Article Figures & Data
Figures
- Figure 1.
An overview of the study process and the suppression protocol. During the experimental night, sleep was disrupted when participants entered the sleep stage of their respective suppression condition (REMSSUPPR / SWSSUPPR). In all playbacks, participants' own singing was played to them via loudspeakers, and physiological and subjective stress responses were measured. In memory retrievals, a subset of encoded metaphors was tested with cued recall.
- Figure 2.
Quantifying SO–spindle coupling. I, Detecting SOs. II, Detecting spindles that peak within the SO cycle. III, After Hilbert transformation, extracting the SO phase at the instant of the spindle peak. IV, Averaging the phase of all SO–spindles and calculating the number of spindles that peak within ±0.5 radian from the mean phase.
- Figure 3.
Skin conductance, suppression condition, and REMS oscillations. The playback of the karaoke recording induced a significant rise in SCL in all measurements (p < 0.001, Bonferroni-corrected; A). Violin plots illustrate individual (gray lines) and condition-wise (blue/orange) distributed SCRs with mean values (black/white dot) across the three measurements (B). Overnight change in SCR was higher in SWSSUPPR than in REMSSUPPR (p = 0.016, Bonferroni-corrected; C). Individual trajectories of SCRs in REMSSUPPR (D) and in SWSSUPPR (E). The scatterplot illustrates the positive association between REMS theta power spectral density (PSD) energy and overnight SCR change (p = 0.002, Bonferroni-corrected), and the heatmap displays Pearson’s correlation strength across 0–30 Hz for REMS (F). Error bars and dashed lines: 95% confidence interval. ***p < 0.001; **p < 0.01; *p < 0.05; Bonferroni-corrected. Extended Data Figure 3-1 displays the excluded outliers.
- Figure 4.
Subjective embarrassment. A violin plot illustrating individual (gray lines) and condition-wise (blue/orange) distributed embarrassment ratings with mean values (black/white dot) across the three measurements.
- Figure 5.
Metaphor recall. A violin plot illustrates individual (gray lines) and condition-wise (blue/orange) distributed metaphor recall percentages with mean values (black/white dot) across the three measurements (A). N2 fast spindle density predicted positively overall metaphor recall (averaged across all retrievals; p = 0.046, Bonferroni-corrected) (B). Dashed lines: 95% confidence interval.
Tables
Mean (SD) p REMSSUPPR (n = 15) SWSSUPPR (n = 14) Age (y) 26.3 (4.4) 25.3 (4.8) 0.571 BDI-II 6.0 (5.7) 6.4 (6.5) 0.877 GAD-7 3.6 (3.4) 3.5 (3.3) 0.936 TOSCA 3 Shame 42.3 (11.4) 41.6 (13.2) 0.892 Education 2.3 (0.8) 2.6 (0.9) 0.348 Singing experience 2.3 (0.9) 3.1 (0.9) 0.015* Sleep characteristics TST baseline night 7:05 (0:38) 7:13 (0:31) 0.539 TST experimental 6:04 (0:27) 6:10 (0:19) 0.548 N1 0:42 (0:16) 0:32 (0:12) 0.052 N2 3:03 (0:30) 3:27 (0:30) 0.039* SWS 1:36 (0:29) 0:48 (0:21) <0.001*** REMS 0:43 (0:19) 1:23 (0:18) <0.001*** WASO 0:53 (0:21) 0:49 (0:20) 