Main

Antenatal glucocorticoid treatment for lung maturation in pregnancies at risk for preterm delivery is an established method to reduce neonatal mortality and morbidity (1). However, there is accumulating evidence that glucocorticoids may alter maturational development and homeostasis of fetal regulatory systems. Repeated antenatal glucocorticoid exposure results in reduced birth weight, increased blood pressure, and glucose levels in adult offspring in different animal models (2).

The cardiac autonomic system has been shown to be sensitive toward antenatal glucocorticoids. Accordingly, fetal heart rate (FHR) variation is altered at least transiently in human fetuses after antenatal glucocorticoids (3,4), and infant heart rate answer to a stress event is significantly increased (5). Putative mechanisms include an altered cardioregulatory innervation and signaling that seem to happen at central levels and at the level of the cardiomyocyte according to rodent models (6,7).

Cardiovascular system regulation strongly depends on sympathetic autonomic control that is believed to play an important role in the pathogenesis of cardiovascular disease in the adult (8,9). Long-term observations in adolescents and young adults with antenatal exposure to glucocorticoids have rather focused on classic signs of disease manifestation such as blood pressure (10,11). However, risk factors for cardiovascular or metabolic disease may transform into apparent clinical manifestation at a more advanced age when stability of these systems can be less compensated by the organism. Therefore, we focused on putative predictive methods characterizing sympathoadrenergic cardiac balance. Heart rate variability (HRV) is a well-established noninvasive measure of cardiac autonomic control (12,13) that is regulated by a complex interplay of sympathetic and parasympathetic branches of the autonomic nervous system. The aim of this study was to analyze the cardiac autonomic balance of healthy newborns at or near term who had received a single course of antenatal betamethasone treatment for imminent preterm delivery before 34 wk of gestation and in whom delivery was delayed for a median of more than 7 wk. Electrophysiological HRV data were supplemented by functional salivary α-amylase measurements in response to a stress stimulus. Salivary α-amylase has been suggested to be a surrogate for cardiovascular autonomic system balance correlating well with HRV parameters (14,15).

METHODS

The study was approved by the Research Ethics Committee of the University of Zurich and conforms with the Declaration of Helsinki. Written parental consent was obtained from all participants.

Healthy infants born after 34 wk of gestation (>238 d postmenstruation) were included in the study. Exclusion criteria were infant intensive care treatment or invasive procedures, malformations, maternal substance abuse (nicotine and alcohol) during pregnancy, or insufficient amounts of saliva collected. No signs of clinical apparent infections were present in both groups. Sonographic measurements of fetal crown-rump length and biparietal diameter during the first trimester served as parameters for accurate determination of gestational age. Because mothers did not always give consent for both ECG and saliva sample collection, the populations were analyzed separately.

HRV measurements.

A cohort of 23 infants with antenatal betamethasone treatment was recruited for 24 h Holter ECG measurements (treatment group). The control group of 23 infants without antenatal betamethasone exposure is part of a cohort previously published (16). A summary of infant and maternal characteristics is shown in Table 1. Three-channel Holter monitors (Lifecard CF; Delmar Reynolds Medical, Hertford, United Kingdom) were placed within the third to fourth postnatal day. Ectopic beats, noisy data, and artifacts were manually identified and excluded from the HRV analysis. For calculation of HRV parameters, the HRV Analysis Software, version 9.3.0 from Nevrokard (www.nevrokard.eu) was applied.

Table 1 Neonatal and maternal baseline characteristics for HRV measurements and salivary α-amylase measurements
Full size table

To account for long-term and for short-term HRV variations, the total monitoring period of 24 h was analyzed for time-domain parameter calculation, and selected 5-min segments were analyzed for frequency-domain parameter calculation to account for different states of activation. According to the recommendations of the Task Force of the European Society of Cardiology and North American Society of Pacing and Electrophysiology (17) and the literature for neonatal HRV measurements (18–21), the following parameters were analyzed.

