Low Doses of Dexamethasone Suppress Pituitary-Adrenal Function but Augment the Glycemic Response to Acute Hypoxemia in Fetal Sheep during Late Gestation

Abstract

Despite the widespread use of antenatal glucocorticoid therapy in obstetric practice, little is known about the effects of synthetic glucocorticoids on the fetal capacity to respond to episodes of acute hypoxemia, such as may occur during labor and delivery. This study investigated the effects of prolonged fetal exposure to low concentrations of dexamethasone on the fetal ACTH, cortisol, and glycemic responses to an episode of acute hypoxemia during the period of dexamethasone treatment in sheep. At 118 d of gestation (term is approximately 145 d), 11 fetal sheep had catheters implanted under halothane anesthesia. From 124 d, five fetuses were infused i.v. continuously with dexamethasone (1.80 ± 0.15 μg·kg⁻¹·h⁻¹ in 0.9% saline at 0.5 mL/h) for 48 h, and the other six fetuses received saline solution i.v. at the same rate. At 45 h of infusion, acute hypoxemia was induced in all fetuses for 1 h by reducing the maternal inspired fraction of oxygen. During glucocorticoid treatment, fetal plasma dexamethasone concentrations increased to 3.9 ± 0.2 nM by 24 h and remained elevated for the rest of the infusion period. During hypoxemia, a similar fall in fetal arterial Po₂ occurred in both saline-infused and dexamethasone-treated fetuses. In control fetuses, significant increases in plasma ACTH and cortisol concentrations and in blood glucose concentrations occurred during hypoxemia. Dexamethasone treatment prevented the increases in fetal plasma ACTH and cortisol, and augmented the blood glucose response, induced by hypoxemia. These data indicate that prolonged fetal exposure to low concentrations of dexamethasone suppresses pituitary-adrenal function, but augments the glycemic response, to acute hypoxemia in fetal sheep during late gestation.

Main

Synthetic glucocorticoids, such as betamethasone and dexamethasone, are used routinely in clinical practice to treat pregnant women at risk of preterm labor (¹). Treatment involves maternal intramuscular injection of the synthetic glucocorticoid, using a variety of dosing regimens (²). This prophylactic treatment is designed to elevate fetal plasma glucocorticoid concentrations and to mimic the maturational effects of the normal endogenous, prepartum increase in fetal plasma cortisol concentrations that occurs close to term in humans and other species (^{3, 4}). The prepartum increase in fetal plasma glucocorticoid concentrations induces tissue maturation and prepares the fetus for the transition to neonatal life (⁵). In the human infant, antenatal glucocorticoid treatment has resulted in a significant reduction in the incidence of conditions, such as respiratory distress syndrome, periventricular hemorrhage, and necrotizing enterocolitis, and in neonatal mortality associated with premature delivery (⁶).

Despite the clear benefits of synthetic glucocorticoid therapy in humans, there are possible adverse side effects to current dosing regimens. For example, maternal glucocorticoid therapy in pregnant women modifies fetal heart rate and fetal breathing movement variability (⁷), both of which are used clinically as indices of fetal well-being (⁸). It also elevates mean arterial blood pressure during the first 3 d after delivery compared with placebo-administered control subjects (⁹). In sheep, repeated maternal betamethasone administration suppresses neuroendocrine and adrenal responsiveness in the preterm neonatal lamb (¹⁰). In addition, administration of betamethasone or dexamethasone to fetal sheep produced pronounced hypertension during 48 h of treatment, which was maintained, albeit at a lower level, after cessation of the infusion (¹¹). In that study (¹¹), betamethasone or dexamethasone was infused directly into the fetal jugular vein at doses designed to produce fetal circulating synthetic glucocorticoid concentrations similar to those measured in human infants at cesarean section after maternal antenatal treatment (¹²).