0.548 REMS Fragmentation % 42.4 (21.3) 7.1 (8.1) <0.001*** SWS fragmentation % 13.5 (9.0) 39.2 (17.3) <0.001*** Forced awakenings, N 10.1 (3.3) 12.7 (2.9) 0.031* Awakening strength 1.9 (0.2) 1.6 (0.3) 0.002** REMS power spectral density energy SO delta 122.2 (42.5) 206.2 (35.2) <0.001*** Theta 104.5 (56.4) 219.4 (49.1) <0.001*** Alpha 109.3 (54.7) 227.8 (51.9) <0.001*** Sigma 143.1 (62.9) 247.4 (49.0) <0.001*** Beta 1 125.3 (48.9) 292.7 (54.0) <0.001*** Beta 2 125.3 (37.5) 252.4 (71.5) <0.001*** SWS power spectral density energy SO delta 311.8 (161.2) 157.8 (92.4) 0.004** Theta 275.6 (133.6) 160.8 (88.1) 0.012* Alpha 277.7 (193.1) 165.2 (91.9) 0.058 Sigma 353.4 (183.4) 207.8 (102.1) 0.014* Beta 1 287.1 (131.9) 182.2 (84.8) 0.018** Beta 2 307.1 (124.7) 192.3 (87.5) 0.008** SWS SO–spindle Fast spindle N 373.5 (142.7) 190.5 (115.7) <0.001*** Fast spindle density 4.0 (1.6) 4.2 (1.7) 0.762 SO–spindle N 82.4 (32.7) 39.1 (25.0) <0.001*** Resultant vector length 0.53 (0.10) 0.55 (0.19) 0.709 N2 SO–spindle Fast spindle N 927.1 (341.1) 1,037.9 (312.2) 0.371 Fast spindle density 5.0 (1.2) 5.0 (1.1) 0.947 SO–spindle N 137.2 (41.9) 149.9 (48.9) 0.458 Resultant vector length 0.37 (0.17) 0.41 (0.20) 0.564 BDI, Beck Depression Inventory; GAD-7, Generalized Anxiety Disorder 7 questionnaire; TOSCA, Test of Self-Conscious Affect; TST, total sleep time; SWS, slow-wave sleep; N1-2, sleep stage 1–2; REMS, rapid eye movement sleep; WASO, wake after sleep onset; SO, slow oscillation.
↵*** <0.001.
↵** <0.01.
↵* <0.05.
- Table 2.
LMM results: emotional response by condition and power spectral density energy values
SCR Subjective embarrassment Main effect X time Main effect X time F p F p F p F p Condition 0.091 0.766 4.209 0.027* 0.216 0.646 2.056 0.148 REMS PSD SO delta 0.010 1.000 4.191 0.165 0.080 1.000 2.548 0.581 Theta 0.019 1.000 7.002 0.023* 0.069 1.000 3.658 0.236 Alpha 0.833 1.000 3.325 0.314 0.249 1.000 2.002 0.927 Sigma 0.316 1.000 1.575 1.000 0.252 1.000 1.635 1.000 Beta1 0.152 1.000 1.062 1.000 0.219 1.000 2.518 0.596 Beta2 0.115 1.000 0.597 1.000 0.299 1.000 2.412 0.652 SWS PSD SO delta 0.017 1.000 0.135 1.000 2.025 1.000 1.589 1.000 Theta 0.004 1.000 0.065 1.000 1.673 1.000 1.878 1.000 Alpha 0.000 1.000 0.217 1.000 1.020 1.000 1.491 1.000 Sigma 0.008 1.000 0.115 1.000 2.266 0.869 2.818 0.468 Beta1 0.005 1.000 0.325 1.000 2.698 0.678 3.872 0.202 Beta2 0.065 1.000 0.397 1.000 2.992 0.576 4.840 0.098 X time, interaction between the examined variable and time (three measurements); REMS, rapid eye movement sleep; SWS, slow-wave sleep; PSD, power spectral density. REMS/SWS PSD p values Bonferroni-corrected across six tests.
↵* p < 0.05. See Extended Data Table 2-1 for nonparametric tests including outliers.
Figure 3-1
Download Figure 3-1, TIF file.
Table 2-1
The effects of suppression condition and power spectral density values on emotional response values/change in non-parametric statistics. Download Table 2-1, DOC file.
Data 1
Download Data 1, ZIP file.
In this issue