Time-domain parameters [calculated from the total monitoring period (24 h)].

1) Parameters as estimate of overall HRV: the SD of all valid NN intervals (SDNN), 2) parameters mostly influenced by parasympathetic activity: The ratio of the number of all pairs of adjacent NN intervals differing by >50 ms and the total number of RR intervals (%) (s-NN50), >27 ms (s-NN27) and >20 ms (s-NN20), the average of the hourly square root of the mean of the sum of the squares of differences between adjacent NN intervals (RMSSD), 3) parameters mostly influenced by sympathetic activity: The SD of the average of valid NN intervals (SDANN) in 5-min segments in the recording, the average of the hourly means of SDs of all NN intervals (SDNNi) in 5-min segments. As geometric index, the HRV triangular index, defined by the total number of all NN intervals divided by the height of the histogram of all NN intervals measured on a discrete scale with bins of 7.8125 ms as estimate of overall HRV was calculated.

Frequency-domain parameters (5 min segments).

Power spectral analysis was calculated by fast Fourier transformation (Hamming window) in three main spectral components. To account for the neonatal physiology, frequency bandwidths were adjusted according to the literature (18,19): high frequency (HF): 0.24 to 1.04 Hz, representing parasympathetic activity; low frequency (LF): 0.04 to 0.24 Hz, representing both, sympathetic and parasympathetic activity; and very low frequency (VLF): 0.003 to 0.04 Hz with a less-defined physiologic role. Total power (ms2) and the ratio of LF/HF considered as a marker of sympathetic-parasympathetic system balance were analyzed. Absolute values (ms2) and normalized units (n.u.) that represent the relative value of each power component in proportion to the total power minus the VLF component were calculated. The presentation of LF and HF in n.u. emphasized the controlled and balanced behavior of the two branches of the autonomic system (17).

α-amylase measurements.

Salivary α-amylase levels were analyzed from a cohort of 23 infants with antenatal betamethasone treatment and from a cohort of 40 controls that have been published previously (22). A summary of infant and maternal characteristics is shown in Table 1.

Samples were collected using a routinely performed blood sampling (heel prick test) 72 to 96 h postpartum as pain-induced stress factor. This procedure has been shown to be a significant stressor for the newborn (23,24), and salivary α-amylase has been suggested as a measure of endogenous adrenergic activity and changes in the autonomic nervous system in general (15,25) and specifically for cardiac autonomic balance (14,15,26). Saliva samples were collected from each infant 10 min before and 5 and 20 min after stress induction. Collection time was based on experiments revealing peak α-amylase responses between 5 and 10 min poststress induction (15,27). A cotton swab was placed in the neonates' mouth for a collection time of 5 min. Samples were placed in saliva collection tubes (Salivette; Sarstedt, Nümbrecht, Germany) and stored frozen at −20°C until further analysis.

The amylase 4,6-ethylidene-p-nitrophenyl-α,d-maltoheptaoside (EPS) method from Roche Diagnostics, Mannheim, Germany, was applied for the measurement of the α-amylase concentration in saliva. The diluted saliva samples (1 + 9) were analyzed on the Integra system 800. The assays showed good performance characteristics [intraassay coefficient of variation (CVs) < 1.0% and interassay CVs ≤ 1.3% at concentrations of 79.9 and 198 U/L].

Statistical analysis.

All statistical analyses were performed with STATA 10 Statistics/Data Analysis Software (Stata Corporation, College Station, TX). Baseline characteristics of treatment and control infants were compared using the Mann-Whitney test and χ2 test when appropriate. Because HRV parameters were not normally distributed as analyzed by the Shapiro-Francia W' test, we compared groups using the Mann-Whitney test. α-Amylase data were log transformed. The Mann-Whitney test was used for comparison of baseline α-amylase levels. Alterations of log-transformed data between baseline and poststimulation levels were analyzed by the Wilcoxon signed-rank test. To account for within-subject variation, a two-way ANOVA for repeated measurements was conducted. A stepwise multiple regression was applied to analyze the impact of gestational age at steroid administration, birth weight percentile, and gestational age at delivery as putative influencing factors on α-amylase values. The level of statistical significance of all analyses was set at p < 0.05.