Few studies have investigated the effects of glucocorticoids on the capacity of the fetus to respond to hypoxemia. This is important because acute fetal hypoxemia may result during labor and delivery, which are likely to occur during, or shortly after, the administration of glucocorticoids in clinical practice. In the sheep fetus, acute hypoxemia evokes integrated cardiovascular, metabolic, endocrine, and behavioral responses that facilitate fetal survival during the period of reduced oxygen availability (¹³). The metabolic response results in an increase in fetal plasma concentrations of glucose and lactate (¹⁴). The endocrine response includes activation of the fetal HPA axis and increases in fetal plasma ACTH and cortisol concentrations (^{15, 16}). Previous studies have shown that treatment of fetal sheep either with i.v.-administered cortisol for 5 h (¹⁷) or with dexamethasone implants (700 μg) adjacent to the paraventricular nucleus (¹⁸) abolishes the ACTH response to acute hypoxemia. However, the effect of prolonged systemic treatment of the fetus with low doses of dexamethasone on the cortisol response to acute hypoxemia in sheep remains unknown. Similarly, no study has investigated the effects of fetal treatment with glucocorticoids on the glycemic response to acute hypoxemia. Hence, in the present study, the metabolic, ACTH, and cortisol responses to an episode of acute hypoxemia have been investigated in fetal sheep treated with low concentrations of dexamethasone for 48 h.

Fetal, as opposed to maternal, treatment with dexamethasone was chosen to assess the direct effects of the synthetic glucocorticoid on fetal cardiovascular, endocrine, and metabolic physiology. In addition, this approach avoided confounding influences caused by possible differences in transplacental passage of glucocorticoids between sheep and primate placentas.

METHODS

Animals

A total of 11 Welsh Mountain sheep fetuses of known gestational age were used. All procedures were approved by the ethical review committee of the University of Cambridge and were licensed under the UK Animals (Scientific Procedures) Act, 1986. Food, but not water, was withheld for 24 h before surgery.

Surgical procedures

Surgery was performed between 116 and 118 dGA (term is approximately 145 d). Anesthesia was induced with sodium thiopentone (20 mg/kg i.v. Intraval Sodium; Rhone Mérieux, Dublin, Ireland) and, after intubation, maintained with halothane (1.5% in 50:50 O₂/N₂O) via positive-pressure ventilation. A Teflon catheter was inserted into a maternal femoral artery and its tip advanced into the descending aorta. The gravid uterus was exposed via a midline abdominal incision, and catheters were implanted in the fetus in two stages. After exteriorization of the fetal head, translucent polyvinyl chloride catheters (inner diameter, 0.58 or 0.86 mm; outer diameter, 0.96 or 1.52 mm, respectively; Critchly Electrical Products, New South Wales, Auburn, Australia) were inserted into a fetal carotid artery, a jugular vein, and the amniotic cavity. Then the fetal hind limbs were exteriorized through a second uterine incision. Catheters were inserted into a fetal femoral artery and vein, and their tips were advanced into the descending aorta and caudal vena cava, respectively. Uterine incisions were closed in layers, and the vascular catheters were filled with heparinized saline and sealed with brass pins. All catheters were exteriorized via a maternal flank and housed in a pouch sutured to the skin.

Postoperative care

Ewes were housed in individual pens, had free access to hay and water, and were fed concentrates twice daily (100 g; Sheep Nuts No. 6; H&C Beart Ltd., Kings Lynn, United Kingdom). Antibiotics were administered postoperatively and daily for 3 d to the ewe (9–12 mg Depocillin intramuscularly; Mycofarm, Cambridge, United Kingdom), to the fetus (300 mg i.v. ampicillin; Penbritin, SmithKline Beecham Animal Health, Surrey, United Kingdom), and into the amniotic cavity (300 mg Penbritin). The ewes also received 2 d of postoperative analgesia (3 g daily intraoral phenylbutazone; Equipalozone Paste E-pp, Arnolds Veterinary Products Ltd., Shropshire, United Kingdom). Vascular catheters were maintained patent by a slow continuous infusion of heparinized saline at 0.5 mL/h (80 IU heparin/mL in 0.9% saline).

