Effects of sleep disturbances during pregnancy on cardiac autonomic modulation in the resting state

International Journal of Gynecology and Obstetrics 119 (2012) 149–153

Contents lists available at SciVerse ScienceDirect

International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo

CLINICAL ARTICLE

Effects of sleep disturbances during pregnancy on cardiac autonomic modulation in the resting state Kuniko Shiga a, b, Katsuyuki Murata c, Hideya Kodama a,⁎ a b c

Department of Maternity Child Nursing, Akita University Graduate School of Medicine and Faculty of Medicine, Akita, Japan Japanese Red Cross Akita College of Nursing, Akita, Japan Department Environmental Health Sciences, Akita University Graduate School of Medicine and Faculty of Medicine, Akita, Japan

a r t i c l e

i n f o

Article history: Received 24 January 2012 Received in revised form 18 May 2012 Accepted 16 July 2012 Keywords: Cardiac autonomic modulation Heart rate variability Pregnancy Sleep disturbances Snoring

a b s t r a c t Objective: To investigate whether sleep disturbances during pregnancy affect cardiac autonomic modulation in the resting state. Methods: A cross-sectional study was conducted of 160 pregnant women at various stages of gestation. Participants were interviewed about their sleep length per night, sleep quality (Pittsburgh Sleep Quality Index, PSQI), daytime sleepiness (Epworth Sleepiness Scale, ESS), snoring habits, and symptoms of restless legs syndrome during the previous few weeks. Cardiac autonomic modulation in the resting state was quantified by spectral analysis of heart rate variability (HRV) from short-term electrocardiogram monitoring. The relationship of HRV to diverse covariates was studied by multiple regression analysis. Results: No significant influences were observed of short sleep duration (b 7 hours per night), poor sleep quality (PSQI score >5), or restless legs syndrome on HRV measures. Participants with excessive daytime sleepiness (ESS score >9) and habitual snorers had a significantly elevated low frequency/high frequency (LF/HF) ratio and LF power in normalized units (LF norm). In multiple regression analyses, habitual snoring was strongly and positively associated with LF/HF ratio (P b 0.001) and LF norm (Pb 0.001). Conclusion: Resting cardiac autonomic modulation was found to shift toward a sympathetic predominant state among pregnant women who are habitual snorers. © 2012 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction Pregnancy alters sleep in many ways, and sleep disturbances are prevalent among pregnant women [1–4]. Shortening of sleep time, frequent night awakenings, difficulty in falling asleep, and loss of deep sleep stages are widespread problems experienced by pregnant women [1,2]. Excessive daytime sleepiness and frequent snoring are also common, and incidences of sleep-disordered breathing, including obstructive sleep apnea, may be increased during pregnancy [3,4]. Restless legs syndrome (RLS), a neurologic sensory motor disorder affecting sleep, has been observed to increase during pregnancy [1,2]. As disturbed sleep has been implicated as a risk factor for adverse pregnancy outcomes, including pre-eclampsia, gestational hypertension, growth retardation of the fetus, preterm birth, long duration of labor, and increased rate of cesarean delivery [4–7], physicians should take advantage of the opportunity to pay careful attention to sleep disturbances during routine health checks of pregnant women. Many investigators have hypothesized that sleep disturbances affect cardiac autonomic modulation in the resting state, because ⁎ Corresponding author at: Department of Maternity Child Nursing, Akita University Graduate School of Medicine and Faculty of Medicine, 1-1-1 Hondo, Akita-shi 010 8543, Japan. Tel.: +81 18 884 6513; fax: +81 18 884 6500. E-mail address: [email protected] (H. Kodama).

insomnia, short sleep duration, sleep-disordered breathing, and RLS are risk factors for hypertension and cardiovascular morbidity [8]. Spectral analysis of heart rate variability (HRV) has gained widespread recognition for quantification of cardiac autonomic modulation [9]. Using this technique, autonomic modulation in patients with primary insomnia was reported to shift toward a sympathetic predominant state during the night [10]. However, a follow-up investigation failed to demonstrate significant alterations in daytime autonomic functions of insomniacs [11]. In patients with obstructive sleep apnea, microneurography studies demonstrated enhanced sympathetic activity owing to hypoxia and hypercapnia during courses of nocturnal apnea, and a few investigators have speculated that such autonomic responses carry over into the daytime [12]. Daytime HRV of patients with uncomplicated obstructive sleep apnea was found to present a sympathetic predominant state [13], and a follow-up study supported this finding [14]. Furthermore, sleep disturbances sometimes bring about disrupted circadian rhythms, and the cardiac autonomic balance of nurses doing shift work was reported to shift toward a sympathetic predominant state [15]. To date, a number of studies have used HRV measures as autonomic indexes of pregnant women [16–18], but these studies have not evaluated HRV measures in relation to sleep disturbances. Underlying cardiac autonomic modulation among pregnant women may be a key biologic pathway through which disturbed sleep causes adverse