RESULTS

Infant HRV.

Betamethasone treatment was conducted at a median gestational age of 219 (178–238) d (corresponding to 31 2/7 wk of gestation), and the median interval between treatment and delivery was 51 (25 to 108) d. Median birth weight and weight percentile were comparable between treatment and control groups (3040 g and 50th percentile versus 3260 g and 54.8th percentile). Infants of the treatment group were delivered ∼2 wk earlier [260 d (37 1/7 wk of gestation) versus 275 d (39 2/7 wk of gestation], just not reaching statistical significance (p = 0.0504). Infant head circumference and 5-min Apgar levels were similar (p = 0.740 and p = 0.757, respectively). A summary of the characteristics of the study population is depicted in Table 1.

For the long-term HRV analysis, the time-domain parameters SDNN and the triangular index representing estimates for overall HRV were not significantly different in the treatment group compared with the controls (p = 0.258 and p = 0.179, respectively). Similarly, parameters, mostly influenced by sympathetic activity, such as SDANN and SDNNi were similar (p = 0.184 and p = 0.784). No significant differences were also found for estimates of vagal tone as RMSSD (p = 1.0) and for the proportion of pairs of adjacent NN intervals differing by >50, >27, and >20 ms (p = 0.852, p = 0.684, and p = 0.818; Table 2). Short-term HRV analysis during active and resting states calculating frequency-domain parameters confirmed these results. Although there was a significant difference between active and resting states within each group as expected (data not shown), comparison of parameters between similar states of activation were not significantly different when calculating absolute and normalized values. Accordingly, HF, representing vagal activity; LF, representing both sympathetic and vagal activity; VLF, representing total power; and LF/HF, representing sympathovagal balance, all were not significantly different from control infants (Table 3). The likelihood to miss a putative significant difference by our sample size is very low [5.2% (α = 0.8, β = 0.05) for RMSSD and LF/HF (n.u.) resting state].

Table 2 Long-term time-domain HRV parameters in control and treated (betamethasone) group
Full size table
Table 3 Short-term frequency-domain HRV parameters in control and treated (betamethasone) group
Full size table

Infant α-amylase levels.

Betamethasone treatment was conducted at a median gestational age of 201 (173–235) d, and the median interval to delivery was 55 (16–102) d. Median birth weight and weight percentile were comparable in both groups (p = 0.269 and p = 0.617, respectively). Gestational age at delivery was 266 (242–284) d in the treatment group and 273 (240–294) d in the control group (p = 0.201). Head circumference and 5-min Apgar scores were similar (p = 0.583 and p = 0.517, respectively). A summary of the characteristics of the study population is shown in Table 1.

Median baseline levels for α-amylase were slightly higher but not significantly different in the treatment group when compared with controls (p = 0.3040, Fig. 1). α-Amylase levels in control infants increased marginally, just reaching the level of statistical significance (p = 0.045) at the time point 5-min poststimulation, whereas in infants with betamethasone treatment, no significant alteration of α-amylase release was observed (p = 0.408). A two-way ANOVA confirmed that there was a significant alteration of α-amylase in the control (p = 0.0266) but not in the study group (p = 0.183) after the stress stimulus. No significant influence of gestational age at steroid administration (p = 0.82), newborn weight (p = 0.079), and gestational age at delivery (p = 0.46) was observed applying a multiple stepwise regression model.

Figure 1
figure 1

Individual log-transformed salivary levels and medians for α-amylase in control (n = 40) and study group (n = 23) newborns before (baseline) and after (5 and 20 min after) application of the stress stimulus. (left = control; right = study group). Horizontal line indicates median levels. *p < 0.05.

