Hypothalamic–pituitary–adrenal axis function in pony foals after neonatal ACTH-induced glucocorticoid overexposure



Reasons for performing the study: The effects of overexposure to glucocorticoids during early life of the foal on the subsequent HPA programming of the hypothalamic–pituitary–adrenal axis are unknown.

Objectives: To test the hypotheses that excess glucocorticoid exposure in early life subsequently increases both basal plasma concentrations of cortisol and the adrenocortical responsiveness to exogenous adrenocorticotropic hormone (ACTH).

Methods: Foals received either saline (0.9% NaCl, n = 9) or long-acting ACTH (0.125 mg i.m. b.i.d., n = 6) for 5 days from Day 1 to increase endogenous cortisol concentrations. Long-term indwelling catheters were inserted under local anaesthesia into the jugular veins of foals aged 2 and 12 weeks. After recovery, short-acting ACTH1-24 was given as a single i.v. injection (2 µg/kg bwt) and blood samples were taken at 5–30 min intervals before and after ACTH administration to measure plasma cortisol concentrations.

Results: Basal plasma cortisol concentrations were higher in ACTH- than in saline-treated foals at age 3 weeks, but not at 13 weeks. There were no significant differences in either the time profile or the area under the cortisol curve in response to ACTH between the 2 groups.

Conclusions: These data suggest that ACTH-induced overexposure to glucocorticoids during early post natal life of the foal does not have a programming effect on HPA axis function at 13 weeks. In foals, the effects of ACTH-induced overexposure to glucocorticoids, if any, may not become apparent until much later in life in a long-lived species such as the horse.

Potential relevance: These studies suggest that clinical and other stressful conditions that raise plasma cortisol concentrations during early life are unlikely to programme cardiovascular and metabolic function in horses in the short term.


The hypothalamic–pituitary–adrenal (HPA) axis is essential for the physiological response to stress with effects on salt and fluid homeostasis, the maintenance of arterial blood pressure and on the provision of nutrients to the tissues. Both environmental and physiological stress induce the release of corticotropin releasing hormone (CRH) from the paraventricular nucleus of the hypothalamus into the hypophyseal portal system where it travels to the anterior pituitary and acts on the corticotrophs to release adrenocorticotropic hormone (ACTH) into the systemic circulation. The ACTH stimulates the adrenal cortices to synthesise and release the stress hormone, cortisol, which then acts via a classical negative feedback loop at the levels of both the hypothalamus and pituitary to down-regulate CRH and ACTH secretion and maintain plasma concentrations of cortisol appropriate to the level of physiological stress.

During fetal life, glucocorticoids have an essential role in maturing tissues in preparation for extrauterine life. They switch the cell cycle from proliferation to differentiation and can alter the expression of enzymes, receptors, growth factors, ion channels and transporters (Fowden 1995; Fowden et al. 2005). Consequently, studies in man and animals have linked exposure to glucocorticoids during critical windows of development to the ‘programming’ of adult tissue function (Fowden and Forhead 2004; Fowden et al. 2005). Alterations in HPA axis function following prenatal exposure to glucocorticoids have been reported in rats (Levitt et al. 1996), guinea pigs (Dean et al. 2001) and sheep (Sloboda et al. 2002). Treatment with dexamethasone during late gestation increased corticotrophin releasing hormone messenger ribonucleic acid (mRNA) in the paraventricular nucleus, reduced hippocampal glucocorticoid mRNA and increased basal corticosterone concentrations in rat offspring at 16 weeks (Levitt et al. 1996). Similarly, in guinea pigs, exposure to dexamethasone during late gestation increased basal cortisol concentrations in male but not female prepubertal offspring and significantly attenuated the adrenocortical response to isolation stress in female pups (Dean et al. 2001). These alterations in HPA axis function were associated with increased and decreased glucocorticoid receptor (GR) expression in male and female guinea pig pups, respectively (Dean et al. 2001), suggesting that glucocorticoid-mediated effects on hypothalamic–pituitary function are sex specific. Furthermore, evidence in sheep suggests that the effects of glucocorticoids on post natal HPA axis function can vary with the post natal age of the animal, since a single maternal injection with betamethasone during late gestation increased both basal and stimulated cortisol concentrations in the offspring at age one year, but not at age 2 and 3 years (Sloboda et al. 2002; Sloboda et al. 2007).