Experimental procedure

Maternal caudal aorta and fetal carotid blood samples were taken daily for analysis of arterial blood gases and acid-base status, and measurement of blood glucose, lactate, and plasma hormone concentrations. Samples were taken at 1000 h on the 2 days before the infusion began (days 1 and 2), on the 2 days of infusion (days 3 and 4), and on the day after infusion (day 5). At least 6 d after surgery, the fetuses were randomly assigned to one of two experimental groups. From 124 dGA, five fetuses were continuously infused i.v. with dexamethasone (dexamethasone sodium phosphate; Merck, Sharp, Dohme Ltd., Herts, United Kingdom) in heparinized saline (80 IU heparin/mL in 0.9% saline) for 48 h at a rate of 4.23 ± 0.28 μg/h delivered at 0.5 mL/h (1.80 ± 0.15 μg·kg⁻¹·h⁻¹, corrected retrospectively for fetal weight measured at the end of the experimental protocol). The remaining six fetuses were infused i.v. with heparinized saline at the same rate (0.5 mL/h) to act as age-matched controls (Fig. 1). At 126 dGA, 45 h after the onset of infusion, a 1-h period of fetal hypoxemia was induced by reducing the maternal fraction of inspired O₂. The protocol comprised 1 h of normoxia, 1 h of hypoxia, and 1 h of recovery, as described previously (¹⁴); Fig. 1). At the start of the protocol, a large, transparent, polythene bag was placed over the ewe's head, and air was passed through at a rate of approximately 40 L/min. After a 1-h control period, fetal hypoxemia was induced for 1 h by switching the gas mixture breathed by the ewe to 9% O₂ in N₂ (18 L/min air; 22 L/min N₂) with small amounts of CO₂ (1.2 L/min) added to the inspirate. This mixture was designed to reduce fetal Pao₂ to 11–13 mm Hg while minimizing changes in fetal Paco₂. After the 1-h period of hypoxemia, the bag was removed from the ewe's head, and the ewe was allowed to breathe room air for the 1-h recovery period.

Figure 1

Timeline for experimental protocol. Dexamethasone-treated fetuses received a constant i.v. infusion of dexamethasone (1.80 ± 0.15 μg·kg⁻¹·h⁻¹) for 48 h from 124 dGA. Saline-infused fetuses received i.v. heparinized saline (80 IU heparin/mL) at the same rate to act as age-matched controls.

Paired maternal aortic and fetal carotid arterial blood samples (0.3 mL) were collected at 15-min intervals throughout the protocol for the measurement of blood gases, acid-base status, and metabolite concentrations. In addition, a fetal carotid arterial blood gas sample was taken 5 min after the onset of hypoxemia to confirm that fetal Pao₂ had fallen to the expected value. Furthermore, paired maternal and fetal arterial blood samples (5 mL) were collected for hormone measurement at 15 min (early) and 45 min (late) of normoxia, at 15 min (early) and 45 min (late) of hypoxia, and at 45 min of recovery (Fig. 1). Blood samples for hormone analyses were collected into K⁺/EDTA-treated tubes kept on ice and were centrifuged at 4000 rpm for 4 min at 4°C. Plasma samples were stored at −70°C until analyses. All hormone measurements were performed within 2 mo of sample collection.

At the end of the experimental protocol, the ewes and fetuses were killed using a lethal dose of sodium pentobarbitone (200 mg/kg Pentoject; Animalcare Ltd., York, United Kingdom). The positioning of catheters was confirmed, and the fetuses were weighed.

Biochemical analyses

Maternal aortic and fetal carotid arterial blood gas status, acid-base status, and hemoglobin concentrations were determined using an ABL5 blood gas analyser and OSM2 hemoximeter (Radiometer, Copenhagen, Denmark). Measurements in maternal and fetal blood were corrected to 38°C and 39.5°C, respectively. Blood glucose and lactate concentrations were measured using an automated analyzer (Yellow Springs 2300 Stat Plus Glucose/Lactate Analyser; YSI Ltd., Farnborough, United Kingdom). Plasma ACTH, cortisol, and dexamethasone concentrations were determined by RIA, validated for use in sheep plasma.

