Main

The human kidney displays a tubular immaturity at birth, with sodium waste, responsible for a negative sodium balance, and an impaired ability to reabsorb water (1), which is aggravated under circumstances such as prematurity, pyelonephritis, and gastroenteritis (2). This inability to maintain homeostatic functions is a major problem encountered by pediatricians, most notably in preterm infants. Indeed, the challenge is to find a balance between the lethal risk of dehydration and the associated morbidity of excessive hydromineral supplementation, contributing to severe complications such as intraventricular hemorrhage and bronchopulmonary dysplasia (3). A better understanding of water and sodium regulation in the neonatal period is a prerequisite to propose new therapeutic strategies for the management of preterm infants.

Sodium reabsorption is mainly controlled by aldosterone, a steroid hormone synthesized in the zona glomerulosa of the adrenal gland, secondary to renin stimulation via angiotensin II and to potassium stimulation (4). In the distal nephron, aldosterone, by binding to its receptor, the mineralocorticoid receptor, a transcription factor (5), tightly regulates the expression and the activity of several transport proteins implicated in sodium, potassium, and water homeostasis, as the alpha subunit of the epithelial sodium channel (6), the Na-K-ATPase (7), and the aquaporin 2 (8). A possible role of a partial and transient tubular unresponsiveness to this hormone has been suggested to account for sodium waste in neonates (9,10).

Indeed, high levels of plasma aldosterone have been reported in the neonatal period throughout the first year of life (10–12), and several cases of transient pseudohypoaldosteronism have been reported in infants <3 mo of age with urinary infection (13,14). This suggests that a preexistent partial hormonal resistance in young infants might aggravate the impaired renal aldosterone sensitivity secondary to urinary infections. This putative aldosterone resistance has been already proposed by many authors (9,10) but has never been fully characterized. This is easily understandable given the difficulty to perform clinical investigations in the neonatal period, owing to ethical considerations concerning blood samples and urine collections in the newborn infant. Few studies have reported plasma aldosterone and renin levels at birth (9,15,16), but no accurate and reliable reference values are available for either plasma aldosterone and renin concentrations or urinary aldosterone levels in newborns.

Herein, we report the first large prospective study evaluating aldosterone sensitivity in newborn infants. We measured aldosterone and renin levels in both maternal and umbilical cords plasma, together with electrolytes concentrations, and compared these values with neonatal urinary aldosterone levels, sodium and potassium excretion. We established neonatal reference values for plasma aldosterone and renin and most notably for urinary aldosterone concentration (Aldou), which seems to be the best index for accurate evaluation of aldosterone sensitivity. We have thereby demonstrated aldosterone unresponsiveness at birth.

PATIENTS AND METHODS

Patients.

Forty-eight mother-neonate couples were included in the maternity ward of Antoine-Beclere Hospital, Clamart, France. Pregnant women aged 18- to 45-y-old (maternal age mean ± SEM = 31.9 ± 0.6) were eligible for this study. The gestational age was determined by a first-trimester ultrasound scan (gestation week = 39 wk 5 ± 1 d). Pregnancies were uncomplicated for diabetes, hypertension, eclampsia, adrenal or pituitary insufficiency, or other chronic diseases. Singleton pregnancies were normal without any tocolytic treatment given within 15 d before delivery. No fluid had been administered to mothers prior vaginal delivery, which had to be spontaneous and eutocic. Full-term newborn infants (birth weight = 3421 ± 69 g) with no respiratory disease or sign of per-partum anoxia (Apgar score <5 at 5 min or cord pH <7.10 or blood lactate >6 mM) were subsequently enrolled. Exclusion criteria were abnormal antenatal ultrasound scan, prenatal corticosteroid treatment, and small for gestational age. Informed and written consent was obtained from all mothers. The study was conducted in accordance with the Declaration of Helsinki and after approval of the local ethical committee (CCPPRB; Comité Consultatif Pour les Personnes en Recherche Biomédicale, Hôpital Antoine-Béclère, Clamart, France).

Blood and urinary samples.

