/home/precedings/apps/precedings/shared/storage/document/2008/2342/npre20082342-1

Overnight weight loss: relationship with sleep

structure and heart rate variability

Moraes, Walter, MD. PhD

waltermoraes@giro.com.br

Poyares, Dalva, MD, PhD

poyares@psicobio.epm.br

Guilleminault, Christian MD, PhD

cguil@stanford.edu

Rosa, Agostinho, PhD

acrosa@laseeb.org

Mello, Marco Túlio, PhD

tmello@psicobio.epm.br

Rueda,Adriana

adriana@psicobio.epm.br

Tufik, Sergio, MD, PhD.

stufik@psicobio.epm.br

1. Psychobiology Department, Universidade Federal de São Paulo, São Paulo,

SP, Brazil

2. Universidade Técnica de Lisboa, Instituto Superior Técnico, Lisbon, Portugal

3. Human Sleep Research Center, Department of Psychiatry and Behavioral

Science, School of Medicine, Stanford University, Stanford, CA, USA

*Mail: R. Marselhesa, 524 Sao Paulo SP Brazil CEP 04020-060

Fax: 55-11-2108-7633 Phone: 55-11-2108-7633- poyares@psicobio.epm.br

Abstract

Background: Weight loss can be caused by a loss of body mass due to

metabolism and by water loss as unsensible water loss, sweating, or excretion in

feces and urine. Although weight loss during sleep is a well-known phenomenon, it

has not yet been studied in relation to sleep structure or autonomic tonus during

sleep. Our study is proposed to be a first step in assessing the relationship

between overnight weight loss, sleep structure, and HRV (heart rate variability)

parameters.

Methods: Twenty-five healthy volunteers received a 487 kcal meal and 200 ml

water before experiment. Volunteers were weighed before and after

polysomnography. Absolute and relative weight indices were calculated. Time and

frequency domain analysis of heart rate variability was assessed during stages 2,

4, and REM. Nonparametric linear regression analysis was performed between

night weight loss parameters, polysomnographic, and HRV ariables.

Results: High Frequency domain correlated positively with weight loss during

stage 4. Slow wave sleep duration correlated positively with weight loss and

weight loss rate. The duration of Stage 2 correlated negatively with absolute and

relative weight loss.

Conclusions: Weight loss during sleep is dependent upon sleep stage duration

and sleep autonomic tonus. Slow-wave sleep and sleep parasympathetic tonus

may be important for weight homeostasis.

Key words: weight loss, sleep stages, sleep structure, HRV, polysomnography

1. Background

Weight loss can be caused by loss of body mass due to metabolism and by

water loss as insensible water loss, sweating, or excretion in feces and urine.

Eighty-three percent of the total weight loss is due to insensible water loss from

airways and skin.

Water loss rate varies according to changes in activity and

ambient temperature and humidity.

Although weight loss during sleep is a well-

known phenomenon, there are no studies relating it to sleep structure or autonomic

tonus during sleep. Studies on the variation of body composition over 24 hours

using bioimpedance methods showed an increase in weight and a reduction in

height during daytime.

3,4

In these studies, bioimpedance was influenced by food

and drink ingestion.

3,4,5

Many studies assessed the effects of sleep debt on

metabolism and weight gain for long periods of time, suggesting that it may cause

obesity and metabolic syndrome but there are no studies for the effects during

short periods.

6,7,8

There is evidence suggesting a homeostatic mechanism for

weight control in animal models and humans.

9,10

This mechanism is thought to be

dependent on energy intake, energy expenditure, and environmental conditions. In

this context, sleep homeostasis could also influence this process.

9,10

Autonomic

tonus varies according to different sleep stages and influences overnight fluid loss,

blood pressure, and heart rate variability.

12,13,14

HRV (heart rate variability) is an

easy-to-access non-invasive marker of the autonomic tonus during sleep.

12,13,14

Our study is proposed as a first step in assessing the relationship between

overnight weight loss, sleep structure, and HRV parameters.

2. Methods

2.1 Experimental subjects

Experiments were performed in the sleep laboratory of the Psychobiology

Department of the Universidade Federal de São Paulo, São Paulo Brazil.

Twenty-eight healthy young adult volunteers who were regular sleepers and free of

medications were selected for the experiment. There were three drop-outs during

the experiment due to data loss. The final sample consisted of 25 subjects, 7

males and 18 females aged 18-29 years. Female subjects were not in their

menstruation period. Eleven were in estrogenic and 7 were in progestagenic

phases.

