Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Mar 24;34(5):e70050. doi: 10.1111/jsr.70050

NREM Sleep Oscillations Are Associated With Anxiety and Negative Affect in Young Adults

Hazal Arpaci 1, Nandita Banik 1, Pinar Kurdoglu Ersoy 2, Ciara Harrington 1, Aycan Kapucu 2, Bengi Baran 1,3,4,✉
PMCID: PMC12353645  PMID: 40129180

ABSTRACT

Non‐rapid eye movement sleep (NREM) oscillations are critical for cognitive and affective processing. While several studies link anxiety and depression symptoms to sleep quality, a critical gap remains in elucidating the role of NREM physiology in sleep‐dependent processing of affect and anxiety symptoms. The goals of the present study were to investigate sleep‐dependent consolidation of emotional memory and the relations of NREM oscillations with state anxiety and affect upon awakening in a non‐clinical sample enriched for trait anxiety. Forty‐two participants were recruited from a larger cohort of college students based on self‐reported high (> 2 SD cohort mean, n = 26) versus moderate‐low levels of trait anxiety (< 2 SD cohort mean, n = 16) for a 2‐h polysomnography monitored mid‐day nap. Memory for negative and neutral picture stimuli was tested over this nap interval. Sleep spindles and slow oscillations (SOs) predicted post‐nap state anxiety and negative affect. Importantly, these were independent relationships in opposing directions such that higher SO activity was associated with reduced negative affect and state anxiety, whereas spindle activity correlated with higher negative affect and anxiety. We observed significantly reduced SO activity in the high‐anxiety group but no associations of anxiety with macro‐features of sleep (sleep duration, latency, efficiency or stage distributions). There were no group differences in emotional memory, nor did sleep parameters correlate with memory performance. These findings reflect that NREM oscillations are uniquely sensitive to both trait and state level variability in anxiety and highlight their potential as a novel target to attenuate anxiety and negative affect.

Keywords: anxiety, emotional memory, negative affect, sleep spindles, slow oscillations

1. Introduction

Sleep serves critical cognitive and affective functions including memory consolidation (Diekelmann and Born 2010; Stickgold and Walker 2013), emotional regulation (Goldstein and Walker 2014; Palmer and Alfano 2017) and decision making (Pace‐Schott et al. 2012). A well‐known consequence of sleep deprivation is disrupted emotional regulation. Sleep loss in adults is shown to disrupt mood, increase anxiety and introduce emotional biases in working memory and cognitive control (Gerhardsson et al. 2019; Gujar et al. 2011; Palmer et al. 2024). These studies of partial or complete sleep disruption highlight the importance of deep and continuous NREM sleep in mood, affect and anxiety (Chellappa and Aeschbach 2022; Finan et al. 2015), although the exact mechanism of how sleep improves emotional regulation is not known. Emerging work, for example, Stokes et al. (2023), identifies alterations in non‐rapid eye movement (NREM) sleep oscillatory activity as a biomarker and potentially a treatment target for psychiatric disorders. The primary goal of the present study was to investigate the effects of NREM sleep oscillations on anxiety and affect upon awakening in a non‐clinical sample of young adults enriched for heightened trait anxiety. We believe this is the critical first step in delineating causal links between NREM oscillations and emotional regulation and devising novel, sleep‐based interventions to alleviate psychopathology symptoms. The second goal was to examine the effects of heightened trait‐level anxiety on sleep architecture, NREM physiology and sleep‐dependent emotional memory consolidation.

A handful of studies have investigated the relationship between NREM physiology and anxiety symptoms. In healthy young adults, Stage 3 NREM sleep duration and slow wave spectral power (0.5–4 Hz, Delta) have been linked to lower levels of state anxiety the next morning (Ben Simon et al. 2020). In another study of 1083 adults, higher state and trait anxiety levels were associated with increased time in Stage 2 NREM sleep and decreased time in Stage 3 NREM sleep (Horváth et al. 2016). Furthermore, a metanalytic study on sleep quality and architecture revealed reduced Stage 3 NREM sleep duration in individuals with anxiety disorders (Baglioni et al. 2016). Children with social anxiety disorder (Wilhelm et al. 2017) or major depressive disorder (Lopez et al. 2010) are also shown to exhibit reduced sleep spindle activity.

Memory retrieval is better if encoding is followed by sleep compared to continuous wakefulness. This ‘sleep benefit’ is thought to rely on the temporal coordination of cardinal NREM sleep oscillations that allows the transfer of newly encoded memory traces from the hippocampus to more stable cortical representations (Rasch and Born 2013). Importantly, sleep‐dependent memory consolidation is thought to involve a triage that prioritises emotionality and future relevance (Stickgold and Walker 2013). Several studies report selective consolidation of emotional stimuli over neutral ones (Hu et al. 2006; Payne et al. 2008; Wagner et al. 2006) or simultaneous sleep‐dependent processing of both the memory and the emotional reactivity for studied material (Baran et al. 2012) and the sleep benefit for emotional memories has been linked to SO and sleep spindle activity (Halonen et al. 2023; Rodheim et al. 2023). Impairments in sleep‐dependent memory consolidation have been shown for several psychiatric disorders (Goerke et al. 2017).

Taken together, separate lines of evidence reveal that NREM sleep is critical for emotional processing and memory consolidation, and that the duration and physiological properties of NREM may be altered in anxiety disorders. However, the exact nature of the relationship between NREM physiology and anxiety symptoms is not known. The primary goal of the present study was to provide converging evidence on the effects of NREM sleep on anxiety and affect. Sleep‐dependent consolidation of emotional memory was probed over a midday nap in young adults with high versus low levels of trait anxiety. We hypothesised that individuals with heightened anxiety would have reduced NREM oscillatory activity and reduced emotional memory performance. Additionally, we examined whether the physiological properties of the nap following this stressful task that involved studying negative and arousing picture stimuli would predict affective ‘state’ (i.e., self‐reports of state anxiety and negative affect).

2. Methods

2.1. Participants

Forty‐two young adults (mean age:19.3 ± 1.8, range: 18–27 years) were recruited through introductory psychology courses at the University of Iowa. Students eligible for course credit completed an online screening form at the beginning of each semester that included the 10‐item trait anxiety measure from the State‐Trait Anxiety Inventory for Adults (STAI; Spielberger 1983). Individuals whose STAI‐trait scores were 2 SD below or above the mean of the entire subject pool of more than 200 students enrolled each semester were invited to participate in the present study (over two consecutive semesters) and were assigned to low (n = 16) versus high anxiety (n = 26) groups, respectively. Groups were matched in age and sex. While these screening data were collected outside of research activities and are not available for reporting, participants completed the same trait anxiety measure at the completion of the study, which revealed a significant group difference (Table 1). Additional inclusion criteria were (1) proficiency in English, (2) normal or corrected‐to‐normal vision, (3) body mass index ≤ 30, (4) no self‐reported history of neurological or sleep disorders, (5) no self‐reported history of psychiatric illness, (6) no history of head injury resulting in loss of consciousness or neurological sequelae, (7) no travel across time zones or shift work in the past 2 weeks, (8) typical total time in bed 6.5 – 9 h. All procedures were approved by the University of Iowa Institutional Review Board. Informed consent was obtained from all participants before experimental procedures started and participants were compensated with course credit or payment.