Maternal blood samples were obtained during preanalgesia peridural checkup, the day of delivery. Immediately after delivery, neonatal venous blood samples were collected from umbilical cords on EDTA and heparinized tubes and immediately transferred to laboratory for centrifugation. Supernatants were stored at −20°°C. Single-spot urinary samples were collected onto a gauze compress settled in the diaper during the first 24 h of life. Compresses not contaminated by meconium were stored at −20°C before subsequent processing. A 24-h urinary collection was not performed for technical as well as ethical reasons.

Hormonal and biochemical analyses.

Plasma aldosterone and renin concentrations as well as electrolytes, proteins, urea, and creatinine levels were measured in maternal and umbilical cord blood. Hemolyzed samples were excluded from the study. Urine samples extracted from compresses were used to determine aldosterone, electrolytes, and creatinine concentrations.

Active plasma renin was assessed using the Renin III generation RIA kit (Cisbio, Gif sur Yvette, France) with a sensitivity of 2.5 pg/mL. The intra-assay and interassay coefficients of variation were 4.5% and 7.4%, respectively. Renin concentrations in plasma of healthy subjects (aged 20-40 y) were 8.1 ± 3.7 (n = 50; range, 3–16) in supine position. Plasma aldosterone concentrations were measured with a RIA (Aldo-Riact, CIS-Bio International, Gif sur Yvette, France) with a sensitivity limit of 7 pg/mL. The intra-assay and interassay coefficients of variation were 6.7% and 8.4%, respectively. Aldosterone was assayed in urine samples after acid hydrolysis of aldosterone 18-glucuronide during 18 h at 30°C. Measurements on random urine samples required a correction of urinary output of aldosterone for creatinine excretion. Electrolytes, bun, total proteins, and creatinine were measured with the automat Modular P 900 (Roche, France). Urinary creatinine concentrations were determined using a rate-blanked Jaffé based method (Roche Diagnostics). The fractional excretion of sodium (FeNa) was defined as the equation = (plasma creatinine × urinary sodium)/(plasma sodium × urinary creatinine) %.

Statistical analyses.

Results are expressed as mean ± SEM. Statistical analyses were performed using a nonparametric Mann-Whitney test. Correlations between two parameters within one sample or between newborn infants and their respective mothers were obtained by Spearman regression analysis (Prism4, Graphpad Software, Inc., San Diego, CA), with significant threshold at 0.05.

RESULTS

Forty-eight pairs were included in the study. We obtained 48 maternal and cord blood samples for hormonal dosages and 47 maternal and cord blood samples for biochemical analyses. Twenty-eight urinary samples were obtained from these 48 newborn infants and permitted to measure both aldosterone and electrolyte concentrations.

Distinct fetal and maternal compartments.

We first compared electrolyte concentrations obtained from maternal blood samples and those obtained from the umbilical cords of their respective newborn. Neonatal and maternal plasma sodium concentrations were in the same range, with mean values of 132.6 ± 0.7 and 132.3 ± 0.7 mM, respectively (Table 1), but surprisingly no correlation was observed between these two parameters (data not shown). This indicates that neonatal natremia is independent from maternal natremia that may be related to pregnancy hemodilution (17) or mineralocorticoid antagonism by high levels of progesterone during late gestation (18). Furthermore, plasma potassium concentrations were significantly higher in newborn infants than in their respective mothers, with mean values reaching 5.7 ± 0.3 versus 3.7 ± 0.1, p < 0.001 (Table 1 and Fig. 1A). Once again, no correlation was found between neonatal and maternal kalemia (Fig. 1A and B), even after excluding the extremely high kalemia values obtained from nonhemolyzed samples of two newborn infants. Plasma protein concentrations also significantly differed, underlining that maternal and fetal compartments are distinct and in part submitted to independent regulations (Table 1). On the other hand, neonatal plasma urea and creatinine concentrations at birth strictly reflected maternal levels, as demonstrated by the very high correlation between maternal and neonatal values (Fig. 1C), in accordance with the previous studies (19). Overall, these results suggest that although creatinine and urea freely diffuse through the placenta during pregnancy, ionic homeostasis in maternal and fetal compartments is at least in part controlled by distinct regulatory mechanisms. Moreover, our findings demonstrate that healthy newborn infants have well-tolerated physiologic hyponatremia and hyperkalemia, which are highly suggestive of a functional hypoaldosteronism.

