Maturation of Heart Rate and Blood Pressure Variability during Sleep in Term-Born Infants

Abstract

Study Objectives:

Abnormal blood pressure control is implicated in the sudden infant death syndrome (SIDS). However, no data exist on normal development of blood pressure control during infancy. This study assessed maturation of autonomic control of blood pressure and heart rate during sleep within the first 6 months of life.

Participants:

Term infants (n = 31) were studied longitudinally at 2-4 weeks, 2-3 months, and 5-6 months postnatal age.

Interventions:

Infants underwent daytime polysomnography at each age studied. Blood pressure and heart rate were recorded during quiet (QS) and active (AS) sleep in undisturbed baseline and head-up tilt conditions.

Measurements and Results:

Autonomic control was assessed using spectral indices of blood pressure and heart rate variability (BPV and HRV) in ranges of low frequency (LF, reflecting sympathetic + parasympathetic activity) and high frequency (HF, parasympathetic activity), total power (LF+HF), and LF/HF ratio (sympathovagal balance).

With increasing postnatal age and predominantly during QS, HRV-LF, HRV-HF, and HRV total power increased, while HRV-LF/HF decreased. BPV-LF/HF also decreased with postnatal age. All changes were evident in both baseline and head-up tilt conditions. BPV-LF and BPV total power during tilts were markedly reduced in QS versus AS at each age.

Conclusions:

In sleeping infants, sympathetic vascular modulation of the circulation decreases with age, while parasympathetic control of heart rate is strengthened. These normative data will aid in the early identification of conditions where autonomic function is impaired, such as in SIDS.

Citation:

Yiallourou SR; Sands SA; Walker AM; Horne RSC. Maturation of heart rate and blood pressure variability during sleep in term-born infants. SLEEP 2012;35(2):177-186.

INTRODUCTION

During infancy regulation of blood pressure and heart rate by the autonomic nervous system is immature and thus infants are frequently exposed to sudden hypotensive or hypertensive events during sleep.¹ Cardiovascular instability in sleep is of particular importance to infant well-being, as sleep time is at a lifetime maximum, with young infants averaging 70% of each 24 hours in sleep. In particular, impaired blood pressure control has been reported in infants born preterm^2,3 and has been implicated in the sudden infant death syndrome (SIDS).^–7

Noninvasive assessment of autonomic control of both heart rate and blood pressure can be achieved by analysis of heart rate variability (HRV) and blood pressure variability (BPV) using power spectral analysis in the frequency domain.⁸ Spectral analysis separates cardiovascular changes into specific frequency components, providing quantitative information on sympathetic and parasympathetic influences on cardiovascular dynamics. Previously, spectral analysis has been used to assess autonomic control of both heart rate and blood pressure at rest and during head-up tilt conditions in adults.^9,10 In infants, however, the use of spectral analysis has been mainly confined to assessment of heart rate control.¹¹ Previous studies of the maturation of HRV in infancy have shown an increasing parasympathetic dominance over the first 6 months of life,^12–15 and greater parasympathetic dominance in quiet sleep (QS) versus active sleep (AS) during this period.^13,14,16 Although infant data on BPV are limited, Andriessen and colleagues¹⁷ assessed BPV in preterm and term neonates at postmenstrual ages ranging from 26-42 weeks and showed that BPV decreased with postmenstrual age, suggesting a reduction in sympathetic vascular modulation with maturation. However, to date there have been no equivalent studies of the postnatal maturation of BPV in term infants within the first 6 months of life, nor have there been assessments of the effects of sleep state on BPV.

Thus, the overall objectives for this study were to assess the maturation of autonomic control of both heart rate and blood pressure during infant sleep. Since heart rate variation contributes significantly to the physiological regulation of blood pressure, we used simultaneous assessment of HRV and BPV to enable a comprehensive assessment of autonomic control. HRV and BPV were analyzed under baseline conditions, reflecting natural variations, and during head-up tilts to impose a challenge to test autonomic regulation of blood pressure. Based on the limited infant data,^17,18 we hypothesized that: (1) sympathetic modulation of the circulation decreases with postnatal age, while parasympathetic control of heart rate increases; and (2) parasympathetic control dominates in QS, while sympathetic control dominates in AS.