0020-7292/$ – see front matter © 2012 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijgo.2012.05.034

150

K. Shiga et al. / International Journal of Gynecology and Obstetrics 119 (2012) 149–153

pregnancy outcomes. The objective of the present study was to investigate the effects of sleep disturbances during pregnancy on cardiac autonomic modulation in the resting state, which was assessed by frequency domain HRV measures during short-term electrocardiogram (ECG) monitoring. 2. Materials and methods A cross-sectional study was conducted that enrolled pregnant women who attended a private prenatal clinic in Akita, Japan. Eligible participants were recruited between August 1, 2010, and July 31, 2011. The study protocol was approved by the Ethics Committee of Akita University Graduate School of Medicine; all women provided written informed consent. The eligibility criteria included singleton pregnancy; stable relationship; Japanese as the native language; and normal course of pregnancy, without obstetric, medical, or psychiatric complications. Women with pregnancy-induced hypertensive disease (verified blood pressure above140/90 mmHg) or gestational diabetes mellitus (degassed by screening glucose challenge tests followed by 75 g glucose tolerance tests) were excluded. Women who drank any alcohol and/or smoked were also excluded. Approximately 60% of women consecutively approached agreed to participate in the present study; in all, 175 women were enrolled. Confirmation was obtained that participants had not eaten, consumed caffeinated beverages, or performed any exercise for at least 2 hours prior to data collection. Participants first completed a questionnaire about demographic data, including age, gestational age, parity, height, and employment status. As part of a routine health check, measurements of body weight, blood pressure, and blood hemoglobin concentrations were performed. Sitting blood pressure was consistently measured using a sphygmomanometer, and the mean of 2 individual measures was usually recorded. To control for circadian variation in autonomic functions, blood pressure data and ECGs were obtained in the morning (10:00–12:00). Sleep state during the previous few weeks was assessed by subjective sleep variables, including sleep length per night, the Pittsburgh Sleep Quality Index (PSQI), and Epworth Sleepiness Scale (ESS). On the basis of these sleep variables, sleep disturbances of participants were categorized into short sleep duration, poor sleep quality, or excessive daytime sleepiness. Sleep length per night was determined by subjective reports on mean sleep onset time, waking time, and waking duration after sleep onset. Participants with less than 7 hours of sleep per night were categorized as having short sleep duration [2]. The PSQI is a self-rated questionnaire that assesses multiple dimensions of sleep quality [19]. The 18 individual items of the PSQI generate 7 component scores: subjective sleep quality; sleep latency; sleep duration; habitual sleep efficiency; sleep disturbance; use of medicine to aid sleep; and daytime dysfunction. The sum of the 7 component scores yields a single global score (range 0–21); a score above 5 indicates poor sleep quality [19]. The ESS is a subjective self-report measurement of sleepiness [20]. This scale comprises 8 typical situations requiring various degrees of vigilance. Each participant was asked how likely she was to fall asleep in these situations, rated on a scale of 0 to 3. A total ESS score ranges from 0 to 24; excessive daytime sleepiness is defined as a total ESS score of 10 or more [20]. Information was also obtained on snoring habits and RLS. Participants were asked about their experience of snoring during the previous few weeks. In cases where snoring occurred, information on how many nights per week they snored was obtained. When participants did not know whether they snored, information on snoring was solicited from their partners. Participants who snored often (≥3 nights per week) were defined as habitual snorers. Symptoms of RLS were identified according to diagnostic criteria defined by the International RLS Study Group [21]. Participants whose frequency of RLS was at least 3 times per week were defined as having RLS.