TABLE 1.

Participant characteristics and the comparison of emotional memory task performance, sleep architecture and post‐sleep mood and affect measures between the high‐ versus low‐anxiety groups.

Low anxiety (n = 16) High anxiety (n = 26) t(40) p Cohen's d
m SD m SD
Age 18.80 0.93 19.60 2.10 −1.32 0.21 0.45
Sex 62.5% female 77% female χ 2 = 0.56 0.76 0.84
Sleep quality (PSQI a ) 5.91 2.94 7.04 3.49 −1.65 0.11 0.52
Chronotype (MEQ b ) 44.94 8.95 46.73 8.95 −0.63 0.53 0.20
Self‐reported sleep duration (min) for the previous night c 484 113 484 81.20 0.01 0.99 0.004
Emotional memory task
AUC d negative 0.97 0.04 0.97 0.03 0.41 0.69 0.13
AUC neutral 0.97 0.03 0.94 0.15 0.92 0.37 0.31
Pre‐nap arousal ratings 5.12 1.33 5.88 1.37 −1.77 0.09 0.56
Pre‐nap valence ratings 2.68 0.56 2.28 0.66 2.03 0.049 0.65
Change in valence ratings −0.05 0.36 −0.02 0.53 −0.16 0.88 0.05
Change in arousal ratings −0.03 0.79 −0.11 1.05 0.29 0.78 0.09
Sleep architecture
Total sleep time (min) 101.72 15.74 99.35 20.96 0.39 0.70 0.12
WASO e (min) 5.91 5.54 7.25 9.24 −0.53 0.60 0.17
Sleep onset latency (min) 11.41 8.53 10.37 5.56 0.48 0.63 0.15
Sleep efficiency (%) 87.50 9.18 86.36 9.76 0.38 0.71 0.12
N1 (%) 22.88 16.93 31.09 20.19 −0.36 0.18 0.43
N2 (%) 53.70 17.08 55.78 17.64 −0.38 0.71 0.12
N3 (%) 9.36 11.86 5.23 9.43 1.25 0.22 0.41
REM (%) 14.06 12.46 7.89 11.88 1.61 0.12 0.51
N1 (min) 24.38 19.89 31.87 22.73 −0.09 0.28 0.35
N2 (min) 53.59 15.98 53.85 18.56 −0.05 0.96 0.01
N3 (min) 9.53 12.13 5.35 9.67 1.25 0.22 0.39
REM (min) 14.22 12.51 8.29 12.77 1.47 0.15 0.47
Post‐nap affect and mood
PANAS f positive 31.40 6.00 30.10 8.12 0.58 0.57 0.18
PANAS f negative 17.90 3.86 22.50 6.77 −2.51 0.02 0.81
State anxiety (STAI g ) 31.31 7.50 38.92 12.46 −2.20 0.03 0.70
Trait anxiety (STAI g ) 37.90 5.98 46.70 13.30 −2.49 0.02 0.79
QIDS h 5.25 3.79 9.58 5.39 −2.81 0.01 0.90
Open in a new tab
a

PSQI, Pittsburgh sleep quality index (Smyth 1999); a higher score indicates worse subjective sleep (range 0–21).

b

MEQ, Morningness‐Eveningness Questionnaire (Horne and Ostberg 1976); a higher score indicates a stronger morning preference (range 16–86).

c

Based on a brief self‐report sleep survey.

d

Area under the curve. Memory accuracy measure that reflects the ability to distinguish between signal and noise. Higher values indicate better memory accuracy (range 0–1).

e

Wake after sleep onset.

f

PANAS, positive and negative affect schedule (Watson et al. 1988); a higher score indicates higher endorsement of positive/negative affect (range 5–50).

g

STAI, State–Trait Anxiety Inventory (Spielberger 1983); a higher score indicates higher anxiety (range 20–80).

h

QIDS, quick inventory of depressive symptomatology (Rush et al. 2003); a higher score indicates higher depressive symptoms (range 0–27).

2.2. Experimental Design

Participants completed a polysomnography (PSG) monitored midday nap and an emotional memory task (Figure 1A). Study procedures started between 11 AM and 1 PM, based on participant availability. Participants were informed about the post‐sleep memory test before starting the encoding phase of the emotional memory task. This was followed by PSG setup, which lasted approximately 30 min. All participants had a two‐hour nap opportunity in a sound and light‐proofed lab bedroom with continuous PSG monitoring. Upon awakening, they were provided with snacks, completed self‐report questionnaires and watched videos of their choice from a list of non‐arousing TV sitcoms and documentaries. Delayed recognition was probed exactly 5 h after initial encoding.

FIGURE 1.

FIGURE 1

Open in a new tab

(A) Study design: Participants took a polysomnography monitored midday nap. Session started with completing the Encoding phase of an emotional memory task, followed by EEG capping and a 2‐h nap opportunity in a private bedroom in the sleep laboratory. Recognition memory was tested 5 h after encoding. Upon awakening participants completed anxiety, mood and affect scales and watched neutral videos for the remainder of the interval. (B) Emotional memory task: during the encoding phase, participants studied 46 negative and 46 neutral images (an exemplar is used in the figure) and completed emotionality ratings. The test phase involved making recognition memory decisions for the 92 target images intermingled with 92 foils using six‐point confidence scale (1 = definitely new, 6 = definitely old). Participants completed emotionality ratings again during this phase.

2.3. EEG Acquisition and Analysis

Naps were monitored with the LiveAmp 32 EEG acquisition system (Brain Vision, NC) and EasyCap EEG caps equipped with active electrodes (EasyCap GmbH, Herrsching, Germany) with 30 EEG electrodes referenced to linked mastoids, 2 electrooculography (EOG), and 2 submental electromyography (EMG) channels. Impedances were kept < 15 kΩ.

Sleep EEG was acquired with a sampling rate of 500 Hz. Data were bandpass filtered at 0.3–35 Hz and notch filtered at 60 Hz. Each 30 s epoch was scored as N1, N2, N3, REM, or wake by an expert sleep scorer according to standard criteria (Berry et al. 2015). The following sleep architecture variables were derived from the sleep scoring: total time in bed (up to 120 min for all participants), total sleep time, wake after sleep onset (WASO; defined as the total minutes spent awake after sleep onset); sleep efficiency (percentage of total sleep time over total time in bed), sleep onset latency (defined as three consecutive sleep epochs) and the duration and percentage of time spent in each sleep stage.

Data underwent visual examination by trained researchers. EEG channels displaying artifactual activity in more than 70% of the recording were interpolated with spherical splines. Subsequently, epochs with any remaining artefacts were removed. Sources of artefacts include major body movements, sweat and recording‐related noise. Groups did not differ in the total duration of artefact‐free NREM N2 and N3 sleep (t(40) = 0.69, p = 0.51, d = 0.22).

