Table 2 Included evidence in SR with EEG waves and measurement units

1

Armitage (1995); EEG

Delta (0.5– < 4 Hz)

Theta (4– < 8 Hz)

Alpha (8– < 12 Hz)

• Delta, theta, and alpha power (µV²); absolute value of the area of the half-wave squared by frequency category

2

Armitage et al. (2000); EEG

Delta (0.5– < 4 Hz)

• Delta power (µV²); derived from power spectral analysis which generated a vector of data describing power in the delta band

• Delta amplitude (µV²); half-wave zero-cross analyses; cumulative squared voltage of all points in the delta zero-cross bin using period-amplitude analysis

3

Baker et al. (2012); EEG

Delta (0.3– < 4 Hz)

Theta (4– < 8 Hz)

Alpha (8– < 12 Hz)

Sigma (12– < 15 Hz)

Beta 1 (15– < 23 Hz)

• Power (µV²⋅s); derived from power spectral analysis

• Amplitude (µV); zero-cross measures using period-amplitude analysis

Power spectra and period amplitude values

were averaged across NREM sleep (Stages 2–4) and REM

sleep separately for the entire night

4

Campbell et al. (2005); EEG

Delta (0.3–3 Hz)

3–4 Hz

Theta (4–6 Hz; 6–8 Hz)

Alpha (8–12 Hz)

Sigma (12–15 Hz)

Beta (15–23 Hz; 23–30 Hz)

30–50 Hz

*Frequencies inferred from study Table 4 and text in results section

• Power density (µV²⋅sec); power per minute of NREM

• Delta amplitude (µV); half-wave detection by zero crossing using period-amplitude analysis

5

Campbell et al. (2012); EEG

Delta (1–4 Hz)

Theta (4–8 Hz)

• Theta and delta power (µV²); spectral analysis

• Theta and delta power decline (µV2); wave power at each semiannual recording plotted against age for each male and each female person. The delta power decline across adolescence was fit with a Gompertz equation:

\(Power=D-A{\bullet e}^{{-e}^{-C(age-M)}}\)

• Delta amplitude (NR): from the upper to lower asymptotes of the delta curve

6

Carrier et al. (2001); EEG

0.25–32z bins were collapsed into 1–Hz frequency ranges (1 Hz: 0.25–1.00 Hz, 2 Hz: 1.25–2.00 Hz, 3 Hz: 2.25–3.00 Hz, etc.)

• Spectral power density (log transformed); power densities were calculated on the 128 Hz signals for consecutive 4-s epochs and 0.25 Hz frequency bandwidths by a Fast Fourier Transform routine. Awake time was excluded based on visual scoring in 60-s epochs. Artifact-laden 4-s epochs were identified by automated procedures. Average power density was calculated as the mean of the artifact-free 4-s sleep epochs for each N-REM period. For comparison of power spectral densities, the 0.25-Hz bins were collapsed into 1-Hz frequency ranges (~ 1 Hz: 0.25–1.00 Hz, 2 Hz:1.25–2.00 Hz, 3 Hz: 2.25–3.00 Hz, etc.)

7

Carrier et al. (2011); EEG

1–4 Hz

 < 4 Hz

*Focuses on slow wave; inferred as Delta

• SW amplitude (µV); difference in voltage between negative and positive peak of unfiltered signal expressed in microvolt

• SW slope (µV/s); slope between negative peak and positive peak

8

Dijk et al. (1989); EEG

0.25–15 Hz

*Frequencies bands inferred from commonly used categorizations

• Power density, expressed as a percentage (females as % of the average value in males); spectral analysis

9

Dorokhov et al. (2024); EEG

Delta (1–4 Hz)

Theta (5–8 Hz)

Alpha (9–12 Hz)

Sigma (13–16 Hz)

• Spectral power density (ln-transformed); calculated from EEG signals recorded at Fz, F4, Cz, Pz, and O2, referenced to M1/M2. Artifact-free epochs were processed using Fast Fourier Transform. Analysis focused on the first 16 frequency bands, from 1 to 16 Hz (e.g., 0.50–1.49 Hz for 1 Hz, 1.50–2.49 Hz for 2 Hz, etc.)