The ECGs for HRV analysis were recorded in a quiet room by 2 trained examiners. Participants were advised to breathe normally, and to inform the examiners if they felt unwell during the recording. After they had been lying in a relaxed, supine position and breathing normally for 10 minutes, the R-R interval was measured 300 times (sampling time, 1 millisecond) using a YPI-01 ECG waveform analyzer (Yokote Precision Industry, Akita, Japan). During ECG recordings, participants were kept in a totally supine position. Subsequently, 100 consecutive R-R intervals with a minimal standard deviation (SD) were automatically extracted from the stored data to calculate frequency domain parameters by spectral analysis [22]. Using an autoregression model, the spectral powers were calculated for the frequency bands of high frequency (HF; 0.15–0.40 Hz), low frequency (LF; 0.04–0.15 Hz), and total frequency (0.00–0.40 Hz). The ratio of LF power to HF power (LF/HF ratio) and the LF power in normalized units (LF norm; defined as LF power/[LF power + HF power]) were determined as quantitative indexes of sympathovagal balance [9]. Results are expressed as the mean ± SD. Owing to skewed distribution, total power, LF power, and HF power were logarithmically transformed. Relations between subjective sleep variables were examined by Pearson correlation coefficients or the Student t test. Group differences in HRV measures by demographic and physical characteristics were examined by the Student t test or 1-way analysis of variance. Differences in HRV measures with respect to each sleep disturbance category were examined by analysis of covariance, controlling for employment status, gestational age, systolic blood pressure, and body mass index (BMI, calculated as weight in kilograms divided by the square of height in meters). Relation of HRV to diverse covariates was studied by multiple regression analysis, in which each HRV measure was taken as the dependent variable, and sleep disturbance categories and potential confounders—including maternal age, parity, employment status, gestational age, systolic blood pressure, BMI, and hemoglobin concentration—as independent variables. All statistical analyses, with 2-sided P values, were performed using the Statistical Package for the Biosciences (Nankodo, Tokyo, Japan). A P value below 0.05 was considered statistically significant.

3. Results None of the 175 participants withdrew from the study; however, 15 women were excluded owing to unstable HRV data (arrhythmia or physical movement during the ECG). Complete data were, therefore, obtained from 160 women. Distribution of the study variables are shown in Table 1. The study population comprised 22 women in the first trimester, 53 women in the second trimester, and 85 women in the third trimester. Of 76 working women, 10 were shift workers (nurses or healthcare workers). Relationships between subjective sleep variables are shown in Table 2. No significant simple correlation was found between PSQI and ESS scores. Sleep length per night was weakly correlated with PSQI score. The PSQI and ESS scores of habitual snorers were significantly higher than those of non-habitual snorers. The PSQI score of participants with RLS was significantly higher than that of participants without RLS. Table 3 shows associations between demographic and physical characteristics and HRV measures in the study population. Employment status, gestational stage, systolic blood pressure, and BMI were all significantly associated with HRV measures. Characteristics of HRV measures in each sleep disturbances category are shown in Table 4. Significant characteristics in HRV measures were not detected in the sleep disturbances categories, including short sleep duration, poor sleep quality, or RLS. Excessive daytime sleepiness was significantly associated with a high LF/HF ratio and LF norm. Habitual snorers had significantly higher heart rates, LF power, LF/HF ratio, and LF norm than non-snorers.

K. Shiga et al. / International Journal of Gynecology and Obstetrics 119 (2012) 149–153 Table 1 Distribution of the study variables among pregnant women (n = 160).a Variable

Distribution

Demographic and physical characteristics Age, y Parity Nulliparous Employed Gestational age, wk Systolic blood pressure, mmHg BMI Hemoglobin, g/dL Subjective sleep variables and sleep disturbance categories Sleep length per night, min PSQI score ESS score Short sleep duration Poor sleep quality Excessive daytime sleepiness Habitual snoring Restless legs syndrome Heart rate variability measures Heart rate, beats/min Total power, ms2 Log [total power, ms2] LF power, ms2 Log [LF power, ms2] HF power, ms2 Log [HF power, ms2] LF/HF ratio LF norm, %

29.3 ± 4.6 0.70 ± 0.81 78 76 27.4 ± 9.1 107 ± 12 23.6 ± 3.2 11.6 ± 0.9

(19–44) (0–4) (48.8) (47.5) (7–40) (78–140) (16.7–34.4) (8.9–14.5)