Table 1 Maternal and neonatal plasma concentrations of various parameters
Full size table
Figure 1
figure 1

Comparison between maternal and neonatal electrolyte concentrations. (A) Plasma potassium concentrations in umbilical cords and maternal blood samples. Each dot represents one value and lines correspond to means of 47 values. Kalemia is significantly higher in newborn infants than in their mothers *p < 0.001. (B) Correlation between maternal and neonatal kalemia. No statistical significance was reached p = 0.38. (C) Correlation between maternal and neonatal plasma creatinine concentration; p < 10−4, r = 0.55 represents the correlation coefficient. (B, C) Correlations between two parameters were obtained by Spearman regression analysis.

Full size image

Renal tubular immaturity in newborns.

Urine samples of newborn infants were collected onto a gauze compress settled in the diaper during their first 24 h of life. This simple, original, and noninvasive method of urine collection was very efficient and allowed us to obtain 200 μL up to 5 mL urinary samples from 28 neonates and enabled us to measure both urinary electrolyte concentrations and steroid hormone excretion. Mean urinary sodium (Nau), potassium (Ku), and creatinine (creatu) concentrations were 40.2 ± 3.8 mM, 32.5 ± 3.2 mM, and 5168 ± 637 μM, respectively. Sodium waste was evaluated by the FeNa, by the Nau/creatu ratio and by the Nau/Ku ratio. Mean FeNa was 0.7 ± 0.3%, mean Nau/creatu was 19.6 ± 8.3, and mean Nau/Ku was 1.99 ± 0.46. When Nau/Ku was compared with urinary concentration, defined by urinary creatinine concentration, a highly significant negative correlation was found (Fig. 2), suggesting a renal tubular immaturity at birth. This clearly indicates that alteration in urinary concentration ability is associated with compromised sodium reabsorption, the latter being in accordance with a possible aldosterone resistance in newborn infants.

Figure 2
figure 2

Neonatal renal immaturity. Correlation between urinary creatinine concentration and sodium/potassium urinary ratio (Nau/Ku) measured in 28 newborn infants. Correlation was obtained by Spearman regression analysis. p < 10−3 and correlation coefficient r = 0.61.

Full size image

Hyperaldosteronism and hyperreninism in newborn infants.

Despite biochemical signs of functional hypoaldosteronism at birth, high plasma aldosterone levels were detected in neonatal samples with mean values of 817.1 ± 77.9 pg/mL, significantly higher than those measured in maternal blood (mean values of 574.7 ± 54.5 pg/mL, p < 0.05) (Table 1 and Fig. 3A). However, when compared with values for nonpregnant 20- to 40-y-old female adults (99 ± 43 pg/mL; range, 42–201), the maternal aldosterone values were elevated, as already described during late pregnancy, associated with increased plasma progesterone levels (18). Even though neonatal and maternal aldosterone concentrations were correlated (p < 10−4, r = 0.57, data not shown) consistent with a partial transplacental passage of aldosterone from the mother to the fetus, as reported previously (20), the significant difference between neonatal and maternal mean aldosterone values provides support to a fetal neosynthesis of aldosterone. In addition, a highly significant hyperreninism was also observed in newborn infants (78.8 ± 9.6 pg/mL) compared with maternal values (14.9 ± 1.5 pg/mL) (Table 1 and Fig. 3B). Finally, as illustrated in Fig. 3C, the high positive correlation between neonatal plasma aldosterone and renin conclusively demonstrates that the renin-angiotensin-aldosterone system was strongly stimulated at birth.

Figure 3
figure 3

Hypersecretion of aldosterone and renin in newborn infants. (A) Plasma aldosterone concentrations in umbilical cords and maternal blood samples. (B) Plasma renin concentrations in umbilical cords and maternal blood samples. (A, B) Each dot represents one concentration value and lines correspond to means of 48 values. **p < 0.01 and *p < 0.001. (C) Correlation between neonatal plasma aldosterone and renin concentrations. Correlation was obtained by Spearman regression analysis. p = 0.008 and correlation coefficient r = 0.38.

Full size image

Aldou is a better index to evaluate mineralocorticoid sensitivity.

