. Mölle M, Marshall L, Gais S, Born J. Grouping of spindle activity during slow oscillations in human non-rapid eye movement sleep. The Journal of neuroscience : the official journal of the Society for Neuroscience 2002;22(24):10941–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Moreira CG, Baumann CR, Scandella M, Nemirovsky SI, Leach S, Huber R, et al. Closed-loop auditory stimulation method to modulate sleep slow waves and motor learning performance in rats. Elife 2021;10. [Google Scholar]
  42. Morin A, Doyon J, Dostie V, Barakat M, Hadj Tahar A, Korman M, et al. Motor sequence learning increases sleep spindles and fast frequencies in post-training sleep. Sleep 2008;31(8):1149–56. [PMC free article] [PubMed] [Google Scholar]
  43. Muehlroth BE, Sander MC, Fandakova Y, Grandy TH, Rasch B, Shing YL, et al. Precise Slow Oscillation–Spindle Coupling Promotes Memory Consolidation in Younger and Older Adults. Scientific reports 2019;9(1):1940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Navarrete M, Arthur S, Treder MS, Lewis PA. Ongoing neural oscillations predict the post-stimulus outcome of closed loop auditory stimulation during slow-wave sleep. Neuroimage 2022;253:119055. [DOI] [PubMed] [Google Scholar]
  45. Ng T, Noh E, Spencer RMC. Does slow oscillation-spindle coupling contribute to sleep-dependent memory consolidation? A Bayesian meta-analysis. eLife Sciences Publications, Ltd; 2024. [Google Scholar]
  46. Ngo HV, Martinetz T, Born J, Molle M. Auditory closed-loop stimulation of the sleep slow oscillation enhances memory. Neuron 2013;78(3):545–53. [DOI] [PubMed] [Google Scholar]
  47. Ngo HV, Miedema A, Faude I, Martinetz T, Mölle M, Born J. Driving sleep slow oscillations by auditory closed-loop stimulation-a self-limiting process. The Journal of neuroscience : the official journal of the Society for Neuroscience 2015;35(17):6630–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Ngo HV, Seibold M, Boche DC, Molle M, Born J. Insights on auditory closed-loop stimulation targeting sleep spindles in slow oscillation up-states. Journal of neuroscience methods 2019;316:117–24. [DOI] [PubMed] [Google Scholar]
  49. Nishida M, Walker MP. Daytime naps, motor memory consolidation and regionally specific sleep spindles. PloS one 2007;2(4):e341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Ong JL, Lo JC, Chee NI, Santostasi G, Paller KA, Zee PC, et al. Effects of phase-locked acoustic stimulation during a nap on EEG spectra and declarative memory consolidation. Sleep Med 2016;20:88–97. [DOI] [PubMed] [Google Scholar]
  51. Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: Open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data. Comput Intell Neurosci 2011;2011:156869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Papalambros NA, Santostasi G, Malkani RG, Braun R, Weintraub S, Paller KA, et al. Acoustic Enhancement of Sleep Slow Oscillations and Concomitant Memory Improvement in Older Adults. Frontiers in human neuroscience 2017;11:109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Piantoni G, Poil SS, Linkenkaer-Hansen K, Verweij IM, Ramautar JR, Van Someren EJ, et al. Individual differences in white matter diffusion affect sleep oscillations. The Journal of neuroscience : the official journal of the Society for Neuroscience 2013;33(1):227–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Prehn-Kristensen A, Ngo HV, Lentfer L, Berghäuser J, Brandes L, Schulze L, et al. Acoustic closed-loop stimulation during sleep improves consolidation of reward-related memory information in healthy children but not in children with attention-deficit hyperactivity disorder. Sleep 2020;43(8). [Google Scholar]
  55. Purcell SM, Manoach DS, Demanuele C, Cade BE, Mariani S, Cox R, et al. Characterizing sleep spindles in 11,630 individuals from the National Sleep Research Resource. Nat Commun 2017;8:15930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Ross EE, Stoyell SM, Kramer MA, Berg AT, Chu CJ. The natural history of seizures and neuropsychiatric symptoms in childhood epilepsy with centrotemporal spikes (CECTS). Epilepsy Behav 2020;103(Pt A):106437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Schabus M, Gruber G, Parapatics S, Sauter C, Klösch G, Anderer P, et al. Sleep spindles and their significance for declarative memory consolidation. Sleep 2004;27(8):1479–85. [DOI] [PubMed] [Google Scholar]