2.2. Ethics

Written informed consent forms were signed by all subjects. The protocol

was approved by the Ethics Committee of the Universidade Federal de São Paulo.

2.3. Procedure

The laboratory procedure is summarized in Figure 1. Subjects arrived in the

sleep laboratory at 8 pm after 4 hours of fasting. They underwent a clinical

interview and completed a standard questionnaire about sleep and physical status.

All subjects were eutrophic and normo-hydrated. Subjects were not allowed to

perform physical activity or to be sleep deprived for at least 48 hrs before the

experiment. They received a standard 487 kcal diet and 200 ml water at 9 pm. All

the following measurements were carried out by a single investigator. A pre-sleep

weighing was performed (initial night weight W1) in a 1 g precision Bod Pod

digital scale with computer interface. After that, their height was measured and

sleep clothes were weighed in order to be subtracted from body weight. After body

measuring, subjects were asked not to ingest liquids or solids, not to use the

restroom, and to wear standard clothing during the entire night of sleep. Body

temperature was measured twice (pre and post-sleep). All subjects slept in a

standard room with controlled temperature and humidity (humidity 40-50%,

temperature 23-24

C). Subjects were also asked not to wash any part of their

bodies until the end of the experiment. Polysomnography recording finished at 7

am and weight and height were measured again (post-sleep weight, W2). The

weight variables analyzed were absolute weight loss in grams (AWL), absolute

weight loss rate in grams per hour (AWLR), relative weight loss (RWL = absolute

weight loss/initial weight x 100), relative weight loss rate (RWLR = relative weight

loss per hour), and height variation (HV). Age, gender, menstrual cycle phase,

body temperature, and room temperature were also analyzed. The present study

was limited to the analysis of body-weight variation during the night. We did not

measure lean and fat mass due to technical limitations.

2.3. Polysomnographic recording and scoring

Polysomnography was performed in equipment Meditron Sonolab

sampling at 256 Hz for 4 EEG, 2 electro-oculogram, 1 chin electromyogram, 1 leg

electromyogram, and 1 tracheal microphone; sampling at 500 Hz for 1

electrocardiogram (EKG); sampling at 16Hz for 1 oronasal thermistor, 1 nasal

pressure transducer, 2 chest and abdominal effort sensors, and 1 oximeter

(Nellcor

oximeter). Polysomnograms lasted at least 7 hours. Recordings were

scored following Rechtshaffen and Kales and American Academy of Sleep

Medicine (AASM) criteria.

15,16,17

The polysomnographic variables analyzed were

total sleep time, sleep efficiency (sleep time/record time x 100), sleep latency, REM

sleep latency, REM and non-REM sleep percentage, respiratory disturbance index

(RDI), microarousals/hour (MAI), periodic leg movements/hour (PLMI), and oxygen

saturation.

2.4. Heart rate variability (HRV)

Analyses were performed on one polysomnography EKG-channel (sampling

rate 500Hz). A careful manual review was performed in order to exclude artifacts or

arrhythmias.

2.4.1 Time domain analysis

In a continuous EKG recording, each QRS complex was detected and the

normal-to-normal (NN) intervals were determined. Five time domain indexes were

derived: the standard deviation of all NN intervals (SDNN); the square root of the

mean squared differences of successive NN intervals (RMS); the standard

deviation of successive differences between adjoining normal cycles (SDSD); and

the proportion of adjacent normal NN intervals differing by

50 msec (NN50 and

pNN50).

2.4.2 Frequency Domain analysis

Artifact-free stable sleep stages 2, 4, and REM, in sleep epochs of 5-minute

duration were selected for analysis. We chose the central 5-minute period of the

longest above-mentioned sleep stages. The power densities in the Very Low

Frequency (VLF, 0.0033-0.04Hz), Low Frequency (LF, 0.04-0.15Hz), and High

Frequency (HF, 0.15-0.4Hz) were calculated by integrating the power spectral

density in the respective frequency bands. Normalized power spectra for LF and

HF and LF/HF ratio were also calculated.

2.5. Statistical analysis

For comparison of repeated measurements, the interval data were

transformed into ordinal data because the precondition of normal distribution was

not fulfilled for all variables. Therefore, comparisons of night weight loss,

polysomnographic, and HRV variables were performed by use of nonparametric

Spearman correlation coefficients. Significance was set at the 0.05 level of

probability (two-tailed). Mann-Whitney U-test was utilized to assess differences

between genders and menstrual cycle phases.