Dexamethasone.

Plasma dexamethasone concentrations were measured after ether extraction using tritium-labeled dexamethasone as tracer. Duplicate 100-μL plasma samples were extracted with 2.5 mL of diethyl ether. After freezing, the ether was decanted and evaporated, and the residue was reconstituted in 500 μL of PBS. Aliquots of varying volumes were removed (depending on expected concentrations determined in pilot studies) made up to 450 μL of PBS and incubated with 100 μL of PBS containing 16 000 dpm [1,2,4,6,7-³H]dexamethasone (Amersham Life Science, Bucks, United Kingdom) and 100 μL of sheep anti-dexamethasone antiserum (Bioclinical Services International, Cardiff, United Kingdom). Bound and free dexamethasone were separated using dextran-coated charcoal. After centrifugation, a 500-μL aliquot was removed for measuring radioactive content. All values were corrected for recovery (86%). The interassay c.v. for three control plasma pools (1.8, 5.4, and 26.7 nM) were 14.6%, 9.3%, and 8.2%, respectively. The lower detection limit of the assay was 0.2 nM. Measurements that were lower than the limit of detection were assigned a value of 0.2 nM. The anti-dexamethasone antiserum showed a 1.6% cross-reactivity against cortisol and cross-reactivities of less than 0.5% against 11-deoxycortisol, corticosterone, testosterone, progesterone, and oestriol.

ACTH.

Maternal and fetal plasma ACTH concentrations were measured using a commercially available double antibody ¹²⁵I RIA kit (Incstar Ltd., Wokingham, United Kingdom). The lower limit of detection of the assay was 15 pg/mL. The interassay c.v. was 8.4%. The intra-assay c.v. for two plasma pools (37 and 150 pg/mL) were 3.6% and 4.1%, respectively. The assay had negligible (<0.01%) cross-reactivities with α-melanocyte-stimulating hormone, β-endorphin, β-lipotropin, leucine enkephalin, methionine enkephalin, bombesin, calcitonin, PTH, FSH, AVP, oxytocin, and substance P.

Cortisol.

Maternal and fetal plasma cortisol concentrations were measured as described previously (¹⁹). The lower limit of detection of the assay was 1.0 ng/mL. The intra- and interassay c.v. were 5.3% and 13.0%, respectively. The antiserum showed no detectable cross-reactivity with dexamethasone. The cross-reactivities of the antiserum at 50% binding with other cortisol-related compounds were 0.5% cortisone, 2.3% corticosterone, 0.3% progesterone, and 4.6% deoxycortisol.

Statistical analyses

Values are expressed as mean ± SEM unless otherwise stated. Statistical significance of any changes in any measured variable within and between treatment groups was assessed using a two-way ANOVA with one repeated measure and the post hoc Tukey test. For all statistical comparisons, significance was accepted when p < 0.05.

RESULTS

Outcome of experimental preparations

Normal feeding patterns were restored within 24–48 h after surgery. None of the ewes went into labor during the study period. All fetuses were normal at delivery. No significant difference in fetal body weight was found between saline-infused (2.44 ± 0.43 kg; mean ± SD;n = 6) and dexamethasone-treated (2.44 ± 0.31 kg;n = 5) fetuses at delivery.

Plasma dexamethasone concentrations

Plasma dexamethasone concentrations remained undetectable in the mothers and saline-infused fetuses throughout the experimental period (Fig. 2). In the dexamethasone-treated fetuses, plasma dexamethasone concentrations increased to 3.9 ± 0.2 nM at 24 h of infusion (p < 0.05, n = 4) and remained elevated for the rest of the infusion period (Fig. 2). These concentrations are approximately one fifth of those measured in human umbilical arterial blood samples taken from infants at cesarean section after maternal antenatal glucocorticoid treatment (¹²). Dexamethasone was undetectable again in the plasma of dexamethasone-infused fetuses 24 h after the infusion ceased.