METHODS

Ethical approval for this study was granted by the Southern Health and Monash University Human Research Ethics Committees. Written parental consent was obtained prior to the commencement of the study for each infant. No monetary incentive was provided for participation.

Subjects

Thirty-one infants (16 female/15 male), born full-term (38-42 weeks of gestational age), with normal birth weights ranging from 2900 g to 4250 g (mean 3590 ± 100 g, mean ± SEM) and APGAR scores of 9-10 (median 9) at 5 minutes were recruited for the study. Each infant was studied longitudinally at 2-4 weeks (mean 3 ± 0.1 weeks), at 2-3 months (mean 10 ± 0.1 weeks), and at 5-6 months (mean 22 ± 0.2 weeks) using daytime polysomnography. All infants enrolled in the study were born to non-smoking mothers, and all infants routinely slept supine at home.

Polysomnography

Polysomnographic recordings included continuous monitoring of electroencephalogram (EEG; (C4/A1; O2/A1); electroocculogram, submental electromyogram, electrocardiogram (ECG), thoracic and abdominal breathing movements (Resp-ez Piezo-electric sensor, EPM Systems, Midlothian, VA, USA), arterial blood oxygen saturation (SpO₂) (Biox 3700e Pulse Oximeter, Ohmeda, Louisville, CO, USA), and abdominal skin temperature (YSI 400 series thermistor, Yellow Springs Instruments, Yellow Springs, OH, USA).

All electrodes and measuring devices for polysomnography were attached during the infant's morning feed. Infants were then allowed to sleep naturally, in the supine position, in a pram in a darkened room at constant temperature (22-23°C).

Polygraphic data were amplified via a polygraph (Grass Instrument Co, Quincy, MA, USA), digitized at a frequency of 400 Hz via a 16 channel Powerlab system (ADInstruments, Sydney, Australia) and recorded onto a computer program for data storage, analysis, and visualization (LabChart 7.2, ADInstruments, Sydney, Australia).

Infants were visually monitored continuously via an infrared camera placed above the pram, and behavioral changes, such as body movements and crying, were recorded.

Recordings were performed during 2 cycles of sleep, with the recording duration per cycle of sleep averaging 21 ± 2 min in QS and 22 ± 2 min in AS at 2-4 weeks, 25 ± 2 min in QS and 21 ± 2 min in AS at 2-3 months, and 24 ± 2 min in QS and 18 ± 1 min in AS at 5-6 months. Total recording time, including awake periods averaged 196 ± 12 min, 200 ± 12 min, and 186 ± 16 min at 2-4 weeks, 2-3 months, and 5-6 months, respectively.

Blood Pressure Measurement

A non-invasive continuous measure of systolic blood pressure (SBP) was made using a photoplethysmographic cuff placed around the infant's wrist (Finometer, FMS, Finapres Medical Systems, The Netherlands) using methods previously described.^5,6,19,20 The Finometer has an inbuilt self-calibration system (physiocal) that is activated at the start of each blood pressure measurement; calibration is effected via an automated algorithm that adjusts the transmural pressure of the artery to equal zero. After self-calibration, continuous SBP measurements of 2 min duration were performed for each recording epoch. To avoid venous pooling in the infant's hand, a 2-min rest period was instituted between each recording period. The blood pressure signal was digitized simultaneously with the other polygraphic recordings at a frequency of 400 Hz via the 16 channel Powerlab system (ADInstruments, Sydney, Australia).

Sleep State

Using EEG, behavioral, heart rate, and breathing pattern criteria,²¹ sleep state was assessed as either QS, AS, or indeterminate sleep. Data from indeterminate sleep were excluded from subsequent analysis.

Assessment of Autonomic Control

To assess autonomic control, measurements of blood pressure and heart rate were recorded during baseline conditions and during a cardiovascular challenge via the head-up tilt method.

Baseline Measurements

During each sleep state and at each age studied, we aimed to achieve n = 3 control measurements of blood pressure, together with heart rate of 2 min duration, that were devoid of movement artifact, sighs, or apneas.

Responses to Cardiovascular Challenge

Head-up tilts were utilized as an autonomic challenge as previously described.^3,5,22 Briefly, at each age and in each sleep state, during physiologically stable conditions, 3 head-up tilts were performed manually from the horizontal position to an angle of 15° over 2-3 s. Each tilt consisted of a baseline period and a tilt period, each of 1-min duration. Between each head-up tilt, a 2-min rest period was allowed. Head-up tilts that resulted in an arousal response, sigh, or apnea were excluded from analysis.