3. Results

3.1. Sleep results

Mean RDI was higher in male subjects (4.1

6.2 vs. 0.4

0.8; U=29,

p=0.04). No snoring was detected for any subject during the night of the

experiment. No significant differences were found according to menstrual cycle

phase for any sleep variable. All individual polysomnography variables are outlined

in Table 1.

3.2. Body-measurement results

All individual body measurements are outlined in Table 2. Body temperature

(before and after the sleep period) ranged from 35.9 to 36.7

C. As seen in Table 2,

overall weight loss (300

68g) during sleep for both genders was a common

phenomenon in our population of healthy and young subjects. There were no

differences in weight loss indices between genders and menstrual cycle phases.

There was an overnight increase in height of 1.0

0.7 cm. Height variation did not

correlate with any sleep parameters.

3.3. HRV results

The mean and standard deviations of HRV parameters for all volunteers

during sleep stages 2 and 4 non-REM sleep and REM sleep are provided in Table

3. SDNN (118.7

48.3 vs. 58.5

29.5; U=15, p=0.003), RMSSD (145.1

81.3 vs.

52.2

34.6; U=18, p=0.06), and SDSD (114.9

58.5 vs. 38.9

27.9; U=14,

p=0.003) were higher in men during REM sleep. No significant differences were

found according to menstrual cycle phase for any HRV variable.

3.4. Weight, HRV and polysomnography correlations

There were significant positive correlations between SWS (more specifically

stage 4) and relative weight loss indices. However we also found a negative

correlation between stage 2 and night weight loss.

The HF (parasympathetic component) correlated positively with all absolute and

relative weight loss indices. Correlation R values are outlined in Table 4.

4. Discussion

To our knowledge, this is the first study to report that overnight weight loss is

dependent upon sleep structure. Although we found that weight loss during sleep

was a universal phenomenon, its magnitude was affected by sleep stages and

autonomic tonus. The literature suggests that sleep is important for weight

homeostasis on a long term since sleep shortage is associated with overweight

status, but overnight weight variation had not yet been studied.

5,6,7,8

In short

periods of time, body-weight is a cyclic phenomenon in which the lowest values are

registered after the sleep period.

3,4

Sleep is not a uniform state and there are

documented differences in cardiovascular function, autonomic tonus,

transepidermal water loss, and many other functions during the distinct sleep

stages, particularly SWS and REM sleep.

18,19,20,21

Considering these facts, we

hypothesized that the overnight weight loss rate is not uniform throughout the sleep

period.

In our study, we found a positive correlation between weight loss and SWS

length. Indeed, SWS is a stage when transepidermal water loss is higher due to

differences in autonomous system activity favoring weight loss.

20,21,22

In support of

this hypothesis, we found that higher SWS length and parasympathetic tonus (HF

component of HRV) were directly related to higher overnight weight loss rates.

The present study was not designed to evaluate hormone secretion, but we

can speculate that the endocrine system is also involved in this phenomenon. A

previous study found that ghrelin promotes SWS in humans.

Since ghrelin levels

tend to be higher during longer fasting periods, higher amounts of SWS in faster

weight losing subjects could be explained by higher ghrelin levels.

23,24

Furthermore, reduced GH secretion due to shorter SWS periods could also be

related to reduced overnight weight loss since most human GH is secreted during

SWS.

GH promotes overnight weight loss through increased lipolysis.

25,26

As far as we are aware, this is the first study relating HRV to overnight

weight changes. In our population, men presented higher sympathetic activity

during REM sleep, a state in which surges of sympathetic tone are found. Gender

differences in HRV have also been reported in previous studies.

However, this

increase in the sympathetic component in male subjects did not correlate or seem

to affect the weight loss rate during sleep. Other variables that increased the

sympathetic component are arousals and apnea/hypopnea, which did not affect

weight changes in our sample.

Our results suggest that the main autonomic

parameter linked to overnight weight loss is the vagal tonus. Autonomic tonus is

proven to be related weight homeostasis in long term studies.

In summary, we have confirmed our initial hypothesis that weight loss during

sleep seems to be dependent on sleep-stage duration and autonomic tonus. SWS

and sleep parasympathetic tonus may be important for weight homeostasis.