From our neonatal urine samples, we were able to quantify Aldou at birth. The Aldou values (mean 106.4 ± 10.3 pg/μg of creatinine with median at 111 pg/μg) followed a Gaussian distribution among 28 newborn infants (Fig. 4A). Surprisingly, we did not find any correlation between plasma aldosterone concentrations and Aldou (Fig. 4B), nor with plasma potassium concentration (Fig. 4C), suggesting that the plasma aldosterone value measured in umbilical cord may not constitute a meaningful hormonal parameter. In sharp contrast, Aldou was negatively correlated with kalemia (Fig. 4D). Because aldosterone regulates potassium balance, Aldou, which probably reflects long-term aldosterone secretion, therefore it seems as the best index to accurately assess mineralocorticoid sensitivity during the postnatal period.

Figure 4
figure 4

Aldou is a better index of aldosterone sensitivity at birth. (A) Distribution of Aldou values in newborn infants. Each dot represents one value and lines correspond to median ± interquartiles. (B) Correlation between urinary and plasma aldosterone concentrations in newborn infants. No statistical significance was reached, p = 0.34. (C) Correlation between plasma aldosterone concentration and kalemia in newborn infants. No statistical significance was reached, p = 0.55. (D) Correlation between Aldou and kaliemia in newborn infants, p = 0.005 and correlation coefficient r = −0.52. (B–D) Correlations between two parameters were obtained by Spearman regression analysis.

Full size image

Physiologic aldosterone resistance in newborn infants.

Because it is well established that aldosterone directly regulates renal sodium and potassium transport, we next examined the relationship between Aldou and urinary electrolytes excretion in newborn infants. Interestingly, Aldou was neither correlated with urinary potassium excretion (Fig. 5A) nor with urinary Nau/Ku ratio (Fig. 5B). This was also true for the FeNa (data not shown). These results, along with hyponatremia, hyperkalemia, and high plasma aldosterone levels, unambiguously demonstrate the physiologic partial aldosterone resistance in newborn infants. Incidentally, the newborn with the highest kalemia (13.4 mM) presented with the highest sodium waste (FeNa = 7.6%) and the most diluted urine (creatu 279 μM). Collectively, our findings suggest a tight association between the severity of renal tubular immaturity and the importance of aldosterone resistance in neonates.

Figure 5
figure 5

Aldosterone resistance in newborn infants. (A, B) Correlation between urinary aldosterone and (A) urinary potassium concentration, no statistical significance was reached p = 0.44 or (B) urinary sodium/potassium ratio at birth, no statistical significance was reached p = 0.28. Correlations between two parameters were obtained by Spearman regression analysis.

Full size image

DISCUSSION

In this large prospective study, we demonstrated that healthy newborn infants exhibit partial aldosterone resistance with high plasma levels of aldosterone and renin, contrasting with biologic signs of functional hypoaldosteronism including hyponatremia, hyperkalemia, and urinary sodium loss. These results are in accordance with previous studies that have reported high levels of aldosterone in neonates (10) compared with their mothers (21). Although aldosterone has been previously shown to cross the placental barrier from the mother to the fetus (20), highest aldosterone levels detected in the cord blood are consistent with de novo synthesis in the fetal adrenal gland, given the very early expression of the aldosterone synthase gene starting from 13 gestational weeks (22). These findings indicate an autonomous regulation of corticosteroid hormone biosynthesis in the fetus, even though the high renin levels in cord blood are likely explained by placental biosynthesis (23). Our study also highlights that ionic balances in the maternal and fetal compartments are not strictly interrelated, most notably hyperkalemia of newborn infants contrasts with maternal values, which brings new interest into the analysis of plasma electrolyte concentrations during the first 24 h of life, because it was assumed they strictly reflected maternal values. The demonstration of a relative neonatal mineralocorticoid insensitivity strongly suggests that it results from the inability of the fetal tubular segments to adequately respond to aldosterone action.