Figure 2

Maternal and fetal plasma dexamethasone concentrations during the experimental protocol. Values shown are mean ± SEM for saline-infused (white bars;n = 4) and dexamethasone-treated (black bars;n = 4) fetuses. a, p < 0.05, differences by post hoc analysis relevant to a significant main effect of time. b, p < 0.05, differences by post hoc analysis relevant to a significant main effect of treatment. Bar represents period of infusion. Dashed line represents the lower detection limit of the assay (0.2 nM).

Effects of dexamethasone on daily basal blood gas, metabolite, and hormone concentrations

Arterial blood gas, acid-base, and metabolite status.

In the control animals, all maternal and fetal blood gas, acid-base, and metabolite variables were within the normal range and did not change during the experimental period. Dexamethasone treatment of the fetuses had no significant effect on daily values for maternal arterial blood gases, acid-base, and metabolite concentrations. In the fetus, hemoglobin concentrations were raised significantly during dexamethasone infusion (from 9.2 ± 0.2 to 10.4 ± 0.3 g/dL;p < 0.05). Similarly, the fetal blood glucose concentration was increased at 48 h of dexamethasone treatment to a value that was significantly higher than that measured in the saline-infused fetuses (1.08 ± 0.04 versus 0.82 ± 0.05 mM;p < 0.05). None of the other fetal blood gases or metabolite values were altered by dexamethasone treatment.

Hormone concentrations.

During both saline and dexamethasone infusion, plasma ACTH and cortisol concentrations remained unchanged from preinfusion values in both groups of fetuses (14.0 ± 11.2 versus 11.6 ± 4.6 ng/mL, saline versus dexamethasone) and mothers (11.3 ± 7.6 versus 13.1 ± 7.4 ng/mL).

Effects of dexamethasone on responses to acute hypoxemia

Arterial blood gas, acid-base, and metabolite concentrations.

Maternal. Maternal arterial blood gas, acid-base, and metabolite concentrations during the hypoxia protocol are shown in Table 1. During normoxia, values for maternal pHa, Paco₂, Pao₂, acid-base excess, hemoglobin saturation and concentration, hematocrit, and blood glucose and lactate concentrations were not significantly different between saline-infused and dexamethasone-treated animals. During hypoxemia, the decreases in maternal Pao₂ and hemoglobin saturation were similar in both groups. In mothers of saline-infused fetuses, a mild reduction in maternal Paco₂ occurred during hypoxemia when compared with normoxic baseline values (p < 0.05, Table 1). However, no differences in maternal Paco₂ were measured between dexamethasone-treated and saline-infused animals during hypoxemia. The small but significant increase in maternal pHa detected at 45 min of hypoxemia was similar in magnitude in both groups of animals (p < 0.05, Table 1). The changes in maternal pHa, Paco₂, and Pao₂ observed during hypoxemia reverted to normoxic levels during recovery in mothers of both saline- and dexamethasone-treated fetuses.

Table 1 Maternal arterial blood gas, acid-base, and metabolite status Values shown are mean ± SEM for mothers of saline-infused (n = 6) or dexamethasone-treated (n = 5) fetuses at 15 (early) and 45 (late) min of normoxia, 15 (early) and 45 (late) min of hypoxemia, and 15 (early) and 45 (late) min of recovery. Maternal blood gas values were corrected to 38°C. Abbreviations: ABE, acid-base excess; So₂, oxygen saturation of Hb; Hct, hematocrit. *p < 0.05, differences by post hoc analysis relevant to a significant main effect of time.