454 ± 74 4.6 ± 1.7 6.0 ± 3.3 41 51 22 40 19

(240–660) (2–9) (1–20) (25.6) (31.9) (13.8) (25.0) (11.9)

77.3 ± 9.6 1061 ± 1236 2.96 ± 0.35 463 ± 497 2.46 ± 0.44 369 ± 438 2.33 ± 0.48 1.83 ± 1.32 56.7 ± 18.8

(53–99) (113–6774) (2.06–3.80) (20–3345) (1.30–3.52) (12–2676) (1.1–3.43) (0.15–6.78) (13.1–87.1)

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); ESS, Epworth Sleepiness Scale; HF, high frequency; LF, low frequency; PSQI, Pittsburgh Sleep Quality Index. a Values are given as mean ± SD (range) or number (percentage).

In the present study, participants with less than 7 hours of sleep were categorized as having short sleep duration, per a previous publication [2]. However, another study [7] chose less than 6 hours of sleep for the threshold for short sleep duration, and we were concerned that our moderate criteria might mask a significant influence of sleep duration. Therefore, data were also examined after less than 6 hours of sleep (only 11 participants [7%] were included in this short sleep duration category) but the results were not affected (data not shown). Results of multiple regression analysis with diverse covariates are presented in Table 5. As colinearity among the independent variables was not observed (the maximum simple correlation coefficient was 0.318), each independent variable was considered to have preserved the independency among them. Habitual snoring was positively associated with heart rate, LF power, LF/HF ratio, and LF norm. Additionally,

Table 2 Relationship between subjective sleep variables among pregnant women (n = 160). Variable

Sleep length per night

PSQI score

Sleep length per night a PSQI score a ESS score a Habitual snoring b Yes (n = 40) No (n = 120) Restless leg syndrome b Yes (n = 19) No (n = 141)

1.000 −0.174 −0.100

−0.174 1.000 0.073

ESS score

444 ± 88 458 ± 69

5.05 ± 1.84 4.41 ± 1.70

c

7.58 ± 3.89 5.45 ± 2.94

463 ± 80 453 ± 74

5.74 ± 1.97 4.41 ± 1.67

d

6.84 ± 1.97 5.87 ± 3.98

c

−0.100 0.073 1.000 d

Abbreviations: ESS, Epworth Sleepiness Scale; PSQI, Pittsburgh Sleep Quality Index. a Pearson correlation coefficients. b Values given as mean ± SD. Group differences were examined via Student t test. c P b 0.05. d P b 0.01.

151

employment was negatively associated with heart rate and LF norm, but positively associated with HF power. Gestational age was negatively associated with total power, LF power, and HF power. Body mass index was positively associated with heart rate and negatively associated with HF power. 4. Discussion In the present study, no significant influences of short sleep duration, poor sleep quality, or RLS on HRV measures were detected among a group of 160 pregnant women. By contrast, however, participants with excessive daytime sleepiness or habitual snoring had significantly increased values for LF/HF ratio and LF norm. Both LF/HF ratio and LF norm are quantitative indexes of sympathovagal balance, with high values indicating an autonomic shift toward relative dominance of sympathetic neural activity [9]. Thus, the findings of the present study reveal the presence of a sympathetic predominant state among pregnant women with excessive daytime sleepiness or habitual snoring. Daytime sleepiness and snoring are frequent symptoms of sleepdisordered breathing [3,4]. In the present study, habitual snorers had a markedly elevated ESS scores, suggesting a parallel association between snoring and daytime sleepiness. However, which of these 2 factors has the most significant impact on HRV still needs to be determined. The multiple regression analysis, which included potential confounders as covariates, validated the relationship between habitual snoring and the LF/HF ratio and LF norm, whereas it did not validate the association between the ESS score and these measures of HRV. These data suggest the relationship of habitual snoring with elevated LF/HF ratio and LF norm are independent of covariates, and significant sympathetic activation was associated with habitual snoring. This observation might imply that a considerable proportion of pregnant women who are habitual snorers are actually experiencing obstructive sleep apnea because a sympathetic predominant state during the daytime was demonstrated in patients with obstructive sleep apnea in a general population [12–14]. Furthermore, the results of the present study may support a previous publication [4] that supported snoring as a sign of pregnancy-induced hypertension and a risk factor for fetal growth retardation, although a study by Köken et al. [23] has revealed conflicting results on fetal outcome. Careful comparison of the present data on habitual snorers to previous studies of patients with obstructive sleep apnea [13,14] reveals a major difference. Although the LF/HF ratio of both populations was increased, HF power was concurrently decreased only among individuals with obstructive sleep apnea. This difference may be attributable to alterations in HRV measures by position changes in pregnant women. When pregnant women turned the body from the left lateral to the supine position, HF power was decreased and the LF/HF ratio increased [17,18]. This response is specific to pregnant women, and probably reflects compression of the enlarged uterus on the vena cava inferior, which causes temporal decreases in cardiac venous return. A similar response was observed among non-pregnant individuals at supine rest and during a 90° head-up tilt [9]. Therefore, when the ECGs of pregnant women were recorded in the supine position in the present study, it is possible that the HF power component was artificially decreased, which might mask specific HF power reductions associated with habitual snoring during pregnancy. In the present study, there was no validated influence of sleep disturbances on HF power, an established index of cardiac vagal modulation reflecting respiratory sinus arrhythmia [9]. The HF power was negatively related to gestational age and BMI, suggesting that respiratory sinus arrhythmia of pregnant women was reduced as a result of restricted breathing owing to the enlarged uterus and deposition of adipose tissue [17,18]. In addition, HF power was positively related to employment status. This relationship may be explained by differences in daily physical activity between working and non-working participants. An increase of physical activities increased vagal modulation in