Consequently, sleep debt and sleep deprivation may negatively affect weight

homeostasis through SWS debt. More precise methods need to be developed in

order to determine the relationship between sleep structure, variations in body

composition, weight loss, autonomic tonus, and hormone secretion.

Acknowledgements

This study was supported by FAPESP (Fundação de Amparo à Pesquisa do

Estado de São Paulo) and AFIP (Associação Fundo de Incentivo à

Psicofarmacologia)

References

1. Weissenberg S. Insensible water loss during sleep: a theoretical exercise.

Advan. physiol. edu. 2005;29:213-215.

2. Henane R, Buquet A, Roussel B, Bittel J. Variations in evaporation and body

temperature during sleep in man. J. Appl. Physiol. 1977;42(1)50-5.

3. Dittmar M, Raschka C, Koch HJ. Effect of diurnal variation on body composition

under consideration of selected chronobiological marker systems. Przegl¹d

Antropologiczny Anthropological Review. 2002;65:17-26.

4. Slinde F, Rossander-Hulthén L. Bioelectrical impedance: effect of 3 identical

meals on diurnal impedance variation and calculation of body composition. Am

J Clin Nutr. 2001;74:4748.

5. Rodríguez Martínez G, Sarría Chueca A, Moreno Aznar L, Fleta Zaragozano J,

Bueno Sánchez M. Diferencias entre la mañana y la tarde en la valoración

nutricional del niño mediante impedancia bioeléctrica. Anales Españoles de

Pediatria. 1998;49(2):145-150.

6. Vioque J, Torres A, Quiles J. Time spent watching television, sleep duration

and obesity in adults living in Valencia, Spain International Journal of Obesity.

2000;24:1683-1688.

7. Sekine M, Yamagami T, Handa K, Saito T, Nanri S, Kawaminami K, Tokui N,

Yoshida K, Kagamimori S A doseresponse relationship between short

sleeping hours and childhood obesity: results of the Toyama Birth Cohort

Study. Child Care, Health & Development. 2002;28:2:163170.

8. Spiegel K, Leproult R, Cauter EV. Impact of sleep debt on metabolic and

endocrine function. Lancet. 1999;354:1435-39.

9. Macias AE. Experimental demonstration of human weight homeostasis:

implications for understanding obesity. British Journal of Nutrition. 2004;

91:479-484

10. Van-Baak M. Adaptative thermogenesis during over and underfeeding in man.

Britsh Journal of Nutrition. 2004;91:329-330.

11. Mathias CJ, Fosbraey P, daCosta DF, Thornley A, Bannister R. The effect of

desmopresin on nocturnal polyuria,overnight weight loss, and morning postural

hypotension in patients with autonomic failure. British Medical Journal.

1986;293:353-354.

12. Akehi Y, Yoshimatsu H, Kurokawa M, Sakata T, Eto H, Ito S, Ono J. VLCD-

induced weight loss improves heart rate variability in moderately obese

Japanese. Exp. Biol. Med. 2001;226(5):440-445

13. Karason K, Molgaard H, Wikstrand J, Sjöström L. Heart rate variability in

obesity and the effect of weight loss. Am. J. Cardiol.1999;83(8):1242-1247.

14. Nault I, Nadreau E, Paquet C, Brassard P, Marceau P, Marceau S, Biron S,

Hould F, Lebel S, Richard D, Poirier P. Impact of bariatric surgery-induced

weight loss on heart rate variability. Metabolism. 2007;56(10):1425-1430.

15. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques

and scoring system for sleep states of human subjects. 1968;US Government

Printing Office.

16. American Sleep Disorders Association. EEG arousals: scoring rules and

examples A preliminary report from Sleep Disorders Atlas Task Force of the

American Sleep. Disorders Association. Sleep. 1992;15:173-184.

17.

AASM Task Force report. Sleep-related breathing disorders in adults:

recommendations for syndrome definition and measurement techniques in

clinical research. Sleep. 1999;22:667-90.

18. Bonnemeier H, Richardt G, Potratz J, Wiegand UK, Brandes A, Kluge N, Katus

HA. Circadian profile of cardiac autonomic nervous modulation in healthy

subjects: differing effects of aging and gender on heart rate variability. J

Cardiovasc Electrophysiol. 2003;14(8):791-799.

19. Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic-nerve activity

during sleep in normal subjects. N Engl J Med. 1993;328:303307.