Custom scripts utilising MATLAB (Mathworks, Natick, MA) and EEGLab (Delorme and Makeig 2004) were used to detect NREM oscillations. These algorithms have been validated against visual detection and other automated detection methods in both healthy and clinical populations (Wamsley et al. 2012; Warby et al. 2014) and are described in greater detail in previous publications (Baran et al. 2019; Demanuele et al. 2017; Denis et al. 2024; Mölle and Born 2011; Mylonas, Baran, et al. 2020). Sleep spindles were detected in Sigma (12–15 Hz) band‐pass‐filtered data utilising an automated wavelet‐based algorithm with a peak frequency of 13.5 Hz. In order to maximise variance differences between spindle versus non‐spindle events (Otsu 1979), and in line with previous work cited above, the amplitude threshold for detecting a spindle event was nine times the median signal amplitude observed in epochs without artefacts, and the duration threshold was 0.4 s. The primary measures of spindle activity were spindle density (number of spindle events per minute of artefact‐free N2 and N3 sleep) and spindle amplitude (maximal voltage). Additional spindle characteristics included spindle duration (half‐height width of the wavelet energy) and spindle frequency (spectral peak of the spindle following Fast Fourier Transformation). To detect SOs, data were filtered in the Delta band (0.5–4 Hz). The threshold of duration for SO detection was 0.8–2 s, and the threshold for SO amplitude was set such that waveforms with amplitudes in the upper 25% were retained as meeting the SO criterion (minimum average amplitude of detected SOs per channel can be found in Table S1). The primary measures of SO activity were SO density (number of detected SOs per minute of artefact‐free N2 and N3 sleep) and SO peak‐to‐peak normalised amplitude. Next, we determined whether a spindle event reached peak amplitude during a detected SO event in order to calculate spindle‐SO coupling features. Sigma band‐filtered data were used to derive the instantaneous amplitude of spindles, and the Hilbert transform was applied to determine the SO phase at spindle peak. Primary measures of SO‐spindle coupling were the percentage of all spindles coupled with SO and coupling strength. Coupling strength corresponds to the consistency with which a spindle peaks at a given phase of the SO (i.e., the mean vector length of phase locking measured on a scale of 0–1).

Power spectral density (PSD) of N2 and N3 sleep across all EEG channels was calculated using the Welch method: data were segmented into overlapping 5‐s Hamming windows with 50% overlap. The temporal derivative of the data was used to minimise the 1/f scaling inherent to EEG signals. We extracted PSD values for Sigma (12–15 Hz) and slow wave activity (SWA; defined as 0.5–4 Hz, i.e., Delta). Both metrics were normalised by dividing by the average spectral power of each channel for each participant.

2.4. Self‐Report Measures

Participants completed the STAI (Spielberger 1983) and the Quick Inventory of Depressive Symptomatology (QIDS; Rush et al. 2003) in the afternoon upon awakening to quantify ‘in the moment’ anxiety and depression symptoms. Affect was quantified with the Positive and Negative Affect Schedule (PANAS; Watson et al. 1988). During this period, participants also filled out the Morningness‐Eveningness Questionnaire; MEQ (Horne and Ostberg 1976) and the Pittsburgh Sleep Quality Index; PSQI (Buysse et al. 1989). Self‐report questionnaires were either presented online on Qualtrics or using paper & pencil forms.

2.5. Emotional Memory Task

This intentional emotional memory task consisted of two phases: Encoding (pre‐nap) and Delayed Recognition (post‐nap; Figure 1B). Stimuli were 368 pictures selected from the Nencki Affective Picture System (NAPS; Marchewka et al. 2014) divided into 92 target (46 negative, 46 neutral) and 92 foil (46 negative, 46 neutral) picture sets. Pictures in each set were equally distributed among the following categories: faces, landscapes, animals and objects. Negative and neutral picture sets were matched in terms of visual complexity (width, height, luminance, contrast, size, colour composition and entropy; all p's > 0.11). Negative and neutral pictures selected for target and foil sets were matched on valence (all p's > 0.75) and arousal (all p's > 0.80). The memory task was developed and presented in OpenSesame (https://www.monash.edu/research‐portal/vlab/applications/opensesame).

During encoding, each target picture was presented in random order at the center of a computer screen for 3 s with a 150 ms inter‐stimulus interval. This was followed by valence (1 = extremely negative, 9 = extremely positive) and arousal (1 = calm/no arousal, 9 = excited/high arousal) ratings that the participants completed at their own pace using the nine‐item self‐assessment manikin (SAM) scales (Bradley and Lang 1994). Delayed Recognition started approximately 5 h after Encoding. Participants were instructed to decide whether a given picture was old or new and to make a confidence rating accordingly (1 = sure new to 6 = sure old). Pictures were presented randomly for 3 s, followed by self‐paced memory judgement and SAM valence and arousal scales. To ensure the participants understood the instructions, each session started with practice trials that did not count towards final performance.

Recognition memory analyses followed the signal detection theory (MacMillan and Creelman 2005). Specifically, memory accuracy was calculated as area under curve (AUC) based on each person's confidence ratings and memory bias was measured with ‘c’ separately for negative and neutral stimuli using the ROC MATLAB toolbox (Koen et al. 2017). Unlike other memory accuracy measures such as d′ or %correct, AUC does not rely on strong or unvalidated assumptions and is robust to response bias. Instead, it directly reflects the model's ability to discriminate between targets and lures (Rotello et al. 2008; Weidemann and Kahana 2016). For ‘c’ lower values reflect more liberal response biases (i.e., higher rates of ‘old’ responses to both targets and foils). These outcomes were calculated using all confidence levels without collapsing responses into binary categories.

2.6. Statistical Analyses

To investigate the associations of NREM oscillations with mood and affect, we conducted linear regressions at each electrode to predict mood and affect based on NREM oscillations. Spindle density, spindle amplitude, SO density and SO amplitude were chosen as predictors, while negative affect and state anxiety were chosen as dependent variables. Differences in NREM oscillatory measures between high‐ and low‐anxiety groups were assessed with independent samples t‐tests at each electrode. Additional linear regressions tested the associations of pre‐nap valence and arousal ratings for negative picture stimuli with NREM oscillations to determine whether the emotionality and arousal levels of these stimuli influenced NREM oscillations.

To correct for multiple comparisons, we implemented a nonparametric cluster‐based method that follows previous work (Mylonas, Baran, et al. 2020) and is based on the approach outlined in (Maris and Oostenveld 2007). Clusters were formed from neighbouring electrodes that surpassed an uncorrected threshold of p ≤ 0.05. Cluster statistics were calculated as the sum of the test statistics for all electrodes within the cluster. Permutation distributions were generated 1000 times for each electrode with random group assignment, and the cluster with the maximum statistic was retained for each permutation. Cluster‐corrected p values of < 0.05 were deemed significant. In line with methodological recommendations for EEG data (Meyer et al. 2021) and considering the size of our electrode arrays, we report maximum the effect within each cluster to enhance replicability of our findings. Effect sizes were calculated for each electrode within the cluster and the electrode with the largest value is reported as Cohen's d for group comparisons that used independent samples t‐tests and Cohen's f 2 for linear regression analyses.

Memory performance was compared with 2 × 2 ANOVAs with factors for group (high vs. low anxiety), valence (negative vs. neutral pictures) and their interaction. Sleep‐dependent changes in emotionality ratings (i.e., SAM valence and arousal scales) were compared with a 2 × 2 × 2 ANOVA with factors for group, valence, sleep (pre‐ vs. post‐nap) and their interactions. Effect sizes for these analyses were reported by calculating partial η 2.