Data Analysis

Heart rate and blood pressure variability

Power spectral analysis:

Power spectral analysis was performed to separate SBP and R-R interval time series into high and low frequency components. Low frequency (LF) changes of HRV are attributed to baroreflex mediated influences and reflect both sympathetic and parasympathetic activity.^10,23,24 The high frequency (HF) components of HRV are attributed to respiratory related changes (sinus arrhythmia) and thus reflect parasympathetic activity.^10,23,24 The LF component of BPV is thought to reflect sympathetic vasomotor modulation^10,24,25 and can be heightened via stimulation with a tilt.^9,10 The HF component of BPV, while influenced by mechanical effects of respiration acting directly on intrathoracic elements of the cardiovascular system, is also influenced via changes in cardiac stroke volume and R-R interval that are affected by parasympathetic activity.^10,23,24 The LF/HF ratio can also be computed to assess sympathovagal balance for HRV and BPV.^10,23,24

Spectral analysis of R-R interval and SBP was performed both on baseline data to reflect natural autonomic variability, and on head-up tilt data to assess autonomic responses during a cardiovascular stress. Spectral measures, in particular the LF power, of head-up tilt responses can provide information on the baroreflex mediated sympathetic outflow to the heart and vasculature²⁶ that are not readily obtained via plots of heart rate and blood pressure trends following the tilt stimulus. While ensamble-averaged plots of blood pressure and heart rate over time provide useful information on the pattern of change in response to head-up tilting, these plots tend to suppress the fluctuations seen in individual traces. Thus, we used both heart rate and blood pressure plots together with spectral analysis to assess these fluctuations in detail.

Frequency ranges:

Based on previous studies in infants, we defined the LF range between 0.04-0.15 Hz.²⁷ The HF range was individualized for each epoch based on individual infant respiratory rates, with the HF range being defined by the 90^th and 10^th centiles of the respiratory rate. HRV and BPV measures were calculated using an automated computer program (MATLAB 7b Mathworks; Natick MA). ECG R waves were identified using peak detection and then resampled at 200 Hz to create an evenly spaced times series. Each 1-2 min epoch was detrended and subdivided into 4 overlapping segments with 75% overlap, with a Hamming window used to reduce spectral leakage; the halving of measured power due to windowing was corrected for.

Data were decomposed into constituent frequency components by computing the average power spectral density functions for each segment. LF, HF, and total power for each epoch were calculated from the area under the power spectral density functions in the appropriate frequency range. These power values represent the square of the amplitude of an oscillatory signal in the frequency range; for example a BPV in the LF range of 4 mmHg² can be interpreted as a 2 mm Hg peak oscillation in the LF band.

Statistical Analysis

Statistical analysis was performed in Sigma Plot 11 (Systat Software Inc, Chicago IL). Data were first tested for normality and equal variance. The effects of postnatal age in each sleep state were tested using a one-way ANOVA, with Student Newman Keuls used for post hoc analysis to maximize the power of the analysis. Effects of sleep state were tested using a Student's paired t-test within each age group. Data are expressed as mean ± SEM. Effects were considered significantly different if P < 0.05.

RESULTS

Of the 31 infants studied, baseline measurements that satisfied the exclusion criteria were met in the following numbers of infants: n = 18 at 2-4 weeks, n = 25 at 2-3 months and n = 24 at 5-6 months during QS; n = 12 at 2-4 weeks, n = 14 at 2-3 months and n = 10 at 5-6 months during AS. The number of infants with movement-free head-up tilt measurements included were n = 22 at 2-4 weeks, n = 24 at 2-3 months and n = 27 at 5-6 months during QS; n = 17 at 2-4 weeks, n = 15 at 2-3 months and n = 17 at 5-6 months during AS.

Effects of Postnatal Age on HRV and BPV

Baseline measurements

Effects of increasing postnatal age on HRV during baseline conditions in both QS and AS are presented in Figures 1A-D. LF power (Figure 1A), HF power (Figure 1B), and total power (Figure 1D) for HRV increased with postnatal age in QS (P < 0.05 for all indices). In contrast, the LF/HF ratio decreased with postnatal age for HRV during QS (P < 0.05; Figure 1C). There were no effects of postnatal age on HRV indices identified during AS.