10

Feinberg et al. (2006); EEG

Delta (0.3–3 Hz)

1–4 Hz

• Delta power density (µV²sec/1000); power in 0.3–3 Hz divided by minutes of artifact-free NREM

11

Fukuda et al. (1999); EEG

Delta (0.5–2 Hz, 2–4 Hz)

• Delta spectral power; calculated by Fast Fourier Transform over 1024 data points with a cosine bell window

12

Hejazi et al. (2024); EEG

Delta (0.5–4 Hz):

Low Delta (0.5–2 Hz)

High Delta (02–4 Hz)

• Delta power spectral density (µV²/Hz); analyzed from C3-A2 and C4-A1 signals (filtered 0.5–30 Hz) during NREM using Welch’s periodogram (0.2 Hz and 1 Hz bins, six 5-s epochs, Hamming window) and a Chebyshev Type II filter

13

Kluge et al. (2010); EEG

Delta (0.5–4 Hz)

Theta (4.5–8 Hz)

Alpha (8.5–12 Hz)

Sigma (12.5–16 Hz)

Beta (16–20 Hz)

• Delta, theta, alpha, sigma, and beta power (µV²); Fast Fourier Transform routine using a rectangular window for consecutive, non-overlapping 2 s miniepochs

14

Latta et al. (2005); EEG

Delta (0.5–4 Hz)

Alpha (8.5–12 Hz)

• Delta and alpha activity (µV²); absolute spectral power in the 0.5–4 Hz and 8.5–12 Hz frequency bands, respectively

• Total power (µV²); calculated over the frequency range 0.5 to 12.5 Hz

• Relative delta and alpha activity (%); percentage of the total power for each 30-s epoch

15

Luo et al. (2024); EEG

Delta (0.5–4 Hz),

Theta (4–8 Hz),

Alpha (8–13 Hz),

Beta (13–30 Hz),

Spindle (11–16 Hz),

Sawtooth (3–7 Hz)

• Functional connectivity, quantified using mutual information (MI); MI evaluated both amplitude and phase information between EEG signals, classified as segments using 0.5–30 Hz band-pass filtering

16

Ma et al. (2011); EEG via PSG

Delta (0.5–4.25 Hz)

Theta (4.25–8 Hz)

Alpha (8–12 Hz)

Sigma (12–15 Hz)

Beta (15–32.5 Hz)

Spindle (11–14 Hz)

• Delta, theta, alpha, sigma, beta, spindle spectral power density, presented as a ratio (females as % difference to males)

17

Markovic et al. (2020); EEG

Delta (1–4.6 Hz)

Theta (4.8–7.8 Hz)

Alpha (8–10.8 Hz)

Sigma (11–16 Hz)

Beta 1 (16.2–20 Hz)

Beta 2 (20.2–24 Hz)

Gamma 1 (24.2–34 Hz)

Gamma 2 (34.2–44 Hz)

Spindle (10–16 Hz):

Slow spindle (10–12 Hz)

Fast spindle (12–16 Hz)

• Absolute power (µV²); calculated per epoch using 5-s windows, Hanning window, no overlap

• Normalized power (µV²); power at each derivation normalized by the total power across derivations

• Spindle amplitude (µV); not defined

• Spindle density (num/s); not defined

• Connectivity, measured with coherence; index of brain connectivity during sleep reflecting interactions between spatially segregated populations of neurons. Calculated between all possible channel pairs (i.e., 1653 connections) as (|P_xy (f)|²)/P_xx (f)P_yy (f), where P_xy(f) is the cross-spectral density and P_xx(f) and P_yy(f) are the auto-spectral density functions of the two

18

Mongrain et al. (2005); EEG via PSG

Delta (0.75–4 Hz)

Theta (4–8 Hz)

Low Sigma (12–14 Hz)

High Sigma (14–16 Hz)

• Spectral power; Fourier transforms performed on 4-s artifact-free sections using a cosine window tapering

19

Mourtazaev et al. (1995); EEG via PSG

Slow Wave (0.5–2 Hz)

*Focuses on slow wave; inferred as Delta

• Slow wave power (µV²); manual and computer-estimated based on the band-pass filter (0.5–2.0 Hz) described by Kemp and Lopes da Silva

20

Pun et al. (2023); EEG via PSG

Spindle (11–16 Hz)

• Spindle density (# events per min); calculated by counting spindles in NREM 2 & 3 separately from the central and frontal electrodes, then dividing by the time spent in each stage. Used computerized detection algorithm (11–16 Hz band-pass)