3. Results

3.1. Participant Characteristics

Groups did not differ in chronotype (MEQ) or habitual sleep quality (PSQI), but the high‐anxiety group scored higher on measures of state and trait anxiety, depression and negative affect (Table 1). In the entire sample, state and trait anxiety levels were positively correlated (r = 0.75, p < 0.001).

3.2. Associations of NREM Oscillations With Anxiety and Negative Affect

All participants were able to nap on command, with an average of 86.8% sleep efficiency (Table 1). In the entire sample, higher SO density during the nap correlated significantly and negatively with subsequent state anxiety (STAI‐state, Figure 2A, seven electrodes, t sum = −18.74, p corrected = 0.046; T7: Cohen's f 2  = 0.24) and negative affect (PANAS, Figure 2B, seven electrodes, t sum = −17.98, p corrected = 0.045; CP5: f 2  = 0.20), reflecting that naps rich in SOs in a wide array of electrodes were associated with less anxiety and negative affect upon awakening. Similar relations were observed for SO amplitude and negative affect (Figure 2C, six electrodes, t sum = −17.41, p corrected = 0.04; Oz: f 2  = 0.36), with higher amplitude SOs being associated with reduced negative affect. These effects were specific to the detected SOs and did not generalise to SWA (i.e., delta power).

FIGURE 2.

FIGURE 2

Open in a new tab

Associations of anxiety and affect with NREM oscillations. (A) Slow oscillation (SO) density and (B) SO amplitude were inversely associated with negative affect, while (C) spindle density showed a positive association. Additionally, (D) SO density was inversely related to state anxiety, whereas (E) spindle amplitude was positively associated with state‐level anxiety symptoms. Left panels show the topographical maps of the correlation strength (Pearson's r). Warm colours represent positive correlations and blue represents negative correlations. Electrodes marked in grey correspond to p uncorrected < 0.05 and electrodes in pink surpass cluster‐level correction for multiple comparisons (p corrected < 0.05). Right panels show scatterplots of these relationships. The regression line is for the entire sample.

In stark contrast, sleep spindle density (Figure 2D, 13 electrodes, t sum = 31.57, p corrected = 0.02; FC1: f 2  = 0.23) and sigma power (Figure S1, 14 electrodes, t sum = 35.05, p corrected = 0.03; FC5: f 2  = 0.25) were positively associated with subsequent negative affect, reflecting that naps rich in detected sleep spindle oscillations and higher sigma power were associated with higher negative affect upon awakening in the entire sample. Similarly, higher spindle amplitude correlated with higher levels of state anxiety in the afternoon (Figure 2E, 13 electrodes, t sum = 31.23, p corrected = 0.03; F7: f 2  = 0.25).

Next, we investigated whether these opposing relations of SO and spindles with anxiety and affect are independent. Adding both SO and spindle densities as predictors to a model that predicts post‐sleep negative affect confirmed that each effect remains significant (SO density: β = −2.67, t(36) = −2.42, p = 0.02, f 2  = 0.16); spindle density: (β = 1.47, t(36) = 2.45, p = 0.02, f 2  = 0.17) and these two predictors do not show multicollinearity (variance inflation factor [VIF] for both predictors was 1.02). A similar model with SO density and spindle amplitude as predictors of state anxiety revealed that the effects of SO density on reduced anxiety remained significant (β = −4.72, t(36) = −2.20, p = 0.03, f 2  = 0.14) while the positive relations between spindle amplitude and anxiety did not survive (β = 0.34, t(36) = 1.88, p = 0.07, f 2  = 0.10); nevertheless, with a VIF of 1.11 indicating no multicollinearity. Altogether, these analyses demonstrate that NREM SO and spindles had independent effects on state anxiety and negative affect in the entire sample. Importantly, all the reported associations remained significant after controlling for artefact‐free minutes of NREM sleep and as reported above, state anxiety and negative affect were not associated with sleep efficiency or onset latency (all p's > 0.63), reflecting that the effects on state anxiety and affect were not due to sleep continuity or quality. There were no associations of NREM oscillations or spectral power measures with trait anxiety or depressive symptoms.

3.3. Group Differences in Sleep Quality, Architecture and NREM Oscillations

While a clinical comparison of sleep physiology in individuals with anxiety disorders versus controls was not within the scope of this study, we explored whether there are meaningful differences in sleep macro and microstructure between young adults with high versus moderate‐to‐low levels of trait anxiety. Sleep onset latency, sleep efficiency, total sleep time or WASO were not different between the high‐ versus low‐anxiety groups (Table 1) and did not correlate with state or trait measures of anxiety (all p's > 0.09) reflecting that the experimental demands and the novelty of sleeping in a lab environment did not differentially impact the macro features of sleep in anxious individuals. Sleep stage distributions of these naps also did not differ between groups (Table 1).

Slow oscillation density was significantly reduced in the high‐anxiety group (Figure 3A, 12 electrodes, t sum = −30.29, p corrected = 0.04; channel Oz: Cohen's d = 1.00). While there was a similar reduction in SO amplitude in the high‐anxiety group, this difference did not surpass correction for multiple comparisons (Figure 3B, six electrodes, t sum = −13.28, p corrected = 0.07, FC6: d = 0.78). Beyond detected SOs, we also observed a significant reduction in SWA in the high‐anxiety group (Figure 3C, 21 electrodes, t sum = −49.31, p corrected = 0.04, O2: d = 1.03). There were no group differences in the density, amplitude, or other characteristics of sleep spindles (Figure S2), or in the percentage of sleep spindles coupled with slow oscillations or coupling strength (Figure S3).

FIGURE 3.

FIGURE 3

Open in a new tab

Group differences in NREM oscillations. Topographical maps of (A) slow oscillation (SO) density, defined as number of detected oscillations per minute of artefact free N2 and N3 sleep, (B) SO amplitude, (C) slow wave activity (SWA; power spectral density at 0.5–4 Hz) in low (left column) versus high (middle column) Anxiety groups. Warm colours represent higher density or amplitude. Topographical maps on the right column represent p values of the group differences with lighter colour corresponding to lower p values. Electrodes marked in grey correspond to p uncorrected < 0.05 and electrodes in pink surpass cluster‐level correction for multiple comparisons (p corrected < 0.05).

3.4. Emotional Memory Task

Memory accuracy (AUC, Table 1 and Figures S4A–5) did not differ by group (F(1,38) = 0.89, p = 0.35, partial η 2  = 0.02) or valence category (F(1,38) = 0.44, p = 0.51, partial η 2  < 0.01). Similarly, response bias (c) also did not differ by group (F(1,38) = 0.93, p = 0.34, partial η 2  = 0.02) or valence category (F(1,38) = 0.59, p = 0.45, partial η 2  < 0.02) (Figure S4B). Pre‐nap valence and arousal ratings or over‐nap change in these ratings did not differ between high‐ versus low‐anxiety groups (all p's > 0.1; see Table 1 for negative stimuli and Figure S4C–F for both negative and neutral stimuli).