The HPA axis matures late in the fetal foal compared with other species (Silver and Fowden 1994; Fowden 1995) and the equine adrenal cortex only becomes responsive to endogenous and exogenous ACTH in the last 15–20 days towards term (Rossdale et al. 1982; Silver and Fowden 1994; Fowden and Silver 1995; Silver and Fowden 1995). Hart et al. (Hart, Slovis & Barton 2009) recently suggested that healthy neonatal foals lack a functionally mature and responsive HPA axis as evidenced by lower adrenocortical responses to endogenous and exogenous ACTH compared with mature horses (Rossdale et al. 1984; Silver et al. 1984; Irvine and Alexander 1987). Additionally, normal ontogenic changes in HPA axis function have been observed in foals during the first week of life (Silver et al. 1984; Ousey et al. 2004; Wong et al. 2009). It therefore seems possible that the HPA axis in foals may remain susceptible to developmental changes in early post natal life. Elevated post natal cortisol concentrations have been reported in foals born premature or dysmature as well as diseased foals with sepsis or hypoxic-ischaemic encephaly (Silver et al. 1984; Hart et al. 2009; Panzani et al. 2009). This is important, since glucocorticoid programming of HPA axis reactivity may ultimately lead to chronic glucocorticoid excess reminiscent of Cushing's syndrome and the ensuing adverse cardiovascular and metabolic disturbances (Seckl and Meaney 2004). However, little is known about the effects of exposure to high concentrations of endogenous glucocorticoids during early life of the foal on the subsequent programming of HPA axis. Therefore we tested the hypotheses that, in foals, exposure to excess glucocorticoids in early post natal life: 1) will increase basal plasma concentrations of cortisol subsequently; 2) will increase the subsequent adrenocortical-responsiveness to exogenous ACTH1-24 at age 3 weeks; and that 3) these alterations will be sex-linked and persist to age 13 weeks. These studies have important implications for the long-term programming of cardiovascular and metabolic function in horses by clinical and other stressful conditions that raise plasma cortisol concentrations during early life.

Materials and methods


Fifteen pony foals that were born spontaneously at full term (325–335 days) were used in the study. All foals were housed in individual boxes with their mothers and suckled ad libitum except during the ACTH challenges when they were separated from the mare for 180 min. All procedures were carried out under the Animals (Scientific Procedures) Act 1986 of the UK Government.

Experimental procedures

After birth, foals were weighed and assigned to one of 2 groups: saline-treated controls received saline (0.9% NaCl, n = 9 of which 5 were female and 4 were male) and ACTH-treated foals received long-acting ACTH (depot synacthena; 0.125 mg i.m. b.i.d.; n = 6, of which 3 were female and 3 were male) for 5 days to maintain an endogenous rise in cortisol concentrations. Jugular vein blood samples (4 ml) were taken by venipuncture immediately before the first injection and daily at 09.00 h for 6 days for measurement of plasma cortisol concentrations.

A long-term catheter (16 gauge central venous catheter)b with an extension tube (Tset)c was inserted into a jugular vein at age 2 weeks. After recovery and at a mean age of 3 weeks, the foals were administered a single, intravenous bolus of short-acting ACTH1-24 (synacthena; 2 µg/kg bwt) and the catheter was immediately flushed with 5 ml saline (0.9% w/v) to assess adrenocortical responsiveness. All ACTH studies were started at approximately 09.00 h to minimise diurnal variations in plasma cortisol concentrations. Blood samples (4 ml) were taken through the jugular catheter at -30, -15 and 0 min (immediately before) and 5, 15, 30, 45, 60, 90 and 120 min after ACTH1-24 infusion for measurement of plasma cortisol concentrations. At the end of the experimental period, the catheter was removed and the foal studied again 10 weeks later using the same protocol. All blood samples were centrifuged immediately at 3000 g and 4°C for 5 min and the plasma was decanted and stored at -20°C. Plasma concentrations of cortisol were determined in unextracted plasma using a commercially available Coat-A-Count cortisol assay kitd that has been validated for equine plasma, and has low cross-reactivity with other naturally occurring steroids (aldosterone 0.03%, corticosterone 0.94%, cortisone 0.98%, deoxycortisol 11.4%, oestriol 0.01%, progesterone 0.02%). Charcoal-stripped horse plasma was spiked with known amounts of cortisol and produced a standard curve that was parallel to that of the human standard solutions provided with the manufacturer's kit. The mean ± s.e.recovery of cortisol added to stripped equine plasma was 100.6 ± 4.3%. The sensitivity of this assay was 0.04 ng/ml, the intra- and interassay coefficients of variation were 5.0 and 2.0%, respectively.