Fetal. Values for fetal carotid arterial blood gas and acid-base status during the hypoxemia protocol are shown in Table 2. These variables were similar in the dexamethasone-treated and saline-infused fetuses during the normoxic period. Fetal carotid Pao₂ fell rapidly to similar values in the dexamethasone-treated (from 20.8 ± 1.1 to 13.0 ± 0.6 mm Hg, p < 0.05) and saline-infused (from 21.3 ± 0.8 to 12.2 ± 1.3 mm Hg, p < 0.05) fetuses 5 min after beginning the hypoxemic challenge. During the hypoxemic challenge, fetal Pao₂ and hemoglobin saturation were reduced to similar levels in both dexamethasone-treated and saline-infused fetuses. Although these were achieved without significant changes in fetal Paco₂ in dexamethasone-treated fetuses, a mild hypocapnia was measured in saline-infused fetuses during acute hypoxemia (p < 0.05, Table 2). There were similar reductions in pHa and acid-base excess in both groups of fetuses during late hypoxemia and the recovery period (Table 2). In both groups of fetuses, blood lactate concentrations were similarly elevated during late hypoxemia and the recovery period (Fig. 3, A ). In saline-infused fetuses, blood glucose concentrations and the fetal-maternal glucose concentration ratio were elevated during late hypoxemia compared with normoxic baseline values (Fig. 3, B – D ). Despite an elevated baseline blood glucose concentration in dexamethasone-treated fetuses, the increment in blood glucose in these fetuses measured during early hypoxemia was enhanced compared with saline-infused fetuses (p < 0.05, Fig. 3, C ).

Table 2 Fetal carotid arterial blood gas, acid-base, and metabolite status Values shown are mean ± SEM for saline-infused (n = 6) or dexamethasone-treated (n = 5) fetuses at 15 (early) and 45 (late) min of normoxia, 15 (early) and 45 (late) min of hypoxemia, and 15 (early) and 45 (late) min of recovery. Fetal blood gas values were corrected to 39.5°C. Abbreviations are defined in Table 1 footnote. *p < 0.05, differences by post hoc analysis relevant to a significant main effect of time. ^†p < 0.05, differences by post hoc analysis relevant to a significant main effect of treatment.

Figure 3

Fetal blood lactate concentrations (A), blood glucose concentrations (B), blood glucose increment (C), and fetal-maternal blood glucose ratio (D) during the acute hypoxemia protocol. Values shown are mean ± SEM for saline-infused (white bars;n = 6) and dexamethasone-treated (black bars;n = 5) fetuses. Glucose increment is calculated with respect to mean normoxic baseline values. a, p < 0.05, differences by post hoc analysis relevant to a significant main effect of time. b, p < 0.05, differences by post hoc analysis relevant to a significant main effect of treatment. Box represents period of hypoxemia.

Endocrine responses.

Maternal. Maternal plasma ACTH and cortisol concentrations were similar in both dexamethasone-treated (38.9 ± 7.6 pg/mL and 18.0 ± 3.1 ng/mL, respectively) and saline-infused animals (46.5 ± 10.7 pg/mL and 16.3 ± 3.6 ng/mL, respectively) and did not change significantly from baseline throughout the hypoxemia protocol in either group of animals.

Fetal. In saline-infused fetuses, significant increases in plasma ACTH and cortisol concentrations occurred during hypoxemia (p < 0.05, Fig. 4). In these fetuses, plasma ACTH and cortisol concentrations returned toward baseline during recovery, but remained elevated 45 min after cessation of the hypoxemic challenge. In contrast, no rise in plasma ACTH or cortisol concentrations occurred during acute hypoxemia in the dexamethasone-treated fetuses (Fig. 4).

Figure 4

Fetal plasma ACTH (A) and cortisol (B) concentrations during the acute hypoxemia protocol. Values shown are mean ± SEM for saline-infused (white bars;n = 6) and dexamethasone-treated (black bars;n = 5) fetuses. a, p < 0.05, differences by post hoc analysis relevant to a significant main effect of time. b, p < 0.05, differences by post hoc analysis relevant to a significant main effect of treatment. Box represents period of hypoxemia.

DISCUSSION

The data reported in this study show that in pregnant sheep during late gestation, fetal treatment with dexamethasone, resulting in lower concentrations than those measured in human obstetric practice, abolishes the fetal plasma cortisol response, in addition to preventing the increase in plasma ACTH in response to acute hypoxemia occurring during dexamethasone exposure. Furthermore, low concentrations of dexamethasone increase basal fetal blood glucose concentrations and enhance the fetal glycemic response to acute hypoxemia.