There was an effect of the nap interval on SAM ratings such that valence (F(1,40) = 31.50, p < 0.001, partial η 2  = 0.44) and arousal (F(1,40) = 17.96, p < 0.001, partial η 2  = 0.31) ratings for negative pictures decreased significantly in all participants, reflecting that negative pictures were rated as less emotional following a nap. There were no group differences or interactions with group (all p's > 0.13).

In exploratory analyses, we examined whether anxiety and negative affect were associated with task performance. Trait anxiety correlated with post‐nap valence ratings for negative stimuli (r = −0.33, p = 0.04) in the entire sample, such that those with higher levels of trait anxiety rated pictures as more negative after the nap. There were no significant relations of anxiety or negative affect with pre‐nap emotionality ratings or nap‐related changes in emotionality ratings. Response bias (c) for negative stimuli correlated with state anxiety (r = −0.32, p = 0.048) such that higher post‐nap state anxiety in the entire sample was associated with adopting a more liberal response bias (i.e., tendency to respond “old” for both old and new stimuli) during the recognition task. There were no other significant relations of memory with anxiety or affect.

3.5. Associations of Task With Sleep Oscillations

There were no significant relations of spindle density or amplitude, SO density or amplitude, spindle–SO coupling, sigma or SWA with memory accuracy, memory bias, nap‐dependent changes in emotionality ratings or post‐nap emotionality ratings (all p corrected > 0.12). Pre‐nap emotionality ratings for the negative pictures were associated with NREM oscillations during subsequent sleep. In the entire sample, SO density correlated significantly and negatively with both arousal (Figure 4A, 23 electrodes, t sum = −61.83, p corrected = 0.01; O1 f 2  = 0.32) and valence (Figure 4B, 28 electrodes, t sum = 71.60, p corrected = 0.01; P8: f 2  = 0.29) such that participants who rated these negative images as more emotional during their first encounter subsequently had fewer SOs during sleep. These relations remain significant when controlling for trait anxiety levels or the duration of artefact‐free minutes of NREM sleep. Sleep efficiency or sleep onset latency did not correlate with pre‐nap emotionality ratings (all p's > 0.49), reflecting that the emotionality of the memory task was associated with SO density but not the duration, quality or continuity of the subsequent nap.

FIGURE 4.

FIGURE 4

Open in a new tab

Associations of task emotionality ratings with NREM oscillations during subsequent sleep. Emotionality ratings for the pictures participants studied during the encoding phase of the memory task were associated with slow oscillation (SO) density during the subsequent nap. (A) Pre‐nap emotionality ratings for the arousal level of stimuli (higher ratings = more arousing) was negatively associated SO density, (B) pre‐nap emotionality ratings for the valence level of stimuli (lower ratings = more negative) were positively associated SO density. Left panels show the topographical maps of the correlation strength (Pearson's r). Warm colours represent positive correlations and blue represents negative correlations. Electrodes marked in grey correspond to p uncorrected < 0.05 and electrodes in pink surpass cluster‐level correction for multiple comparisons (p corrected < 0.05). Right panels show scatterplots of these relationships. The regression line is for the entire sample.

4. Discussion

In a non‐clinical sample enriched for the presence of anxiety, we demonstrate links between NREM oscillations and subsequent state anxiety and negative affect. We identified individuals with high versus moderate‐to‐low trait anxiety among a large, non‐help‐seeking sample of young adults (i.e., college freshmen taking an introductory psychology course). Participants were provided with a 2‐h midday nap opportunity following an emotional memory task that involved encoding arousing negative pictures intermingled with neutral, non‐arousing pictures. The amplitude and density of SO detected during Stages 2 and 3 NREM sleep were associated with subsequently reduced state anxiety and reduced negative affect upon awakening. The comparison of sleep physiology between the high versus moderate‐to‐low trait anxiety groups revealed significant differences in NREM slow waves in the context of no differences in macro features of sleep (i.e., sleep duration, sleep onset latency or sleep stage distributions). The amplitude and density of detected SOs as well as SWA were significantly reduced in the high‐anxiety group. Further, both these group differences and the associations of NREM oscillations and affect and anxiety emerged in a wide array of electrodes, not localised to a specific topography. Habitual sleep quality or chronotype quantified with self‐report scales were similar between high‐ versus low‐anxiety groups. These findings highlight that NREM sleep oscillations that play a critical role in cognition and emotional regulation may be altered in individuals with elevated trait anxiety that would go undetected with standard sleep measurements. Importantly, our results suggest that SO activity may be a biomarker for trait anxiety and is sensitive to state anxiety and negative affect upon awakening. These novel findings warrant a more systematic investigation of NREM oscillations in individuals diagnosed with anxiety disorders.

Our finding of the correlations between slow oscillatory activity and reduced state anxiety and negative affect lends support to an emerging literature on the effects of NREM oscillations on emotional regulation. For instance, a previous study has demonstrated a similar negative relation of nocturnal sleep slow wave duration and SWA with morning anxiety in young adults, leading these researchers to conclude that SWA plays an anxiolytic role (Ben Simon et al. 2020). Critically, we observed this relationship only for slow oscillation density and amplitude and not SWA broadly. While data from neither group are sufficient to attribute causality, our findings highlight the links between NREM oscillations and anxiety upon awakening. Considering that our participants completed an emotional memory task prior to the nap, we examined whether the perceived emotionality of these negative stimuli was associated with NREM oscillations. In fact, individuals who rated the pictures as more negative and more arousing, in turn, had reduced SO activity during subsequent sleep. These associations of SO activity with pre‐sleep emotionality ratings and post‐sleep anxiety and affect lead us to speculate that SO activity is a potential target to enhance emotional regulation. Non‐invasive brain stimulation methods have been successfully used during sleep to mimic the electrophysiological properties of the slow wave upstate and downstates (Fehér et al. 2021). Closed‐loop approaches that time the delivery of transcranial alternating current stimulation (TACS) or auditory stimulation to naturally occurring SOs have been shown to enhance sleep‐dependent memory consolidation in healthy adults (Jones et al. 2023; Marshall et al. 2006; Ngo et al. 2013). As also proposed by other researchers (Chellappa and Aeschbach 2022), similar non‐invasive approaches that enhance slow waves could potentially reduce anxiety but, to our knowledge, have not been used to test changes in affect or mood symptoms.

In contrast to this putative emotion regulatory effect of SO, higher spindle density and amplitude were associated with increased negative affect and state anxiety. While seemingly controversial, these findings are in line with a recent study that demonstrates increased spindle activity following a lab‐based stress manipulation in adults with PTSD in the context of similar sleep duration across stress and control conditions (Natraj et al. 2023). Speculatively, this positive relationship between spindles and post‐nap negative affect may be the byproduct of an emotion regulation process triggered by pre‐nap exposure to negative stimuli. However, a recent study revealed that both the anticipation of post‐sleep stress task and performing a pre‐sleep stress task (compared to a relaxation condition) resulted in reduced sleep spindles and SWA in the stress conditions (Beck et al. 2023). While seemingly contradictory, these reductions in NREM oscillatory activity were observed in the context of reduced sleep time, increased sleep onset latency and reduced N3. Nevertheless, these findings suggest that the relations of NREM oscillations with negative or stressful episodes are complex, warranting future work to elucidate these dynamics. Further, we found that the effects of SO and spindles on state anxiety and negative affect were statistically independent from each other, suggesting that these are independent affective processes. Our finding of reduced SO density in individuals who rated the negative pictures as more emotional lends support to the speculation that task exposure may have created a stressor and triggered a sleep‐dependent emotional regulatory process.