Figure 1.

Postnatal changes in HRV and BPV assessed under baseline conditions at 2-4 weeks, 2-3 months, and 5-6 months postnatal age during QS (black diamonds) and AS (white diamonds). Spectral indices for HRV were calculated for (A) low frequency power, LF RR; (B) high frequency power, HF RR; (C) low frequency/high frequency ratio, LF/HF RR; and (D) total power, Total RR. Spectral indices for BPV were calculated for (E) low frequency power, LF SBP; (F) high frequency power, HF SBP; (G) low frequency/high frequency ratio, LF/HF SBP; and (H) total power, Total SBP. LF power was calculated in a frequency range of 0.04-0.15 Hz, and HF power was calculated in a frequency range individualized for each infant's respiratory frequency.*P < 0.05.

Figures 1E-H present the effects of increasing postnatal age on BPV assessed during baseline conditions in QS and AS. For BPV, the LF/HF ratio decreased with advancing postnatal age during AS (P < 0.05; Figure 1G). There was a strong trend for LF power to decrease with postnatal age during AS (P = 0.056; Figure 1E).

Head-up tilt measurements

Average plots of the SBP and heart rate profiles during QS (Figure 2A-C) and AS (Figure 2D-F) at each age studied are presented in Figure 2. Typically, the response was biphasic, resulting in an initial rise in heart rate and SBP within the first 15 beats followed by a fall in SBP and compensatory rise in heart rate. Table 1 presents the effects of increasing postnatal age on HRV and BPV assessed during head-up tilt measurements in QS and AS. For HRV, LF power increased with postnatal age in QS (P < 0.05), and HF and total power increased in both QS (P < 0.05) and AS (P < 0.05). In contrast, the LF/HF ratio decreased with postnatal age during both QS and AS (P < 0.05). For BPV, the LF/HF ratio decreased with postnatal age in QS only (P < 0.05), with no effects of postnatal age being identified on any of the other spectral indices.

Figure 2.

Average beat-to-beat plots of heart rate (open circles) and SBP (closed circles) responses to head-up tilt during QS (left panel) at (A) 2-4 weeks, (B) 2-3 months, and (C) 5-6 months postnatal age and during AS (right panel) at (D) 2-4 weeks, (E) 2-3 months, and (F) 5-6 months postnatal age. The tilt is indicated by the dotted line. Data are expressed as the percentage change from the baseline value represented by the average of 30 beats prior to the tilt.

Table 1.

	HRV
	Postnatal Age	Quiet Sleep	Active Sleep
LF RR (ms2)	2-4 weeks	238 ± 58	491 ± 93
	2-3 months	400 ± 86	477 ± 70
	5-6 months	634 ± 111*	720 ± 146
HF RR (ms2)	2-4 weeks	81 ± 33	34 ± 15
	2-3 months	144 ± 42	58 ± 18
	5-6 months	439 ± 121*†	279 ± 98*†
LF/HF RR	2-4 weeks	10 ± 2	32 ± 7
	2-3 months	9 ± 2	20 ± 5
	5-6 months	3 ± 1†	7 ± 2*
TOTAL RR (ms2)	2-4 weeks	492 ± 117	665 ± 139
	2-3 months	896 ± 318	683 ± 96
	5-6 months	1449 ± 254*	1349 ± 326†
	BPV
	Postnatal Age	Quiet Sleep	Active Sleep
LF SBP (mmHg2)	2-4 weeks	4.5 ± 0.6	7.0 ± 0.9
	2-3 months	4.6 ± 0.5	8.0 ± 1.5
	5-6 months	3.6 ± 0.5	9.0 ± 1.7
HF SBP (mmHg2)	2-4 weeks	0.6 ± 0.1	1.0 ± 0.2
	2-3 months	0.7 ± 0.2	1.1 ± 0.3
	5-6 months	0.9 ± 0.1	1.8 ± 0.5
LF/HF SBP	2-4 weeks	15 ± 3	17 ± 4
	2-3 months	16 ± 4	14 ± 3
	5-6 months	7 ± 1*†	9 ± 2
TOTAL SBP (mmHg2)	2-4 weeks	6.1 ± 0.7	10.2 ± 1.5
	2-3 months	6.6 ± 0.9	11.4 ± 2.1
	5-6 months	5.8 ± 0.7	14.2 ± 2.4

Spectral indices for HRV and BPV assessed during Head-up Tilt in QS and AS at 2-4 weeks, 2-3 months, and 5-6 months postnatal age.