Data analysis

Values are means ± s.e. For each experiment, the area under the cortisol curve was calculated as the integrated plasma concentration after administration of ACTH1-24, from 0–120 min above the baseline concentration at 0 min for all positive values. The slopes of the relationships between birth weights and the plasma cortisol concentrations were assessed using linear regression. Statistical analyses were performed using SigmaStat Statistical Software version 2.0e. Statistical significance was assessed using Student's t test, one-way analysis of variance (ANOVA) and 2-way repeated measures (RM) ANOVA, as appropriate. For all statistical tests, significance was accepted at P<0.05.


All the foals were born spontaneously within the normal range of gestational ages considered full term for pony mares (Rossdale 1976). Mean bodyweights at birth were not significantly different between males or females subsequently treated with saline (female: 26 ± 2 kg; male: 31 ± 5 kg) or ACTH (female: 30 ± 2 kg; male: 29 ± 5 kg) groups. No further clinical tests of foal maturity were carried out since there were no physical indications of immaturity. The times to stand and to suck, as well as basal cortisol concentrations on the first day of life were within the range of normal full term pony foals (Ousey et al. 2004). Before treatment on Day 1, there were no differences in the basal plasma concentrations of cortisol between the saline and ACTH-treated foals (Fig 1). On Day 2, plasma cortisol concentrations were significantly higher in ACTH (207.2 ± 49.3 ng/ml) than saline-treated (20.4 ± 1.8 ng/ml; P<0.05) foals and remained significantly elevated during the 5 day treatment period (Fig 1). On Day 6, cortisol concentrations were significantly lower than baseline values in the ACTH-treated group (13.1 ng/ml) as well as significantly lower than values in saline-treated foals at Day 6 (22.3 ± 2.2 ng/ml; P<0.05).

Figure 1.

Mean ± s.e. basal concentrations of plasma cortisol in saline- (age 3 weeks: n = 9; age 13 weeks: n = 7) and adrenocorticotropic hormone (ACTH)-treated (3 weeks: n = 6; 13 weeks: n = 6) foals (on the first day of life [Day 1], during treatment from Days 2–5 and at age 3 and 13 weeks, respectively). *P<0.05 with respect to post natal age.

Adrenal function at age 3 and 13 weeks

At age 3 weeks, the basal plasma cortisol concentrations were significantly higher in ACTH- than saline-treated foals (Fig 1). However, by age 13 weeks, the basal cortisol concentrations were not different between the 2 groups (Fig 1). Administration of exogenous ACTH1-24 elicited a gradual increase in endogenous plasma cortisol concentrations (Fig 2a,b). The increase in plasma cortisol was significantly higher than baseline 15 min after ACTH1-24 administration and remained elevated for the duration of the sampling period, reaching maximal values between 90 and 120 min in all groups of foals (Fig 2a,b). There were no significant differences in the increments in plasma cortisol (Fig 2a,b) or in the areas under the cortisol curves (Fig 2c) between the saline- and ACTH-treated foals at either age. When the data were analysed for the effects of sex, there were no differences in either the time profile or the area under the cortisol curve to exogenous ACTH1-24 in male and female foals either within or between groups at both 3 and 13 weeks (P = 0.726 and P = 0.807, respectively; data not shown). At the times of the ACTH challenges, foal bodyweights were not different between the saline- (44 ± 3 and 92 ± 5 kg at age 3 and 13 weeks, respectively) and ACTH-treated (45 ± 4 and 96 ± 8 kg at 3 and 13 weeks, respectively) groups. In addition, no significant relationships were found between foal birth weights and the basal plasma concentrations of cortisol or the area under the cortisol curve during the ACTH challenge in any of the groups (data not shown).

Figure 2.

Change in plasma concentrations of cortisol from baseline at 0 min in response to exogenous ACTH1-24 (↓) in saline- (open circles) and ACTH-treated (filled circles) foals at (a) age 3 weeks (saline n = 9; ACTH n = 6) and (b) age 13 weeks (saline n = 7; ACTH n = 6) and (c) Area under the cortisol curve in saline- (□) and ACTH-treated (inline image) foals at the 2 ages. Values are mean ± s.e. ACTH = adrenocorticotropic hormone; AUC = area under the curve.). *P<0.05 compared to -30, -15 and 0 min within a group.