  58. Schabus M, Hödlmoser K, Gruber G, Sauter C, Anderer P, Klösch G, et al. Sleep spindle-related activity in the human EEG and its relation to general cognitive and learning abilities. The European journal of neuroscience 2006;23(7):1738–46. [DOI] [PubMed] [Google Scholar]
  59. Schmidt C, Peigneux P, Muto V, Schenkel M, Knoblauch V, Munch M, et al. Encoding difficulty promotes postlearning changes in sleep spindle activity during napping. The Journal of neuroscience : the official journal of the Society for Neuroscience 2006;26(35):8976–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Schreiner T, Petzka M, Staudigl T, Staresina BP. Endogenous memory reactivation during sleep in humans is clocked by slow oscillation-spindle complexes. Nat Commun 2021;12(1):3112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Seibt J, Richard CJ, Sigl-Glockner J, Takahashi N, Kaplan DI, Doron G, et al. Cortical dendritic activity correlates with spindle-rich oscillations during sleep in rodents. Nat Commun 2017;8(1):684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Spencer ER, Chinappen D, Emerton BC, Morgan AK, Hämäläinen MS, Manoach DS, et al. Source EEG reveals that Rolandic epilepsy is a regional epileptic encephalopathy. NeuroImage Clinical 2022;33:102956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Staresina BP, Bergmann TO, Bonnefond M, van der Meij R, Jensen O, Deuker L, et al. Hierarchical nesting of slow oscillations, spindles and ripples in the human hippocampus during sleep. Nat Neurosci 2015;18(11):1679–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Steriade M, Domich L, Oakson G, Deschenes M. The deafferented reticular thalamic nucleus generates spindle rhythmicity. J Neurophysiol 1987;57(1):260–73. [DOI] [PubMed] [Google Scholar]
  65. Steriade M, Nunez A, Amzica F. A novel slow (< 1 Hz) oscillation of neocortical neurons in vivo: depolarizing and hyperpolarizing components. The Journal of neuroscience : the official journal of the Society for Neuroscience 1993;13(8):3252–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Tamminen J, Payne JD, Stickgold R, Wamsley EJ, Gaskell MG. Sleep spindle activity is associated with the integration of new memories and existing knowledge. The Journal of neuroscience : the official journal of the Society for Neuroscience 2010;30(43):14356–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Tononi G, Cirelli C. Sleep and the price of plasticity: from synaptic and cellular homeostasis to memory consolidation and integration. Neuron 2014;81(1):12–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Wamsley EJ, Tucker MA, Shinn AK, Ono KE, McKinley SK, Ely AV, et al. Reduced sleep spindles and spindle coherence in schizophrenia: mechanisms of impaired memory consolidation? Biological psychiatry 2012;71(2):154–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Weigenand A, Molle M, Werner F, Martinetz T, Marshall L. Timing matters: open-loop stimulation does not improve overnight consolidation of word pairs in humans. The European journal of neuroscience 2016;44(6):2357–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Wickens S, Bowden SC, D’Souza W. Cognitive functioning in children with self-limited epilepsy with centrotemporal spikes: A systematic review and meta-analysis. Epilepsia 2017;58(10):1673–85. [DOI] [PubMed] [Google Scholar]
  71. Wodeyar A, Chinappen D, Mylonas D, Baxter B, Manoach DS, Eden UT, et al. Thalamic epileptic spikes disrupt sleep spindles in patients with epileptic encephalopathy. Brain : a journal of neurology 2024;147(8):2803–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Wunderlin M, Züst MA, Hertenstein E, Fehér KD, Schneider CL, Klöppel S, et al. Modulating overnight memory consolidation by acoustic stimulation during slow-wave sleep: a systematic review and meta-analysis. Sleep 2021;44(7). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material
NIHMS2129584-supplement-Supplementary_Material.pdf (778.5KB, pdf)

Data Availability Statement

Raw data were generated at Massachusetts General Hospital and the Athinoula A. Martinos Center for Biomedical Imaging. Derived data supporting the findings of this study are available from the corresponding author on request. The spindle detection method is available at https://github.com/Mark-Kramer/Spindle-Detector-Method.

RESOURCES