Full size image

DISCUSSION

We have shown that cardiac autonomic system balance is preserved in the neonate after a single course of antenatal maternal betamethasone treatment for fetal lung maturation in pregnancies at risk for preterm delivery.

There is experimental evidence that intrauterine exposure to synthetic glucocorticoids affects balance and activity of the autonomic cardiovascular system. In adult male guinea pigs with repeated prenatal exposure to dexamethasone, arterial blood pressure was significantly elevated (28). Even a single course of glucocorticoid treatment produced a sustained elevation of blood pressure and altered baroreceptor heart rate response in fetal sheep and the baboon (29–32) and increased central and peripheral vascular resistance (29,30,33). However, others only found transient hypertension after maternal dexamethasone treatment in fetal sheep (34). In the nonhuman primate, multiple antenatal dexamethasone exposure resulted in elevated systolic and diastolic blood pressure at the age of 12 mo (35).

The connection of prenatal glucocorticoid treatment and increased blood pressure in the human is less clear. In the neonate (36) and in a cohort of preterm children at the age of 14 y (37), blood pressure was significantly increased, whereas other data have shown that blood pressure alterations may rather be transient and normalize within the first few weeks of life (38). Large follow-up studies did not find significant alterations in blood pressure in nonstrained conditions at the age of 20 and 30 y (10,11).

Data addressing more specifically the cardiac autonomic system reveal persistent abnormalities of cardiac noradrenergic innervation after intrauterine exposure to exogenous steroids in the rat (6) and alterations of cardiac intracellular β-adrenoceptor signaling (7). Furthermore, centrally located systems involved in cardioregulation such as the brainstem norepinephrine system seem to undergo premature maturation in response to prenatal synthetic glucocorticiod exposure (39). There is some evidence that betamethasone may even have a direct effect on the cardiac pacemaker (40,41). It has been shown that infants with antenatal betamethasone treatment display a significant increase in heart rate in response to a stressor when compared with control infants (5).

To our knowledge, a systematic analysis of HRV parameters according to standardized methods (17) so far has not been performed in individuals exposed to prenatal glucocorticoid treatment. HRV is a well-established noninvasive measure of cardiac autonomic control that has been shown to be related to hypertension (12,13) and to predict future adverse cardiovascular events in adults (42). HRV analyses have been suggested for putative prognostic use in children (43). We did not find significant alterations of any HRV parameter analyzed in neonates after antenatal betamethasone treatment. Studies measuring intrauterine FHR variation after steroid administration have produced inconsistent results. Mulder et al. (3), analyzing FHR short-term and long-term variation, found reductions in FHR variation after maternal betamethasone administration. In contrast, treatment with dexamethasone resulted in a significant rise in short-term FHR variation (4). However, both effects were only transiently observed.

We conducted our study in the neonatal period because our aim was to find out whether putative intrauterine steroid effects ultimately transform into permanent postnatal alterations of the cardiac autonomic system and to exclude postnatal-induced influencing factors of system adaptation. This aspect is even more important in consideration that diseases susceptible to intrauterine origin such as hypertension and type 2 diabetes will exhibit multiple system alterations at the time when they become apparent. For the development of prevention programs for arterial hypertension and type 2 diabetes, the initiating systems have to be identified and discriminated from systems secondarily affected. In this study, putative short-term effects could be excluded as the median interval between betamethasone treatment and delivery was more than 7 wk. Our findings provide new evidence that at the level of the cardiac autonomic subunit, at least, a single course of betamethasone does not primarily affect sympathetic cardiac balance.