20. Yosipovitch G, Xiong GL, Haus E, Sackett-Lundeen L, Ashkenazi I, Maibach HI.

Time-dependent variations of the skin barrier function in humans:

transepidermal water loss, stratum corneum hydration, skin surface pH, and

skin temperature. J Invest Dermatol. 1998;110(1):20-3.

21. Yosipovitch G, Sackett-Lundeen L, Goon A, Yiong Huak C, Leok Goh C, Haus

E. Circadian and ultradian (12 h) variations of skin blood flow and barrier

function in non-irritated and irritated skin-effect of topical corticosteroids. J

Invest Dermatol. 2004;122(3):824-9.

22. Kahn A, Van de Merckt C, Dramaix M, Magrez P, Blum D, Rebuffat E, Montauk

L. Transepidermal water loss during sleep in infants at risk for sudden death.

Pediatrics. 1987;80(2):245-50.

23. Weikel JC, Wichniak A, Ising M, Brunner H, Friess E, Held K, Mathias D,

Schmid A, Uhr M, Steiger A. Ghrelin promotes slow-wave sleep in humans. Am

J Physiol Endocrinol Metab, 2003, 284:407-415.

24. Korner J, Leibel RL. To Eat or Not to Eat - How the Gut Talks to the Brain. N

Engl J Med, 2003, 349(10):926-8.

25. Steiger A. Sleep and endocrinology. J Int Med, 2003, 254;13-22.

26. Cauter EV, Leproult R, Plat L. Age-related changes in slow wave sleep and

REM sleep and relationship with growth hormone and cortisol levels in healthy

men. JAMA, 2000, 284(7):861-868.

27. Antelmi I, de Paula RS, Shinzato AR, Peres CA, Mansur AJ, Grupi CJ.

Influence of age, gender, body mass index, and functional capacity on heart

rate variability in a cohort of subjects without heart disease. Am J Cardiol, 2004,

1;93(3):381-385.

Table 1: Polysomnographic results

Subject Sleep

latency
(min)

REM
sleep
latency
(min)

Total
sleep
time
(min)

Sleep
efficiency
(%)

S1% S2% S3% S4% REM%

RDI

(/h)

MAI
(/h)

PLMI (/h) Mean

oxygen
saturation
(%)

1 21.5 80.0 383.0

90.0 4.0 49.1 13.4 10.4 23.0 0.0 9.7 0.0 96.5

2 5.5 138.5

421.0

88.9

13.2

55.1

8.2 12.1 11.4 0.0 10.8 16.1 97.9

3 8.0 155.0

356.0

84.4

6.7 58.7

10.8 5.9 17.8 0.2 20.4 0.0 96.3

4 5.0 153.0

437.5

96.5

3.1 54.2

4.1 22.7 15.9 0.0 17.7 3.2 97.9

5 4.0 177.0

392.5

94.4

6.6 57.3

4.1 15.3 16.7 2.1 21.2 0.0 97.7

6 3.5 117.0

405.5

96.6

11.0

52.4

7.2 19.7

9.7 7.6 18.6

0.0 95.4

7 5.0 77.0 364.5

94.8 4.3 37.6 7.0 17.6

33.0 0.0 7.9 0.0 95.8

8 5.5 178.5

375.5

96.0

4.4 52.3

10.5 16.4 16.4 2.2 4.3 0.0 93.4

9 14.5 102.5 327.0 85.4 6.1 38.1 14.2 19.4 22.2 4.0 4.8 0.0 98.4
10 11.0 64.5 409.5