^*P < 0.05, 5-6 months vs 2-4 weeks;

^†P < 0.05, 5-6 months vs 2-3 months.

Effects of Sleep State on HRV and BPV

Baseline measurements

The effects of sleep state on HRV (Figures 3A-D) and BPV (Figures 3E-H) assessed during baseline measurements at each postnatal age are presented in Figure 3. The LF power (Figure 3A), LF/HF ratio (Figure 3C), and total power (Figure 3D) were generally higher in AS than QS at 2-4 weeks (LF power, P < 0.05 and total power, P < 0.05) and at 2-3 months (LF power, P < 0.05 and LF/HR ratio, P < 0.05). In contrast, for HF power (Figure 3B), no differences were identified between sleep states at either 2-4 weeks or 2-3 months, however at 5-6 months HF power was lower in AS than QS (P < 0.05). Similarly, BPV spectral indices were generally higher in AS than QS for all indices at all ages studied, though significant differences were only identified at 2-4 weeks for LF power (P < 0.05; Figure 3E) and total power (P < 0.05; Figure 3H).

Figure 3.

Sleep state differences in HRV and BPV assessed under baseline conditions at 2-4 weeks, 2-3 months, and 5-6 months postnatal age during QS (black bars) and AS (white bars). Spectral indices for HRV were calculated for (A) low frequency power, LF RR; (B) high frequency power, HF RR; (C) low frequency/high frequency ratio, LF/HF RR; and (D) total power, Total RR. Spectral indices for BPV were calculated for (E) low frequency power, LF SBP; (F) high frequency power, HF SBP; (G) low frequency/high frequency ratio, LF/HF SBP and (H) total power, Total SBP. LF power was calculated in a frequency range of 0.04-0.15 Hz and HF power was calculated in a frequency range individualized for respiratory frequency.*P < 0.05 QS vs AS.

Head-up tilt measurements

Figure 4 shows the effects of sleep state on HRV (Figure 4A-D) and BPV (Figure 4E-H) assessed during head-up tilt conditions at each postnatal age. Sleep state effects on both HRV and BPV were similar to those identified for baseline measurements. Overall for HRV, LF power, the LF/HF ratio and total power averaged higher in AS than QS, with this difference reaching significance for the LF/HF ratio at 2-4 weeks (P < 0.05) and 2-3 months (P < 0.05; Figure 4C). In contrast, HF power was lower in AS than QS at 5-6 months (P < 0.05; Figure 4B).

Figure 4.

Sleep state differences in HRV and BPV assessed under head-up tilt conditions at 2-4 weeks, 2-3 months, and 5-6 months postnatal age during QS (black bars) and AS (white bars). Spectral indices for HRV were calculated for (A)low frequency power, LF RR; (B) high frequency power, HF RR; (C) low frequency/high frequency ratio, LF/HF RR; and (D)total power, Total RR. Spectral indices for BPV were calculated for (E) low frequency power, LF SBP; (F)high frequency power, HF SBP; (G) low frequency/high frequency ratio, LF/HF SBP; and (H) total power, Total SBP. LF power was calculated in a frequency range of 0.04-0.15 Hz and HF power was calculated in a frequency range individualized for respiratory frequency.*P < 0.05 QS vs AS.

For BPV, all spectral power indices were higher in AS than QS; these differences were significant at 2-4 weeks (LF power, P < 0.05; HF power, P < 0.05; and total power, P < 0.05), 2-3 months (LF power, P < 0.05 and total power, P < 0.05), and 5-6 months (LF power, P < 0.05; LF/HF ratio, P < 0.05; and total power, P < 0.05) postnatal age.

DISCUSSION

This study is the first to provide combined data on BPV and HRV during both infant sleep states across the first six months of infancy.^4,7 The novel assessment of BPV in this study has revealed a low frequency dominance that reflects greater sympathetic modulation of blood pressure in early infancy, which subsequently decreases with postnatal age. At the same time, our HRV data confirm previous findings that cardiac sympathovagal balance shifts in favor of a strong parasympathetic dominance.^15,18 Furthermore, under both rest and stimulated conditions we have identified that sleep state has a marked effect on both HRV and BPV, where AS is described by a greater cardiac and vascular sympathetic presence.