The data from the current study suggest that early post natal exposure to endogenous glucocorticoids induced by ACTH-stimulation of the adrenal gland does not programme alterations in foal HPA axis reactivity. There were no differences in basal plasma concentrations of cortisol at age 13 weeks between the saline- and ACTH-treated groups. In addition, no differences were seen in the adrenocortical response to a supraphysiological dose of ACTH1-24 at age either 3 or 13 weeks, suggesting a lack of difference in stimulated adrenocortical function after ACTH-induced glucocorticoid overexposure. However, there were differences in the baseline cortisol concentrations with treatment at age 3 weeks, which suggests that effects of early glucocorticoid exposure persisted for about 2 weeks after cessation of treatment.

The baseline plasma cortisol concentrations reported for saline-treated controls in the current study are consistent with those reported previously in age-matched foals (Hart et al. 2009; Wong et al. 2009). This suggests that, despite separation from the mare and periodic restraint of the foal, the test procedure did not cause significant endogenous release of cortisol. During the first 5 days of life, basal plasma concentrations of cortisol in ACTH-treated foals were 8–9 times higher than in control foals and were similar to those reported in foals hospitalised for prematurity/dysmaturity, septicaemia and hypoxic-ischaemic encephaly (Hart et al. 2009; Panzani et al. 2009). In ACTH-treated foals at age 3 weeks, a small, but significant increase in plasma cortisol concentrations remained that may be accounted for by: 1) a rebound in plasma cortisol following the withdrawal of the high concentrations of ACTH and cortisol negative feedback on the hypothalamus and pituitary during the treatment period of the foals; or 2) maturation or transient modification of HPA axis function (Fletcher et al. 2004). Such alterations to HPA axis function may include adrenal growth and changes to the steroidogenic pathways in the adrenal cortex (Challis et al. 1985), changes in the responsiveness of the adrenal cortex to ACTH stimulation (Louis et al. 1976; Ratter et al. 1979; Brooks 1992; Brooks et al. 1994) and/or changes to the negative feedback effects of cortisol by altering the expression of GRs in the hippocampus, hypothalamus and pituitary (Matthews et al. 1995; Sloboda et al. 2000; Seckl 2001). Whatever the mechanism, the alteration to baseline adrenocortical function was short lived, as no differences in plasma concentrations of cortisol were found in 13-week-old foals during basal conditions.

Numerous studies in neonatal horses have described the dynamic nature of the HPA axis during the first week of life (Rossdale et al. 1984; Silver et al. 1984; Silver et al. 1987; Silver and Fowden 1994; Ousey et al. 2004; O'Connor et al. 2005; Hart et al. 2009; Wong et al. 2009). In brief, both basal and stimulated HPA axis function appear to be high immediately after birth with a gradual decrease over the first post natal week (Hart et al. 2009; Wong et al. 2009). The lack of significant differences in the increments in plasma cortisol as well as in the areas under the cortisol curves between age 3 and 13 weeks in the present study suggests that no further ontogenic changes in HPA axis reactivity to a supraphysiological dose of ACTH1-24 (2 µg/kg bwt) occur over this time period. These data contrast with the study of Wong et al. (Wong et al. 2009), in which plasma cortisol concentrations following administration of a low dose of cosyntropin (0.1 µg/kg bwt) were significantly higher at 12 weeks than at age 2 and 3 weeks in foals. The low, physiological dose of cosyntropin used in the latter study may have detected a more subtle alteration in adrenocortical responsiveness; alternatively, differences in the breed of horse (Thoroughbred and Quarter Horse foals vs. pony foals) may explain the difference between the latter and our study.

Whilst a wealth of information is available on prenatal programming, relatively little is known about the programming effects of the early post natal environment, when the neonate is adapting to extrauterine life. Studies in rodents have shown that maternal care and maternal-separation stress have opposing effects on the post natal development of the HPA axis (Macrìet al. 2008). In rat pups, short periods of handling or touch stimulation decrease HPA axis reactivity to a variety of stressors through the life of the animal (Meaney et al. 1993, 1996). These rats have lower ACTH and corticosterone concentrations during and after stress, have increased hippocampal GR gene expression and are more sensitive to glucocorticoid negative feedback regulation (Meaney et al. 1989; Viau et al. 1993; O'Donnell et al. 1994; Jutapakdeegul et al. 2003). In contrast, separation of a mother and her pups for 24 h during the post natal stress hyporesponsive period in both rats and mice increases HPA axis reactivity in the mature offspring, resulting in lower hippocampal GR mRNA expression and prolonged corticosterone responses to stress (Meaney et al. 1993; Enthoven et al. 2008). Similarly, daily parental separation stress has also been shown to increase basal cortisol concentrations and reduce hippocampal GR expression in infant-adolescent marmosets (Pryce et al. 2010). Taken together, these studies suggest that the early post natal environment can influence the long-term functioning of the HPA axis.