As early effects may only become apparent during events with increased demand for adaptive regulation, we supplemented our electrophysiological studies with a neuroendocrinological approach by analyzing salivary α-amylase levels during resting conditions and after a stressful stimulus. Salivary α-amylase is secreted by acinar cells in the salivary glands that are richly innervated by both sympathetic and parasympathetic nerve fibers, influencing the release of α-amylase by classic neurotransmitters (44). Studies in humans and animals have suggested that the activation of the autonomic nervous system leads to a high activity of salivary α-amylase (25,45–47). Furthermore, α-amylase levels have been found to be associated with cardiovascular physiology and are suggested to be a surrogate for cardiovascular autonomic system balance (26). Bosch et al. (14) found a significant negative correlation between the parasympathetic-influenced HRV parameter RMSSD and α-amylase levels during stress induction in adults. Furthermore, a positive correlation between α-amylase levels and LF/HF as surrogate for sympathetic tone has been shown (15), thereby making this parameter a promising indicator for cardiac autonomic function. Although studies on cardiovascular autonomic physiology using α-amylase measurements have not been validated in neonates, several analyses from children support the strong relationship between salivary α-amylase and sympathetic/parasympathetic nervous system activation in younger individuals (26). Interestingly, we found stimulated α-amylase levels in betamethasone-exposed neonates rather decreased when compared with controls although statistical significance was only marginal (p = 0.045). One rather had expected increases in α-amylase levels as response to an increased sympathetic stress reaction if betamethasone treatment had affected the system.

The slightly attenuated α-amylase response might be explained by previous findings about the stress-regulating hypothalamo-pituitary-adrenal (HPA) axis. Indeed, we and others have shown that a single course of antenatal betamethasone administration causes sustained suppressive effects of HPA reactivity in these infants (5,22,48). We speculate a nonsignificant increase in α-amylase levels may be the result of inadequate HPA system sensitivity rather than direct effects of steroid exposure on the sympathetic system because these systems are closely interconnected. A stepwise multiple regression was conducted to exclude putative influencing factors such as gestational age at steroid administration, newborn weight, and gestational age at delivery, not revealing a significant influence on α-amylase results.

We analyzed a relatively homogenous collective excluding isolated effects of severe prematurity, pregnancy complications such as intrauterine growth restriction or preeclampsia in addition to maternal substance abuse and diseases. Nevertheless, although not statistically significant, there was a difference of 15 d in gestational age between the control and the treatment group. This may, even though near or at term, entail a difference in perinatal maturation with a putative influence on the cardiac autonomic system. It has been shown that with increasing maturation of the sympathetic nervous system during the course of pregnancy, there is a shift of the sympathetic/parasympathetic balance toward parasympathetic tone (49–51). The time-domain parameter RMSSD has been shown to demonstrate highest correlation to gestational age (51), and for frequency-domain parameters, gestational age-dependent alterations of LF and HF parameters have only been found to be significant during resting and not during active states (49). Therefore, we analyzed our data for a putative influence of gestational age on these parameters and conducted a regression analysis. We found adjusted R-squared values for RMSSD of 0.10, for LF/HF (resting state) of 0.00, and for LF/HF (active) of 0.02, indicating no relevant effect of the difference in gestational age in our groups. Thus, our data strongly support the notion that at least the cardiac autonomic subunit of the sympathetic system is not permanently affected in utero by antenatal steroid treatment for lung maturation.

It is important to note that the sympathetic nervous system is composed of multiple function-specific subunits (52) and the programming of sympathetic nervous system function is believed to occur regionally rather than on a global basis, suggesting that subdivisions of the sympathetic nervous system may be influenced by different sets of environmental variables (53), thus aspects such as vascular bed regulation may be affected differently.

We cannot exclude that our methods were not sensitive enough to expose small alterations in cardiac sympathetic system balance that may only become apparent with increasing system maturation during infant and adolescent development. Even though long-term analyses of antenatal steroid treatment in young adults did not find sympathoadrenergic connected pathologies (10), these diseases may not be present until later ages at the time they usually become apparent. In fact, putative early markers of sympathetic cardiac activation such as HRV parameters have not been applied in these studies.

In conclusion, despite evidence from animal models, it seems that the neonatal cardiac autonomic system is primarily unaffected by a single course of betamethasone treatment for lung maturation. However, the impact of multiple doses is not known.