Sleep spindles and their temporal coordination with other NREM oscillations are thought to play a causal role in memory consolidation (Inostroza and Born 2013), and optogenetic manipulations in animal models provide evidence for this (Latchoumane et al. 2017). However, the exact mechanistic role of the coordination of NREM oscillations on emotional memories is not fully understood. While previous work demonstrates both negative (Denis et al. 2022) and positive (Rodheim et al. 2023) correlations between spindle‐SO coupling and recognition memory for emotional memory in young adults, we did not find any associations of memory performance with spindles, SO or their temporal coordination. Beyond methodological differences, for example, Denis et al. (2022) exposed participants to a stress manipulation prior to encoding and Rodheim et al. (2023) presented negative and positively valenced stimuli together, recognition accuracy in our sample was particularly high, hindering our ability to detect relations of sleep with individual variability in memory performance.

In summary, our study demonstrates a specific reduction in SO in individuals with heightened trait anxiety and identifies a sleep‐dependent emotional regulation function that is driven by NREM oscillations. However, these novel findings come with several limitations. First, we examined these oscillations during a midday nap. While previous work established that spindle activity is a stable and heritable trait that can be evaluated with a short midday nap in typical and clinical samples (Mylonas, Tocci, et al. 2020), no such reliability study compared SO activity during naps and nocturnal sleep. Second, while we identified and recruited our sample based on trait anxiety, that scale was completed outside of our experimental protocol as part of a college course experiential activity and was not available to researchers to report in this manuscript. Nevertheless, participants completed the same trait anxiety measure at the end of the study, which confirmed the group difference. Considering that recruitment was through a college campus, this limits the generalisability of our findings to other sociodemographic groups. Further, we did not measure anxiety and affect before starting the emotional memory task and before sleep and did not include a wake control group, which limits our interpretation that the post‐nap anxiety and affect was influenced by the task and subsequent sleep. Another potential limitation is the utilisation of a continuous variable (i.e., STAI trait anxiety scores) for binary group assignment and the unequal sample size across groups. The high sleep efficiency we observed in our sample may reflect a self‐selection bias with respect to flexibility of schedules and propensity for napping. Finally, ceiling‐level memory performance in our sample hindered our ability to investigate the relationship between NREM oscillations and sleep‐dependent memory consolidation and the effect of trait anxiety.

Author Contributions

Hazal Arpaci: formal analysis, writing – original draft, visualization, methodology, data curation. Nandita Banik: investigation, formal analysis, project administration, writing – review and editing, data curation. Pinar Kurdoglu Ersoy: writing – review and editing, methodology, formal analysis, data curation. Ciara Harrington: writing – review and editing, investigation. Aycan Kapucu: writing – review and editing, methodology, conceptualization. Bengi Baran: conceptualization, investigation, funding acquisition, writing – original draft, writing – review and editing, methodology, formal analysis, supervision, resources.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1.Supporting Information.

JSR-34-e70050-s001.docx (1.3MB, docx)

Acknowledgements

This work was supported by the National Institute of Mental Health (K01MH114012) to Bengi Baran. Hazal Arpaci is the recipient of fellowship support through the National Institute of General Medical Sciences predoctoral training grants T32GM108540 and T32GM149386 (MPIs: Lutgendorf/Voss/Tranel). The authors would like to thank Dr. Teresa Treat for statistical consultation, Valentine Soto for help with sleep EEG scoring and Pinar Emirahmetoglu for data collection.

Arpaci, H. , Banik N., Ersoy P. K., Harrington C., Kapucu A., and Baran B.. 2025. “ NREM Sleep Oscillations Are Associated With Anxiety and Negative Affect in Young Adults.” Journal of Sleep Research 34, no. 5: e70050. 10.1111/jsr.70050.

Funding: This work was supported by the National Institute of Mental Health (K01MH114012) to Bengi Baran. Hazal Arpaci is the recipient of fellowship support through the National Institute of General Medical Sciences (Grants T32GM108540 and T32GM149386) (MPIs: Lutgendorf/Voss/Tranel).