21

Ricci et al. (2021); EEG

Sigma (11–16 Hz)

Spindle (10–16 Hz)

Fast Spindle (12–16 Hz)

• Sigma power (µV²); computed by summing the power density data (including lower and upper limits of the frequency band), adjusting for rejected epochs, averaging for C3 and C4

• Spindle density (spindles/min); total number of spindles in N2 sleep and divided by the time in minutes of N2 sleep

• Spindle power (µV²); highest power S within the spindle

• Fast spindle percent (%); percent of fast spindles out of all identified spindles

• Percent change in spindle/sigma activity: calculated as a percent change with the formula [(follow-up value—baseline value)/baseline value]*100. Linear models wee used to calculate age-related percent change as a function of groups: 12–14 years, 15–17 years, and 18–22 years

22

Ringli et al. (2013); EEG

Slow Wave (0.75–4.5 Hz)*

*Focuses on slow wave; inferred as Delta

• Power, measured as slow wave activity (µV²); calculated as mean power in the range of 1–4.5 Hz during the first 60 min of NREM sleep stages 2 and 3

23

Rosinvil et al. (2021); EEG via PSG

Slow Wave (PtP ≥ 75 μV and NegA ≥ 40 μV, negative to positive deflection ≥ 125 to ≤ 1,500 ≤ 1,000 ms)

*Focuses on slow wave; inferred as Delta

• Slow wave density (number of slow waves per minute of NREM sleep, N2–N3 combined)

• Slow wave peak-to-peak amplitude (µV); difference in voltage between negative and positive peak of filtered signal, averaged over all-night NREM with N2–N3 combined

• Slow wave slope (µV/sec); ratio between the peak-to-peak amplitude and the delay between the two peaks

24

Ujma et al. (2019); EEG via PSG

Sigma (11–15.75 Hz)

Alpha (8–10.75 Hz)

Beta (16–25 Hz)

Gamma (25.25-48 Hz)

• Functional connectivity, measured as weighted phase-lag index (WPLI); calculated between all possible electrode pairs from the resulting time–frequency data and finally averaged across all data segments and electrode pairs belonging to the same cluster, respectively

25

Ujma et al. (2022); EEG

Beta (16–30 Hz)

Low Sigma (10–12.5 Hz)

High Sigma (12.5–16 Hz)

(0.01 Hz–4 Hz w 0.01 Hz increments)

• Amplitude; not defined

26

Ventura et al. (2022); EEG

Spindle (~ 11–15 Hz)

• Spindle spectral power, median (µV²); calculated from power spectral density estimates of each spindle

• Spindle density (spindles/min); number of sleep spindles per minute of NREM

27

Yoon et al. (2021); EEG via PSG

Delta (0.75–4 Hz)

Theta (4–8 Hz)

Alpha (8–12 Hz)

Sigma (12–14 Hz)

Beta (14–30 Hz)

• Absolute spectral power (µV², ln transformed); calculated the spectral band power for each 30-s epoch using Fast Fourier Transform with a Hanning window (3-s sliding window with 50% overlap). Spectral power was calculated for the whole night of sleep, REM and NREM sleep, and each stage of NREM including N1, N2 and N3

• Relative spectral power (%); calculated for each frequency band (delta [0.75–4 Hz], theta [4–8 Hz], alpha [8–12 Hz], sigma [12–14 Hz] and beta [14–30 Hz]) by dividing the power of each band by the total EEG power (0.75–30 Hz) at C4 electrode

28

Yuksel et al. (2021); EEG

Delta (0.3–4 Hz)

Theta (> 4–8 Hz)

Alpha (> 8–12 Hz)

Sigma (> 12–15 Hz)

Low Beta (> 15–23 Hz)

High Beta (> 23–30 Hz)

• Relative power (ratios); calculated for each frequency band as a function of the EEG total power (1–30 Hz), and then averaged across C3 and C4

29

Zhang et al. (2021); EEG

Spindle (12–15 Hz)

• Spindle amplitude (µV); calculated the peak-trough amplitude as the averaged amplitude difference between each sorted amplitude of peak and trough

• Spindle density (#spindles/min); calculated as the average number of spindles per minute of NREM artifact-free sleep