152

K. Shiga et al. / International Journal of Gynecology and Obstetrics 119 (2012) 149–153

Table 3 Associations between demographic and physical characteristics and heart rate variability measures among pregnant women (n = 160).a,b Demographic and physical characteristics Age, y ≤24 (n = 25) 25–34 (n = 110) ≥35 (n = 25) Parity Nulliparous (n = 78) Primiparous or Multiparous (n = 82) Employment status Working (n = 76) Not working (n = 84) Gestational stage First trimester (n = 22) Second trimester (n = 53) Third trimester (n = 85) Systolic blood pressure, mmHg b125 (n = 149) 125–140 (n = 11) BMI b25 (n = 116) 25–29 (n = 37) ≥30 (n = 7) Hemoglobin, g/dL b11 (n = 22) ≥11 (n = 138)

Heart rate, beats/min

78.8 ± 11.0 77.0 ± 9.6 77.0 ± 8.1

P value

Log [total power (ms2)]

0.679

3.05 ± 0.28 2.95 ± 0.38 2.89 ± 0.29

77.2 ± 10.2 77.3 ± 9.0

0.949

75.7 ± 9.8 78.7 ± 9.2

P value

Log [LF power (ms2)]

P value

Log [HF power (ms2)]

P value

LF/HF ratio

P value

LF norm, %

P value

0.671

1.87 ± 1.26 1.85 ± 1.34 1.72 ± 1.32

0.899

57.2 ± 20.0 57.0 ± 18.6 54.9 ± 19.5

0.866

0.277

2.54 ± 0.38 2.46 ± 0.45 2.39 ± 0.45

0.453

2.40 ± 0.42 2.32 ± 0.48 2.29 ± 0.52

3.00 ± 0.32 2.92 ± 0.37

0.154

2.48 ± 0.42 2.44 ± 0.46

0.515

2.35 ± 0.48 2.31 ± 0.48

0.567

1.87 ± 1.39 1.80 ± 1.25

0.751

56.7 ± 19.2 56.7 ± 18.6

0.992

0.047

3.02 ± 0.35 2.90 ± 0.35

0.035

2.52 ± 0.46 2.40 ± 0.41

0.083

2.43 ± 0.47 2.23 ± 0.47

0.008

1.75 ± 1.41 1.91 ± 1.23

0.443

54.4 ± 19.9 58.8 ± 17.7

0.145

76.0 ± 9.3 75.4 ± 9.0 78.8 ± 9.9

0.097

3.19 ± 0.37 2.94 ± 0.30 2.90 ± 0.36

0.002

2.85 ± 0.40 2.46 ± 0.37 2.36 ± 0.44

b0.0001

2.67 ± 0.46 2.40 ± 0.41 2.20 ± 0.48

b0.0001

2.08 ± 1.41 1.48 ± 1.01 1.99 ± 1.43

0.059

59.1 ± 20.5 53.4 ± 17.2 58.2 ± 19.3

0.291

76.5 ± 9.2 87.5 ± 9.6

0.0002

2.97 ± 0.35 2.73 ± 0.37

0.026

2.47 ± 0.43 2.27 ± 0.49

0.131

2.36 ± 0.46 1.87 ± 0.50

0.001

1.77 ± 1.31 2.68 ± 1.09

0.026

55.7 ± 19.0 70.5 ± 9.0

0.012

75.4 ± 8.8 80.8 ± 9.6 89.6 ± 8.6

b0.0001

2.97 ± 0.35 2.99 ± 0.33 2.58 ± 0.24

0.014

2.50 ± 0.44 2.39 ± 0.42 2.11 ± 0.42

0.037

2.40 ± 0.46 2.23 ± 0.44 1.67 ± 0.40

0.0001

1.75 ± 1.32 1.87 ± 1.28 2.94 ± 1.09

0.067

55.2 ± 19.4 58.5 ± 17.1 72.3 ± 9.6

0.053

76.1 ± 10.2 77.4 ± 9.5

0.548

2.97 ± 0.33 2.95 ± 0.36

0.848

2.42 ± 0.36 2.47 ± 0.45