In sufficient amounts, synthetic glucocorticoids induce parturition in the sheep (^{20, 21}). The dose of dexamethasone given to the sheep fetus is therefore critical in determining whether labor is initiated in this species. In the present study, the dose of dexamethasone given to the fetus was the highest compatible with maintaining pregnancy in Welsh Mountain ewes, and was only 60% of the weight-specific dose used previously to treat fetuses of Rambouillet-Columbia ewes without inducing labor (²¹). Fetal administration of dexamethasone to Welsh Mountain sheep at the higher dose induced parturition by 48 h of treatment (Fletcher AJW, Giussani DA, unpublished observations). Fetal plasma concentrations of dexamethasone measured during infusion in the present study (3.2 ± 0.8 nM; mean ± SEM during the 48-h treatment regimen for all treated fetuses) were therefore lower than those observed previously (20 nM; see Derks et al. (²¹) and were 20% of the mean value found in umbilical arterial blood samples taken from human infants at cesarean section 12 h after the completion of a course of maternal antenatal glucocorticoid treatment (5 mg dexamethasone intramuscularly every 12 h for 48 h; see Kream et al. (¹²). However, the National Institutes of Health advise routine administration ranging from two to four doses over 48 h (²), and some of the more common recommended dosing regimens currently used (e.g. two doses at 24-h intervals) may produce lower circulating concentrations of dexamethasone at delivery than those reported by Kream et al. (¹²). Furthermore, clinical dosing regimens will expose the human fetus to initially high, but then rapidly decreasing, concentrations of synthetic steroid, whereas in the present study the sheep fetuses were exposed to a continuous infusion of dexamethasone during the 48-h treatment period. The half-life of dexamethasone in fetal sheep is approximately 6–8 h (²²).

In the present study, the increments in plasma ACTH and cortisol during acute hypoxemia in control fetuses were similar in magnitude to those reported previously for similar degrees of hypoxemia imposed in fetuses at the same stage of gestation (^{15, 17, 23}). These fetal plasma ACTH and cortisol responses to acute hypoxemia were completely abolished by the comparatively low concentrations of dexamethasone achieved in the present study. The observations are consistent with previous findings on the effects of dexamethasone on pituitary-adrenal responses in both fetal and adult animals (^{18, 24}). In the sheep fetus, dexamethasone implants close to the paraventricular nucleus inhibit fetal ACTH secretion in response to acute hypoxemia (¹⁸). Similarly, short-term cortisol administration for 5 h to the sheep fetus abolishes the ACTH response to this stimulus (¹⁷). The present study extends these findings to demonstrate that prolonged systemic administration of dexamethasone to the fetus inhibits not only ACTH but also cortisol secretion in response to acute hypoxemia. It is unknown whether this suppression of the fetal HPA axis is transient or persists after the end of the treatment period. However, clinical studies in humans indicate that fetal HPA axis suppression is transient (²⁵). These observations have important implications for the dosing regimens used for antenatal treatment with synthetic glucocorticoids in clinical practice.

In addition to suppressing HPA axis activity via glucocorticoid receptor negative feedback mechanisms, dexamethasone, and other glucocorticoids such as cortisol, may exert maturational effects on components of the axis, particularly if administered for prolonged periods. For example, ovine fetal adrenocortical cells cultured in the presence of dexamethasone for 48 h exhibit enhanced cAMP production in response to exogenous ACTH administration (²⁶). Furthermore, glucocorticoids may potentiate adrenal responsiveness to ACTH, both directly (²⁷) and indirectly, for example by stimulating production of prostaglandins (²⁸), which have been reported to activate the fetal HPA axis (^{29, 30}). Glucocorticoid treatment may also alter the gain of neural influences on the adrenal cortex, which have been demonstrated to be important regulators of adrenocortical sensitivity to ACTH in the sheep both before (¹⁵) and after birth (³¹). However, the current observations indicate that the suppressive effects of glucocorticoid negative feedback rather than the maturational effects of dexamethasone are the predominant influence on the fetal HPA axis during acute hypoxemia.