95.1 4.4 38.7 13.2 18.4 25.3 1.2 3.4 0.0 96.2

11 5.0 70.5 320.5

93.6 5.9 46.3 9.7 15.1 22.9 0.0 11.8 0.0 96.1

12 13.0 74.5 382.0

93.5 6.2 45.4 6.9 15.2

26.3 2.0 6.4 0.0 96.9

13 8.0 146.5

326.5

80.9 8.7 45.6 9.8 26.3

9.5 4.2 18.0

0.0 96.0

14 9.0 302.0

305.5

79.7 12.1 49.3 7.5 25.0

6.1 1.8 13.2

0.0 95.0

15 13.0 129.5 360.5 83.4 11.0 33.6 8.9 31.3

15.3 0.0 7.7 0.0 95.9

16 21.5 147.5 404.5 89.6 6.9 56.5 5.4 22.7

8.4 0.0 4.7 0.0 98.1

17 32.0 129.0 264.0 76.3 13.6 46.2 8.9 24.8

6.4 0.7 7.3 0.0 98.3

18 8.0 108.0

346.5

94.7 4.6 47.9 4.3 31.5 11.7 0.0 18.9 0.0 96.2

19 16.5 13.5 358.5

87.0 6.0 40.6 8.4 29.8 15.2 0.0 10.9 0.0 96.0

20 28.5 178.5 368.5 85.7 7.7 54.8 3.5 18.6 15.3 0.0 15.1 0.0 92.2
21 5.5 187.5

402.5

97.5 1.1 37.6 12.3 29.8 19.1 0.0 8.8 0.0 97.9

22 41.5 126.5 298.0 75.3 6.9 48.5 4.5 25.3

14.8 0.0 7.7 0.0 96.0

23 2.0 179.5

404.0

98.2 1.4 32.7 17.1 35.5 13.4 0.0 2.8 0.0 97.1

24 46.5 155.0 296.5 71.2 16.4 42.8 6.7 8.8 25.3 0.0 14.2

0.0 97.5

25 11.5 201.5 340 85.9 8.5 49.0 10.7 17.6 14.1 0.0 3.9 0.0 95.9
min= minutes /h= per hour

Table 2: Individual body measurement data

Subject Age

(years)

Gender Height

(cm)

Absolute
initial weight
(g)

AWL (g)

AWLR (g/h)

RWL (%)

RWLR (%/h)

HV (cm)

1 22

f 158

46,800

220 23.0 0.47 0.049

1.5

2 28

164

54,573

244 26.6 0.45 0.049

1.5

3 24

f 163

53,139

261 27.6 0.49 0.052

1.5

4 24

f 163

54,756

294 31.0 0.54 0.057

2.0

5 22

f 164

59,235

303 33.4 0.51 0.056

2.5

6 20

169

56,937

255 27.6 0.45 0.048

2.0

7 28

f 170

54,841

269 28.5 0.49 0.052

2.0

8 18

f 160

58,293

262 30.5 0.45 0.052

0.5

0 21

179

68,909

316 36.7 0.46 0.053

2.0

10 20 m 174

64,725

307 33.7 0.47 0.052 0.5

11 24 f 165

54,062

196 21.8 0.36 0.040 1.0

12 25 f 180

56,674

263 29.0 0.46 0.051 0.5

13 21 m 167

57,646

341 37.9 0.59 0.066 1.5

14 23 m 181

78,728

304 33.6 0.39 0.043 0.0

15 22 f 161

51,634

469 49.5 0.91 0.096 1.5

16 24 f 162

53,145

289 31.3 0.54 0.059 1.0

17 23 f 157

57,175

236 27.1 0.41 0.047 1.0

18 21 f 153

47,196

419 47.5 0.89 0.101 1.0

19 20 f 158

50,934

327 33.8 0.64 0.066 1.5

20 29 f 162

72,307

262 29.1 0.36 0.040 1.0

21 18 f 166

53,862

323 33.9 0.60 0.063 1.0

22 19 f 163

73,779

265 28.3 0.36 0.038 0.3

23 20 f 152

47,755

318 30.5 0.67 0.064 0.8

24 21 m 181

74,716

480 46.4 0.64 0.062 0.1

25 24 f 165

55,711

303 30.1 0.54 0.054 0.3

g= grams cm= centimeters h= hours f= female m= male

Table 3: Mean Heart Rate Variability data

REM

Mean NN interval

807.3

243.4 884.3

187.7 853.6

249.8

SDNN

58.6

38.8 62.5

42.7 75.3

44.3

RMSSD

61.2

43.8 72.6

57.2 78.2

65.6

SDSD

45.2

35.8 54.0

52.2 60.2

51.1

NN50

18.0

14.1 19.7

14.7 14.5

11.3

PNN50

27.9

23.2 32.4

24.1 25.4

22.1

VLF

1770.5

984.9 1628.3

1318.5 4242.5

3399.1

2870.2

1948.8 2542.2

1798.8 3265.7

1652.7

2478.9

1612.4 2826.8

1623.8 2101.3

1013.1

Total power

7326.6

3877.0 6873.1

3404.7 9858.7

4998.8

LF/HF

1.4

0.8 1.1

0.7 1.7

1.0

Mean

standard deviation