0.616

2.30 ± 0.41 2.33 ± 0.49

0.788

1.76 ± 1.31 1.84 ± 1.32

0.773

55.6 ± 19.0 56.9 ± 18.9

0.773

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); HF, high frequency; LF, low frequency. a Values are given as mean ± SD. b Group differences were examined via Student t test or 1-way analysis of variance.

pregnant women with obesity [24], and working women are likely more physically active than their non-working counterparts. A few limitations of the present study warrant discussion. First, all sleep-related parameters were obtained from subjective reports, and objective assessment of sleep using actigraphy or polysomnography may provide different results. Second, the mental stress levels of the participants were not evaluated. Cardiac autonomic modulations in the resting state may be influenced by daily worry, emotional strain, or anxiety [25], which also affect sleep state among pregnant women. Third, HRV measures were obtained briefly during the day. Ambulatory 24-hour ECG recordings may elucidate the influence of

sleep disturbances on the nighttime HRV measures that may have clinical importance. In conclusion, the present study demonstrated that cardiac autonomic modulation in the resting state of pregnant women who are habitual snorers shifted toward a sympathetic predominant state, as indicated by elevated LF/HF ratio and LF norm. This preliminary epidemiologic survey calls for further investigations. Although obstructive sleep apnea has been implicated as an important health threat among pregnant women, correct diagnosis of this condition is not easy because of the lack of specific symptoms and necessity for nocturnal polysomnographic examination [3,4]. Thus, polysomnography

Table 4 Characteristics of heart rate variability measures in each sleep disturbance category among pregnant women (n = 160).a,b Sleep disturbance category

Short sleep duration (sleep length b7 h/night) Yes (n = 41) No (n = 119) Poor sleep quality (PSQI score >5) Yes (n = 51) No (n = 109) Excessive daytime sleepiness (ESS score >9) Yes (n = 22) No (n = 138) Habitual snoring Yes (n = 40) No (n = 120) Restless legs syndrome Yes (n = 19) No (n = 141)

Heart rate, beats/min

P value

Log [total power (ms2)]

P value

Log [LF power (ms2)]

P value

Log [HF power (ms2)]