In the present study, despite suppression of the ACTH and cortisol responses to acute hypoxemia, dexamethasone treatment had little effect on basal fetal plasma ACTH and cortisol concentrations, which in both groups of animals were within the normal range reported for sheep fetuses of similar gestational ages (^{5, 15}). In addition, maternal antenatal glucocorticoid treatment has been reported to reduce fetal body weight in rats (³²) and in sheep (³³). In the present study, fetal treatment with dexamethasone in low concentrations did not affect fetal body weight. These findings are in keeping with those of Jobe et al. (³⁴), which showed fetal growth impediment after administration of glucocorticoids to the mother, but not to the fetus, at comparable ages of gestation in sheep.

In contrast to the fetus, the degree of hypoxemia achieved in the ewes in this study failed to elicit significant maternal pituitary-adrenal responses. This may reflect differences between maternal and fetal HPA axis threshold or sensitivity to acute hypoxemia. Indeed, differences between fetal and maternal sensitivity to hypoxemia have been demonstrated for plasma vasopressin responses to acute inhalational hypoxia in pregnant ewes (³⁵). Additionally, it is not known whether adult pituitary-adrenal sensitivity to acute hypoxemia is altered by pregnancy. However, increases in adult basal plasma ACTH and cortisol concentrations have been observed in ovine pregnancy compared with the nonpregnant state, and this may indicate modification of adult HPA axis function between different reproductive states in the ewe (³⁶).

In common with previous findings (³⁷), baseline blood glucose concentrations were elevated in the sheep fetus by dexamethasone treatment. The current results extend these findings to show that in the dexamethasone-treated fetuses, the blood glucose concentrations reach higher levels in early hypoxemia compared with control fetuses. The increases in basal blood glucose and in the hyperglycemic response to hypoxemia may have been related to either suppression of fetal glucose utilization or activation of endogenous glucose production in response to reducing oxygen availability. Glucocorticoids have been shown to increase hepatic glycogen content (³⁸), hepatic and renal gluconeogenic enzyme activities (³⁹), and hepatic adrenoceptor density (⁴⁰). Dexamethasone is therefore likely to have increased the fetal capacity for hepatic glucose production. Certainly, administration of cortisol to sheep fetuses close to term has been shown to stimulate endogenous glucose production (⁴¹). Glucocorticoids therefore appear to have an important role in increasing the availability of glucose to the fetal tissues during periods of adverse intrauterine conditions such as during hypoxemia.

In the present study, fetal treatment with dexamethasone enhanced fetal blood hemoglobin concentration. These findings are consistent with those of Anwar et al. (⁴²), who reported that fetal treatment with betamethasone also led to an increase in fetal erythrocyte count and basal hemoglobin concentration. The mechanism of the synthetic glucocorticoid-induced increase in fetal hemoglobin concentration is unclear. It is possible that dexamethasone and betamethasone may stimulate fetal erythropoiesis by affecting erythroprogenitor cells or fetal erythropoietin production (⁴³). However, the available literature suggests that both endogenous and exogenous glucocorticoids suppress fetal erythropoietin production and gene expression in a tissue-specific manner (^{44, 45}). In addition, glucocorticoids may promote an increase in the rate of switching from fetal to adult hemoglobin synthesis (⁴⁶) and changes in mean erythrocyte corpuscular volume (⁴⁷), both of which may alter basal hemoglobin concentration.

In conclusion, this article reports for the first time that fetal treatment with dexamethasone, at relatively low doses, abolishes the cortisol response, but enhances the glycemic response, to an episode of acute hypoxemia in fetal sheep during late gestation. The consequences of these effects on the capacity of the fetus to respond to acute hypoxemia remain uncertain. However, inasmuch as the dexamethasone concentrations achieved in the present study were only one fifth of those seen in obstetric practice, these findings have important implications for the clinical use of antenatal glucocorticoid therapy in pregnant women.