One of the main questions remains why and how such a physiologic hormonal unresponsiveness occurs at birth. The perfect integrity of the renin-angiotensin-aldosterone system with adaptive responses of the fetal adrenal gland sharply contrasts with the inability of fetal kidney to appropriately control sodium and water reabsorption. Whether hyperaldosteronism is the direct consequence of renal tubular immaturity, which leads to sodium loss and subsequent activation of the renin-angiotensin-aldosterone system, remains to be elucidated. Alternatively, impaired ability of nephronic segments to adequately respond to aldosterone action may constitute a protective mechanism against high plasma aldosterone levels. The neonatal weight loss related to extracellular fluid loss (24), which occurs during the first week after birth, is likely to be the consequence of this tubular insensitivity. However, the physiologic trigger and the role of this transient sodium and water negative balance in newborn infants are still unclear. Transient aldosterone unresponsiveness observed at birth may be part of an adaptive process from intrauterine aquatic life, where renal sodium reabsorption is dispensable, to extrauterine terrestrial life in which a strict sodium handling controlled by the kidney becomes essential. In this context, the low sodium content of the maternal breast milk (25) could contribute to delay the normalizing hydroelectrolytic homeostasis. On the other hand, a high plasma level of aldosterone could have a functional significance at birth, most notably in some aldosterone target tissues such as endothelial (26) and vascular smooth muscle cells in which high aldosterone levels may be required for rapid vasoactive responses through nongenomic effects, as described recently (27). Finally, a role of prostaglandins could be also evoked owing to the major increase in prostaglandin production in the neonatal period (28) and their inhibitory potential on hydroelectrolytic homeostasis (29).

From a pathophysiological perspective, it has been reported that plasma aldosterone levels and potassium concentration values are higher in low birth weight infants than normal weight infants (30). Similarly, preterm infants have higher aldosterone levels and hyperkalemia than full-term infants (31,32). Given that low birth weight infants have greater risks of developing adult hypertension (33), high levels of plasma aldosterone and kalemia may thus serve as important biochemical parameters and represent risk factors for dysregulated renin-angiotensin-aldosterone system and early onset of a high blood pressure. It is very likely that neonatal sensitivity to aldosterone action varies with birth weight and could be a predictive factor of future chronic diseases with a possible involvement of gene developmental programming as recently described for glucocorticoid sensitivity (34). Finally, high aldosterone values gradually reach normal adult levels during the first year of life in full-term infants (11), somehow following the renal maturation process (1), and suggesting a progressive normalization of mineralocorticoid tubular responsiveness with age.

This physiologic aldosterone unresponsiveness identified in healthy full-term infants is only partial and transient with a rapid normalization of sodium balance at variance with that observed in pseudohypoaldosteronism type I patients, secondary to heterozygous inactivating mutation in the mineralocorticoid receptor gene (35), or in infants with aldosterone synthase deficiency (36). These genetically affected children have to be sodium supplemented during their first years of life because of a high risk of failure to thrive and dehydration, a condition similar to preterm infants.

Given the major physiologic impact of aldosterone action most notably during the critical neonatal period, it is essential to explore hormonal sensitivity at birth by means of precise and reliable biologic criteria. As one of the direct applications of this study, we propose that urinary aldosterone measurement is the best index for accurate evaluation of the mineralocorticoid effector mechanisms. Plasma aldosterone does not constitute a good parameter at birth in accordance with previous studies, which already reported the lack of correlation between kalemia and plasma aldosterone concentrations (37,38). Instead, the highly significant negative relationship found between urinary aldosterone and plasma potassium concentrations establishes that aldosterone participates to the control of potassium balance. However, the slope of this linear regression, which reflects the hormonal sensitivity, may largely depend on the age and birth weight. We anticipate that a steeper correlation slope should be observed as the renal maturation and aldosterone responsiveness increase with age. Thus, determination of urinary aldosterone levels on a spot urinary collection and its subsequent comparison with Nau/Ku ratio or the urinary potassium excretion, characteristic of hormonal resistance, constitutes an interesting noninvasive method to investigate aldosterone sensitivity in preterm and full-term infants.

Further clinical investigations need to be conducted to evaluate aldosterone resistance and the regulation of the renin-angiotensin-aldosterone system in preterm, full-term, low birth weight and normal birth weight infants, as well as in children and adults, to determine its evolution with time and to establish new therapeutic strategies, most notably for preterm neonates in whom sodium loss is a major problem.