The first two authors contributed equally to this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. Baglioni, C. , Nanovska S., Regen W., et al. 2016. “Sleep and Mental Disorders: A Meta‐Analysis of Polysomnographic Research.” Psychological Bulletin 142, no. 9: 969–990. 10.1037/bul0000053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baran, B. , Karahanoğlu F. I., Mylonas D., et al. 2019. “Increased Thalamocortical Connectivity in Schizophrenia Correlates With Sleep Spindle Deficits: Evidence for a Common Pathophysiology.” Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 4, no. 8: 706–714. 10.1016/j.bpsc.2019.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baran, B. , Pace‐Schott E. F., Ericson C., and Spencer R. M. C.. 2012. “Processing of Emotional Reactivity and Emotional Memory Over Sleep.” Journal of Neuroscience 32, no. 3: 1035–1042. 10.1523/JNEUROSCI.2532-11.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beck, J. , Loretz E., and Rasch B.. 2023. “Stress Dynamically Reduces Sleep Depth: Temporal Proximity to the Stressor Is Crucial.” Cerebral Cortex 33, no. 1: 96–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ben Simon, E. , Rossi A., Harvey A. G., and Walker M. P.. 2020. “Overanxious and Underslept.” Nature Human Behaviour 4, no. 1: 100–110. 10.1038/s41562-019-0754-8. [DOI] [PubMed] [Google Scholar]
  6. Berry, R. B. , Brooks R., Gamaldo C., et al. 2015. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. American Academy of Sleep Medicine. www.aasmnet.org. [Google Scholar]
  7. Bradley, M. M. , and Lang P. J.. 1994. “Measuring Emotion: The Self‐Assessment Manikin and the Semantic Differential.” Journal of Behavior Therapy and Experimental Psychiatry 25, no. 1: 49–59. 10.1016/0005-7916(94)90063-9. [DOI] [PubMed] [Google Scholar]
  8. Buysse, D. J. , Reynolds C. F. 3rd, Monk T. H., Berman S. R., and Kupfer D. J.. 1989. “The Pittsburgh Sleep Quality Index: A New Instrument for Psychiatric Practice and Research.” Psychiatry Research 28, no. 2: 193–213. 10.1016/0165-1781(89)90047-4. [DOI] [PubMed] [Google Scholar]
  9. Chellappa, S. L. , and Aeschbach D.. 2022. “Sleep and Anxiety: From Mechanisms to Interventions.” Sleep Medicine Reviews 61: 101583. 10.1016/j.smrv.2021.101583. [DOI] [PubMed] [Google Scholar]
  10. Delorme, A. , and Makeig S.. 2004. “EEGLAB: An Open Source Toolbox for Analysis of Single‐Trial EEG Dynamics Including Independent Component Analysis.” Journal of Neuroscience Methods 134, no. 1: 9–21. 10.1016/j.jneumeth.2003.10.009. [DOI] [PubMed] [Google Scholar]
  11. Demanuele, C. , Bartsch U., Baran B., et al. 2017. “Coordination of Slow Waves With Sleep Spindles Predicts Sleep‐Dependent Memory Consolidation in Schizophrenia.” Sleep 40, no. 1: zsw013. 10.1093/sleep/zsw013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Denis, D. , Baran B., Mylonas D., et al. 2024. “NREM Sleep Oscillations and Their Relations With Sleep‐Dependent Memory Consolidation in Early Course Psychosis and First‐Degree Relatives.” Schizophrenia Research 274: 473–485. 10.1016/j.schres.2024.10.026. [DOI] [PubMed] [Google Scholar]
  13. Denis, D. , Kim S. Y., Kark S. M., Daley R. T., Kensinger E. A., and Payne J. D.. 2022. “Slow Oscillation‐Spindle Coupling Is Negatively Associated With Emotional Memory Formation Following Stress.” European Journal of Neuroscience 55, no. 9–10: 2632–2650. 10.1111/ejn.15132. [DOI] [PubMed] [Google Scholar]
  14. Diekelmann, S. , and Born J.. 2010. “The Memory Function of Sleep.” Nature Reviews. Neuroscience 11, no. 2: 114–126. 10.1038/nrn2762. [DOI] [PubMed] [Google Scholar]
  15. Fehér, K. D. , Wunderlin M., Maier J. G., et al. 2021. “Shaping the Slow Waves of Sleep: A Systematic and Integrative Review of Sleep Slow Wave Modulation in Humans Using Non‐invasive Brain Stimulation.” Sleep Medicine Reviews 58: 101438. 10.1016/j.smrv.2021.101438. [DOI] [PubMed] [Google Scholar]
  16. Finan, P. H. , Quartana P. J., and Smith M. T.. 2015. “The Effects of Sleep Continuity Disruption on Positive Mood and Sleep Architecture in Healthy Adults.” Sleep 38, no. 11: 1735–1742. 10.5665/sleep.5154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gerhardsson, A. , Åkerstedt T., Axelsson J., Fischer H., Lekander M., and Schwarz J.. 2019. “Effect of Sleep Deprivation on Emotional Working Memory.” Journal of Sleep Research 28, no. 1: e12744. 10.1111/jsr.12744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Goerke, M. , Müller N. G., and Cohrs S.. 2017. “Sleep‐Dependent Memory Consolidation and Its Implications for Psychiatry.” Journal of Neural Transmission (Vienna) 124, no. Suppl 1: 163–178. 10.1007/s00702-015-1476-3. [DOI] [PubMed] [Google Scholar]
  19. Goldstein, A. N. , and Walker M. P.. 2014. “The Role of Sleep in Emotional Brain Function.” Annual Review of Clinical Psychology 10: 679–708. 10.1146/annurev-clinpsy-032813-153716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gujar, N. , Yoo S. S., Hu P., and Walker M. P.. 2011. “Sleep Deprivation Amplifies Reactivity of Brain Reward Networks, Biasing the Appraisal of Positive Emotional Experiences.” Journal of Neuroscience 31, no. 12: 4466–4474. 10.1523/jneurosci.3220-10.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Halonen, R. , Luokkala S., Kuula L., Antila M., and Pesonen A. K.. 2023. “Right‐Lateralized Sleep Spindles Are Associated With Neutral Over Emotional Bias in Picture Recognition: An Overnight Study.” Cognitive, Affective, & Behavioral Neuroscience 23, no. 5: 1445–1459. 10.3758/s13415-023-01113-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Horne, J. A. , and Ostberg O.. 1976. “A Self‐Assessment Questionnaire to Determine Morningness‐Eveningness in Human Circadian Rhythms.” International Journal of Chronobiology 4, no. 2: 97–110. https://www.ncbi.nlm.nih.gov/pubmed/1027738. [PubMed] [Google Scholar]
  23. Horváth, A. , Montana X., Lanquart J.‐P., et al. 2016. “Effects of State and Trait Anxiety on Sleep Structure: A Polysomnographic Study in 1083 Subjects.” Psychiatry Research 244: 279–283. [DOI] [PubMed] [Google Scholar]
  24. Hu, P. , Stylos‐Allan M., and Walker M. P.. 2006. “Sleep Facilitates Consolidation of Emotional Declarative Memory.” Psychological Science 17, no. 10: 891–898. 10.1111/j.1467-9280.2006.01799.x. [DOI] [PubMed] [Google Scholar]
  25. Inostroza, M. , and Born J.. 2013. “Sleep for Preserving and Transforming Episodic Memory.” Annual Review of Neuroscience 36: 79–102. 10.1146/annurev-neuro-062012-170429. [DOI] [PubMed] [Google Scholar]
  26. Jones, A. P. , Bryant N. B., Robert B. M., et al. 2023. “Closed‐Loop tACS Delivered During Slow‐Wave Sleep Reduces Retroactive Interference on a Paired‐Associates Learning Task.” Brain Sciences 13, no. 3: 468. 10.3390/brainsci13030468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Koen, J. D. , Barrett F. S., Harlow I. M., and Yonelinas A. P.. 2017. “The ROC Toolbox: A Toolbox for Analyzing Receiver‐Operating Characteristics Derived From Confidence Ratings.” Behavior Research Methods 49, no. 4: 1399–1406. 10.3758/s13428-016-0796-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Latchoumane, C. V. , Ngo H. V., Born J., and Shin H. S.. 2017. “Thalamic Spindles Promote Memory Formation During Sleep Through Triple Phase‐Locking of Cortical, Thalamic, and Hippocampal Rhythms.” Neuron 95, no. 2: 424–435. 10.1016/j.neuron.2017.06.025. [DOI] [PubMed] [Google Scholar]