Maturation of Autonomic Cardiovascular Control

Confirming previous findings,^15,18 our study identified that HF power of HRV increased with postnatal age during QS under baseline conditions. An increase in HF power with development reflects maturation of the parasympathetic nervous system and cardiorespiratory control. In support of this, studies performed longitudinally in preterm infants between 27 and 40 weeks post-conceptional age showed a reduction in LF/HF ratio as infants approach term equivalent age.¹¹ The current study identified that the LF/HF ratio continues to decrease with advancing postnatal age up until 6 months, reflecting an increase in parasympathetic contribution to heart rate with maturation. We also identified that LF HRV during baseline conditions, that reflect natural variability, increased with postnatal age in QS. This finding supports previous studies using time domain techniques, which also identified that changes in heart rate in the LF range increase from 1 month to 6 months.¹³ LF HRV reflects both sympathetic and parasympathetic contributions to heart rate variability, and therefore the interpretation of the increase in LF power with development is problematic. However, given that the HF HRV increased and the LF/HF power decreased with age, it is reasonable to assume that the increase in LF power predominantly reflects an increase in parasympathetic tone. Such an increase in parasympathetic activity with age is in keeping with the reduction in baseline heart rate that occurs with development. An increase in LF power with postnatal age also suggests that compensatory heart rate changes mediated by the baroreflex mechanism are minimal within the newborn period and become more prominent with advancing postnatal age.

We also employed the head-up tilt, a test that is commonly used to assess the sympathetic response to a hypotensive challenge.^9,10 An advantage of the head-up tilt is that it allows enhanced isolation of the sympathetic contribution to HRV within the LF range. In these healthy term infants, we identified a postnatal increase in LF and HF HRV in response to the head-up tilt, which suggests that, despite a clear increase in baseline parasympathetic modulation, the sympathetic response to an induced cardiovascular stress may also increase with postnatal age.

The novel aspect of this study was to provide data on BPV during the post-neonatal period in term-born infants. Previous assessment of resting BPV in preterm infants, aged 28 to 42 weeks during QS, revealed that preterm infants have greater total and HF power compared to term infants.¹⁷ In the present study, baseline BPV at rest showed a clear reduction in the LF/HF ratio with postnatal age. As a trend towards reduced postnatal LF and Total power was apparent, the reduction in the LF/HF can be mainly attributed to the fall in LF power with maturation. Reduced LF dominance in resting BPV is most typically ascribed to a reduction in sympathetic modulation of the vasculature,^10,23,24 potentially arising from either fewer “central commands” occurring in sleep²⁸ or a less reactive sympathetic vascular response.²⁹ An age-related reduction in natural, baseline sympathetic vascular tone is in keeping with animal studies showing that sympathetic tone is high during the transitional phase of circulatory adaptation to postnatal life, before reducing with age.³⁰ Reduced LF dominance of resting BPV with age is also in accordance with previous studies, by our group, that have shown that baroreflex sensitivity is low in the newborn period and then increases with postnatal age.²⁰ When taken together, these observations are likely to reflect a shift in strategy from sympathetic vascular control of blood pressure to that of predominantly parasympathetic cardiac control. Such a view is consistent with our use of a tilt challenge to elicit a sympathetic vascular reaction, which also demonstrated a postnatal reduction in LF/HF BPV at 5-6 months, in association with a progressively greater overall cardiac response.