Exposure to excess glucocorticoids prenatally is known to affect HPA axis function in later life (McMillen and Robinson 2005). Programmed alterations in the expression of GR and mineralocorticoid receptor (MR) in the hippocampus and hypothalamus have been reported in rat (Lesage et al. 2001) and sheep (Bloomfield et al. 2003) offspring after exposure during late gestation to adverse intrauterine conditions, such as maternal undernutrition, that are known to increase fetal exposure to endogenous glucocorticoids (Edwards et al. 2001; McMillen and Robinson 2005). Similarly, fetal exposure to synthetic glucocorticoids during late gestation has been shown to alter hippocampal GR and MR expression as well as basal cortisol concentrations in rats (Levitt et al. 1996) and guinea pigs (Dean et al. 2001). In sheep, however, treatment with betamethasone during mid-late gestation increased responsiveness of the HPA axis to CRH and arginine vasopressin at age 6 and 12 months (Sloboda et al. 2002). Although the results of the current study do not compare directly with prenatal glucocorticoid exposure, it should be noted that in almost all of these studies, the glucocorticoid-mediated effect was observed in offspring that were older than in the current study.

Only limited data are available on the effects of sex on adrenocortical function in foals. One study has shown a significantly greater change in plasma cortisol in male compared with female foals elicited by a physiological dose of exogenous ACTH at age 5–7 days (10 µg; [Hart et al. 2009]). Hart et al. (2009) also reported that the same male foals had a larger change in plasma cortisol than female foals in response to a supraphysiological dose of ACTH (100 µg); however, this did not reach statistical significance. In the present study, there were no differences between male and female foals in either their basal plasma cortisol concentrations during the first week of life and at 3 and 13 weeks or in the areas under the cortisol curves in response to a supraphysiological dose of ACTH1-24 (2 µg/kg bwt) in any of the groups. However, further investigation is required to determine whether foal sex influences adrenocortical responsiveness and whether HPA axis function is differentially affected by glucocorticoid exposure in male and female foals, since intrauterine programming of mature tissue function by exposure to glucocorticoids is not sex-linked until post puberty in laboratory species (Kapoor et al. 2006).

In the present study, depot synacthen was used to stimulate hypercortisolaemia, since it is known to cause prolonged adrenal activation and to increase endogenous plasma cortisol concentrations for at least 24 h (Silver et al. 1984). This ACTH-induced cortisol release provides a good model of HPA axis activation, as would occur naturally during illness. One limitation to the method, however, is that, in addition to glucocorticoid over exposure, foals were exposed to supraphysiological concentrations of ACTH during the treatment period. ACTH acting on the melanocortin 2 receptor stimulates testicular androgen production in fetal-type Leydig cells of the neonatal testis in mice (Johnston et al. 2007). In addition, ACTH may have a role in metabolic function, since ACTH can promote acute insulin resistance in murine adipocytes (Iwen et al. 2008) and produces variable anorexic and orexigenic effects when administered directly into the rat brain, depending on the N-or C- terminal fragment of ACTH studied (Mountjoy 2010). Whether over exposure to ACTH in addition to glucocorticoid has long-lived programming effects on fetal and/or neonatal physiology is currently unknown.

In conclusion, the above data suggest that exposure to excess ACTH-induced glucocorticoids during early post natal life of the foal does not have a programming effect on HPA axis function, since neither basal nor stimulated adrenocortical responsiveness was altered in foals at age 13 weeks in the present study. However, the horse is a long-lived species compared with those more extensively used to study early life programming of mature phenotype. Consequently, the effects of overexposure to endogenous glucocorticoids, if any, may not become apparent until much later in life in the horse.

Conflicts of interest

No conflicts of interest have been declared.

Source of funding

This study was funded by the Horserace Betting Levy Board.

Manufacturers' addresses

a Alliance Pharmaceuticals Ltd, Chippenham, Wiltshire, UK.

b Arrow International Inc., Reading, Pennsylvania, USA.

c Mila International Inc., Florence, Kentucky, USA.

d Siemens Medical Solutions Diagnostics, Los Angeles, California, USA.

e Systat Software Inc., Point Richmond, California, USA.

Author contributions ALF contributed to the conception and design of the study. NBH gave veterinary assistance and catheter placement. JKJ, VLA, AJF and ALF contributed to the acquisition of data, analysis and interpretation. JKJ, VLA, NBH, AJF and ALF contributed to the drafting, critique and final approval of the manuscript for publication.