  29. Lopez, J. , Hoffmann R., and Armitage R.. 2010. “Reduced Sleep Spindle Activity in Early‐Onset and Elevated Risk for Depression.” Journal of the American Academy of Child & Adolescent Psychiatry 49, no. 9: 934–943. 10.1016/j.jaac.2010.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. MacMillan, N. A. , and Creelman C. D.. 2005. Detection Theory: A User's Guide. Cambridge University Press. [Google Scholar]
  31. Marchewka, A. , Zurawski Ł., Jednoróg K., and Grabowska A.. 2014. “The Nencki Affective Picture System (NAPS): Introduction to a Novel, Standardized, Wide‐Range, High‐Quality, Realistic Picture Database.” Behavior Research Methods 46, no. 2: 596–610. 10.3758/s13428-013-0379-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Maris, E. , and Oostenveld R.. 2007. “Nonparametric Statistical Testing of EEG‐ and MEG‐Data.” Journal of Neuroscience Methods 164, no. 1: 177–190. 10.1016/j.jneumeth.2007.03.024. [DOI] [PubMed] [Google Scholar]
  33. Marshall, L. , Helgadóttir H., Mölle M., and Born J.. 2006. “Boosting Slow Oscillations During Sleep Potentiates Memory.” Nature 444, no. 7119: 610–613. 10.1038/nature05278. [DOI] [PubMed] [Google Scholar]
  34. Meyer, M. , Lamers D., Kayhan E., Hunnius S., and Oostenveld R.. 2021. “Enhancing Reproducibility in Developmental EEG Research: BIDS, Cluster‐Based Permutation Tests, and Effect Sizes.” Developmental Cognitive Neuroscience 52: 101036. 10.1016/j.dcn.2021.101036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mölle, M. , and Born J.. 2011. “Slow Oscillations Orchestrating Fast Oscillations and Memory Consolidation.” Progress in Brain Research 193: 93–110. 10.1016/b978-0-444-53839-0.00007-7. [DOI] [PubMed] [Google Scholar]
  36. Mylonas, D. , Baran B., Demanuele C., et al. 2020. “The Effects of Eszopiclone on Sleep Spindles and Memory Consolidation in Schizophrenia: A Randomized Clinical Trial.” Neuropsychopharmacology 45, no. 13: 2189–2197. 10.1038/s41386-020-00833-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mylonas, D. , Tocci C., Coon W. G., et al. 2020. “Naps Reliably Estimate Nocturnal Sleep Spindle Density in Health and Schizophrenia.” Journal of Sleep Research 29, no. 5: e12968. 10.1111/jsr.12968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Natraj, N. , Neylan T. C., Yack L. M., et al. 2023. “Sleep Spindles Favor Emotion Regulation Over Memory Consolidation of Stressors in Posttraumatic Stress Disorder.” Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 8, no. 9: 899–908. 10.1016/j.bpsc.2023.02.007. [DOI] [PubMed] [Google Scholar]
  39. Ngo, H. V. , Martinetz T., Born J., and Mölle M.. 2013. “Auditory Closed‐Loop Stimulation of the Sleep Slow Oscillation Enhances Memory.” Neuron 78, no. 3: 545–553. 10.1016/j.neuron.2013.03.006. [DOI] [PubMed] [Google Scholar]
  40. Otsu, N. 1979. Time Frequency and Wavelets in Biomedical Signal Processing. Wiley‐IEEE Press. [Google Scholar]
  41. Pace‐Schott, E. F. , Nave G., Morgan A., and Spencer R. M.. 2012. “Sleep‐Dependent Modulation of Affectively Guided Decision‐Making.” Journal of Sleep Research 21, no. 1: 30–39. 10.1111/j.1365-2869.2011.00921.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Palmer, C. A. , and Alfano C. A.. 2017. “Sleep and Emotion Regulation: An Organizing, Integrative Review.” Sleep Medicine Reviews 31: 6–16. 10.1016/j.smrv.2015.12.006. [DOI] [PubMed] [Google Scholar]
  43. Palmer, C. A. , Bower J. L., Cho K. W., et al. 2024. “Sleep Loss and Emotion: A Systematic Review and Meta‐Analysis of Over 50 Years of Experimental Research.” Psychological Bulletin 150, no. 4: 440–463. 10.1037/bul0000410. [DOI] [PubMed] [Google Scholar]
  44. Payne, J. D. , Stickgold R., Swanberg K., and Kensinger E. A.. 2008. “Sleep Preferentially Enhances Memory for Emotional Components of Scenes.” Psychological Science 19, no. 8: 781–788. 10.1111/j.1467-9280.2008.02157.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Rasch, B. , and Born J.. 2013. “About sleep's Role in Memory.” Physiological Reviews 93, no. 2: 681–766. 10.1152/physrev.00032.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Rodheim, K. , Kainec K., Noh E., Jones B., and Spencer R. M. C.. 2023. “Emotional Memory Consolidation During Sleep Is Associated With Slow Oscillation‐Spindle Coupling Strength in Young and Older Adults.” Learning & Memory 30, no. 9: 237–244. 10.1101/lm.053685.122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Rotello, C. M. , Masson M. E., and Verde M. F.. 2008. “Type I Error Rates and Power Analyses for Single‐Point Sensitivity Measures.” Perception & Psychophysics 70, no. 2: 389–401. 10.3758/pp.70.2.389. [DOI] [PubMed] [Google Scholar]
  48. Rush, A. J. , Trivedi M. H., Ibrahim H. M., et al. 2003. “The 16‐Item Quick Inventory of Depressive Symptomatology (QIDS), clinician Rating (QIDS‐C), and Self‐Report (QIDS‐SR): A Psychometric Evaluation in Patients With Chronic Major Depression.” Biological Psychiatry 54, no. 5: 573–583. 10.1016/s0006-3223(02)01866-8. [DOI] [PubMed] [Google Scholar]
  49. Smyth, C. 1999. “The Pittsburgh Sleep Quality Index (PSQI).” Journal of Gerontological Nursing 25, no. 12: 10–11. 10.3928/0098-9134-19991201-10. [DOI] [PubMed] [Google Scholar]
  50. Spielberger, C. D. 1983. State‐Trait Anxiety Inventory for Adults. Sampler Set. Mind Garden. [Google Scholar]
  51. Stickgold, R. , and Walker M. P.. 2013. “Sleep‐Dependent Memory Triage: Evolving Generalization Through Selective Processing.” Nature Neuroscience 16, no. 2: 139–145. 10.1038/nn.3303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Stokes, P. A. , Rath P., Possidente T., et al. 2023. “Transient Oscillation Dynamics During Sleep Provide a Robust Basis for Electroencephalographic Phenotyping and Biomarker Identification.” Sleep 46, no. 1: zsac223. 10.1093/sleep/zsac223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wagner, U. , Hallschmid M., Rasch B., and Born J.. 2006. “Brief Sleep After Learning Keeps Emotional Memories Alive for Years.” Biological Psychiatry 60, no. 7: 788–790. 10.1016/j.biopsych.2006.03.061. [DOI] [PubMed] [Google Scholar]
  54. Wamsley, E. J. , Tucker M. A., Shinn A. K., et al. 2012. “Reduced Sleep Spindles and Spindle Coherence in Schizophrenia: Mechanisms of Impaired Memory Consolidation?” Biological Psychiatry 71, no. 2: 154–161. 10.1016/j.biopsych.2011.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Warby, S. C. , Wendt S. L., Welinder P., et al. 2014. “Sleep‐Spindle Detection: Crowdsourcing and Evaluating Performance of Experts, Non‐experts and Automated Methods.” Nature Methods 11, no. 4: 385–392. 10.1038/nmeth.2855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Watson, D. , Clark L. A., and Tellegen A.. 1988. “Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS Scales.” Journal of Personality and Social Psychology 54, no. 6: 1063–1070. 10.1037//0022-3514.54.6.1063. [DOI] [PubMed] [Google Scholar]
  57. Weidemann, C. T. , and Kahana M. J.. 2016. “Assessing Recognition Memory Using Confidence Ratings and Response Times.” Royal Society Open Science 3, no. 4: 150670. 10.1098/rsos.150670. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wilhelm, I. , Groch S., Preiss A., Walitza S., and Huber R.. 2017. “Widespread Reduction in Sleep Spindle Activity in Socially Anxious Children and Adolescents.” Journal of Psychiatric Research 88: 47–55. 10.1016/j.jpsychires.2016.12.018. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Data S1.Supporting Information.

JSR-34-e70050-s001.docx (1.3MB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Journal of Sleep Research are provided here courtesy of Wiley

RESOURCES