P value

LF/HF ratio

P value

LF norm, %

P value

76.2 ± 8.6 77.6 ± 8.8

0.402

2.95 ± 0.30 2.96 ± 0.35

0.840

2.49 ± 0.35 2.45 ± 0.43

0.661

2.30 ± 0.40 2.34 ± 0.43

0.659

1.92 ± 1.15 1.80 ± 1.32

0.623

59.4 ± 16.3 55.8 ± 18.5

0.300

78.2 ± 9.0 76.8 ± 8.6

0.358

3.01 ± 0.35 2.93 ± 0.32

0.192

2.54 ± 0.46 2.42 ± 0.39

0.097

2.36 ± 0.45 2.32 ± 0.40

0.590

1.92 ± 1.19 1.79 ± 1.32

0.562

59.7 ± 17.4 55.3 ± 18.1

0.160

79.0 ± 7.8 77.0 ± 8.9

0.337

2.96 ± 0.30 2.96 ± 0.34

0.994

2.61 ± 0.33 2.44 ± 0.42

0.074

2.30 ± 0.38 2.33 ± 0.43

0.763

2.37 ± 1.21 1.75 ± 1.27

0.039

65.9 ± 14.1 55.3 ± 18.2

0.012

80.7 ± 8.7 76.2 ± 8.6

0.009

2.92 ± 0.27 2.97 ± 0.35

0.537

2.61 ± 0.35 2.41 ± 0.42

0.016

2.27 ± 0.35 2.35 ± 0.44

0.373

2.57 ± 1.21 1.85 ± 1.22

b0.0001

67.2 ± 15.2 53.3 ± 17.7

b0.0001

77.7 ± 7.4 77.2 ± 8.9

0.834

2.96 ± 0.38 2.96 ± 0.33

0.923

2.46 ± 0.45 2.46 ± 0.41

0.978

2.27 ± 0.42 2.34 ± 0.42

0.552

1.97 ± 1.29 1.81 ± 1.28

0.618

59.7 ± 17.2 56.3 ± 18.1

0.452

Abbreviations: ESS, Epworth Sleepiness Scale; HF, high frequency; LF, low frequency; PSQI, Pittsburgh Sleep Quality Index. a Values are given as mean ± SD. b Analysis of covariance, adjusted for employment status, gestational age, systolic blood pressure, and body mass index (BMI, calculated as weight in kilograms divided by the square of height in meters).

K. Shiga et al. / International Journal of Gynecology and Obstetrics 119 (2012) 149–153

153

Table 5 Multiple regression analysis of the relationship of sleep disturbance categories and demographic and physical characteristics to heart rate variability measures among pregnant women (n = 160). Variable Multiple correlation coefficient Independent variables a Sleep disturbance categories Short sleep duration Poor sleep quality Excessive daytime sleepiness Habitual snoring Restless leg symptoms Demographic and physical characteristics Age Nulliparous Employed Gestational age Systolic blood pressure BMI Hemoglobin

Heart rate 0.476

0.117 0.062 0.031 0.218 −0.011 −0.056 0.021 −0.169 0.118 0.124 0.236 0.077

b

b

c

b

Log [total power] 0.373

c

Log [LF power] 0.438

0.044 0.097 −0.005 −0.051 0.002

0.021 0.111 0.106 0.173 −0.027

−0.127 −0.042 0.157 −0.256 −0.122 −0.024 0.018

−0.096 0.063 0.097 −0.208 −0.039 −0.142 0.059

b

b

c

c

Log [HF power] 0.513

b

0.041 0.048 −0.014 −0.054 −0.024 −0.136 0.036 0.202 −0.306 −0.112 −0.223 −0.020

LF/HF ratio 0.415

0.013 0.020 0.108 0.301 −0.008

b d

b

b

d

0.052 −0.001 −0.094 0.154 0.048 0.107 0.096

LF norm 0.468

−0.027 0.079 0.147 0.280 −0.003 0.067 0.032 −0.154 0.149 0.100 0.121 0.098

b

d

c

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by the square of height in meters); HF, high frequency; LF, low frequency. a Standardized regression coefficients. b P b 0.01. c P b 0.05. d P b 0.001.