The specific elements of the neural-humoral control mechanisms that may be responsible for the developmental changes of autonomic function are unclear. Developmental changes could occur within afferent/efferent pathways, central integration centers involved in cardiovascular control, or end-organ (heart and vasculature) maturation. For example, neuronal maturation of the several nuclei within the brainstem responsible for cardiorespiratory control after birth has been observed.³¹ Specifically, normal maturational loss of dendritic spines occurring within the dorsal vagal nucleus, the solitary nucleus, and the medullary reticular formation occurs with increasing postnatal age.³² Furthermore, postmortem studies suggest that active myelination of the vagus nerve occurs within the first 9 months of life, and a slow transition from unmyelinated to myelinated fibers during the first year of life, particularly during the first 3 months also occurs.³³ However, as the present study did not explore the maturation of the neural circuitry, it is difficult to speculate where or when these exact developmental changes occur. Previous studies have described developmental changes in HRV within the first month of life and have shown that HRV decreases from one week to one month postnatal age, after which it increases again, particularly during QS.¹³ Based on these early neonatal changes, it is thought that reorganization of central autonomic control mechanisms may occur during the first month after birth, when parasympathetic contribution is low and sympathetic contribution is possibly high. As this study was designed to surround the peak incidence of SIDS, we did not assess cardiovascular control at these early postnatal ages. Despite this, the data do indicate that within the neonatal period, autonomic regulation of both the heart and circulation is immature and undergoes functional maturation within the first 6 months of infancy.

Effects of Sleep State

Our study confirms previous findings that resting HRV is increased during AS compared to QS in term infants^13,18,34 and that resting HF spectral power is reduced in AS compared to QS.¹⁸ In complementary findings, we have also shown that LF and total spectral power is higher in AS, particularly in early postnatal life. These findings suggest that during QS, autonomic cardiac control is dominated by parasympathetic tone in older infancy, while in AS it is dominated by sympathetic activity in early infancy. These findings in infants are in accordance with adult studies, which have also shown that parasympathetic activity is dominant in NREM compared to REM sleep (which are the mature forms of infant QS and AS, respectively).³⁵ To our knowledge, beat-beat variability of blood pressure has only been studied previously at one age (3 months) in the time domain in term infants; in accordance with our spectral analysis, this study also showed that BPV was higher in AS than QS.³⁶ Using spectral techniques our study has extended this work by demonstrating that autonomic blood pressure control is sleep-state dependent from birth. We identified that the LF power and total power of BPV were strikingly lower in QS than AS, both at baseline (2-4 weeks) and head-up tilt conditions (at all ages studied). The reduced LF BPV in QS versus AS leads us to conclude that sympathetic modulation of vascular tone is reduced in amplitude in QS. Alternatively, a reduced baroreflex sensitivity in AS might provide for less effective damping of blood pressure fluctuations in AS compared with QS²²; however, this explanation is unlikely since our previous study revealed that baroreflex sensitivity was only reduced in AS at 5-6 months. Additionally, during the head-up tilt challenges, greater LF/HF ratio in AS at 5-6 months is also in keeping with sympathetic dominance.

Clinical Implications

An understanding of the normal maturation of both HRV and BPV in infants, as provided by this study, is fundamental to a fuller understanding of pathologic states where impaired autonomic control has been implicated. Specifically, normative data on HRV and BPV may provide reference values for future studies of autonomic impairment in infancy, and possibly of SIDS. Impaired autonomic control and circulatory failure is thought to be one of the underlying mechanisms involved in SIDS,^4,6,7,37 and future SIDS victims have reduced respiratory sinus arrhythmia compared to aged-matched controls, suggesting a reduced parasympathetic activity or a shift from parasympathetic to sympathetic dominance prior to death.³⁸ Furthermore, impaired development of cardiovascular control in infancy may explain the predisposition of preterm infants to significant cardiovascular consequences later in life,^2,3 as epidemiological studies have identified that infants born preterm have increased risk for developing high blood pressure in adulthood.³⁹ Comparison of BPV data of term-born infants of this study with those of preterm infants gathered in future studies may provide insight into this predisposition.

CONCLUSIONS

This is the first study to describe the maturational changes of autonomic control on both heart rate and blood pressure during sleep in infants. The data confirm that parasympathetic dominance on heart rate control increases with postnatal age, and is the first to demonstrate that sympathetic modulation of blood pressure is higher in newborns compared to older infants. Understanding the normal development of cardiovascular variability will aid in the early identification of conditions in which autonomic function is impaired, such as in preterm infants or infants at increased risk for SIDS.

DICSLOSURE STATEMENT

This was not an industry supported study. The authors have indicated no financial conflicts of interest.

ABBREVIATIONS

SIDS

sudden infant death syndrome

HRV

heart rate variability

BPV

blood pressure variability

QS

quiet sleep

AS

active sleep

ECG

electrocardiogram

SBP

systolic blood pressure

LF

low frequency

HF

high frequency