of pregnant women who are habitual snorers should be evaluated in relation to daytime LF/HF ratio or LF norm to determine the diagnostic value of these measures. In addition, following a previous publication [4], predictive values of these HRV measures with snoring for adverse pregnancy outcome should be assessed in a prospective cohort study. Acknowledgments The present study was supported by Grant-in-Aid for Scientific Research (No. 23593281) from the Japan Society for the Promotion of Science. Conflict of interest The authors have no conflicts of interest. References [1] Hedman C, Pohjasvaara T, Tolonen U, Suhonen-Malm AS, Myllylä VV. Effects of pregnancy on mothers' sleep. Sleep Med 2002;3(1):37-42. [2] Facco FL, Kramer J, Ho KH, Zee PC, Grobman WA. Sleep disturbances in pregnancy. Obstet Gynecol 2010;115(1):77-83. [3] Venkata C, Venkateshiah SB. Sleep-disordered breathing during pregnancy. J Am Board Fam Med 2009;22(2):158-68. [4] Franklin KA, Holmgren PA, Jönsson F, Poromaa N, Stenlund H, Svanborg E. Snoring, pregnancy-induced hypertension, and growth retardation of the fetus. Chest 2000;117(1):137-41. [5] Edwards N, Blyton CM, Kesby GJ, Wilcox I, Sullivan CE. Pre-eclampsia is associated with marked alterations in sleep architecture. Sleep 2000;23(5):619-25. [6] Okun ML, Hall M, Coussons-Read ME. Sleep disturbances increase interleukin-6 production during pregnancy: implications for pregnancy complications. Reprod Sci 2007;14(6):560-7. [7] Lee KA, Gay CL. Sleep in late pregnancy predicts length of labor and type of delivery. Am J Obstet Gynecol 2004;191(6):2041-6. [8] Malhotra A, Loscalzo J. Sleep and cardiovascular disease: an overview. Prog Cardiovasc Dis 2009;51(4):279-84. [9] Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Eur Heart J 1996;17(3): 354-81.

[10] Bonnet MH, Arand DL. Heart rate variability in insomniacs and matched normal sleepers. Psychosom Med 1998;60(5):610-5. [11] Fang SC, Huang CJ, Yang TT, Tsai PS. Heart rate variability and daytime functioning in insomniacs and normal sleepers: preliminary results. J Psychosom Res 2008;65(1): 23-30. [12] Carlson JT, Hedner J, Elam M, Ejnell H, Sellgren J, Wallin BG. Augmented resting sympathetic activity in awake patients with obstructive sleep apnea. Chest 1993;103(6): 1763-8. [13] Narkiewicz K, Montano N, Cogliati C, van de Borne PJ, Dyken ME, Somers VK. Altered cardiovascular variability in obstructive sleep apnea. Circulation 1998;98(11): 1071-7. [14] Hilton MF, Chappell MJ, Bartlett WA, Malhotra A, Beattie JM, Cayton RM. The sleep apnoea/hypopnoea syndrome depresses waking vagal tone independent of sympathetic activation. Eur Respir J 2001;17(6):1258-66. [15] Ishii N, Dakeishi M, Sasaki M, Iwata T, Murata K. Cardiac autonomic imbalance in female nurses with shift work. Auton Neurosci 2005;122(1–2):94-9. [16] Yang CC, Chao TC, Kuo TB, Yin CS, Chen HI. Preeclamptic pregnancy is associated with increased sympathetic and decreased parasympathetic control of HR. Am J Physiol Heart Circ Physiol 2000;278(4):H1269-73. [17] Kuo CD, Chen GY, Yang MJ, Lo HM, Tsai YS. Biphasic changes in autonomic nervous activity during pregnancy. Br J Anaesth 2000;84(3):323-9. [18] Matsuo H, Inoue K, Hapsari ED, Kitano K, Shiotani H. Change of autonomic nervous activity during pregnancy and its modulation of labor assessed by spectral heart rate variability analysis. Clin Exp Obstet Gynecol 2007;34(2):73-9. [19] Buysse DJ, Reynolds III CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989;28(2):193-213. [20] Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 1991;14(6):540-5. [21] Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, Montplaisi J, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology. A report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med 2003;4(2):101-19. [22] Murata K, Landrigan PJ, Araki S. Effects of age, heart rate, gender, tobacco and alcohol ingestion on R-R interval variability in human ECG. J Auton Nerv Syst 1992;37(3):199-206. [23] Köken G, Sahin FK, Cosar E, Saylan F, Yilmaz N, Altuntas I, et al. Oxidative stress markers in pregnant women who snore and fetal outcome: a case control study. Acta Obstet Gynecol Scand 2007;86(11):1317-21. [24] Stutzman SS, Brown CA, Hains SM, Godwin M, Smith GN, Parlow JL, et al. The effects of exercise conditioning in normal and overweight pregnant women on blood pressure and heart rate variability. Biol Res Nurs 2010;12(2):137-48. [25] Brosschot JF, Van Dijk E, Thayer JF. Daily worry is related to low heart rate variability during waking and the subsequent nocturnal sleep period. Int J Psychophysiol 2007;63(1):39-47.