. 22
( 51 .)



Area (ng/min/ml)
80 6000
Cortisol (ng/ml)

Saline M1 M4




0 20 40 60 80 100 120 140 160 180 200 220 240
Time (minutes)
Figure 4.2 One single dose of betamethasone administered to ewes at 104 days of gestation
resulted in signi¬cant increases in cortisol responsiveness to a CRH with AVP challenge
in their lambs at 1 year of postnatal age. M1 (shaded squares) represents animals that
received one single dose of betamethasone at 104 days of gestation followed by 3 weekly
injections of saline. M4 (black diamonds) represents animals that received 4 weekly doses
of maternal betamethasone beginning at 104 days of gestation. MS (open triangles)
represents animals that received four doses of saline at weekly intervals starting at 104
days of gestation. Histograms represent the area under the cortisol response curves.
All values are expressed at mean SEM. Adapted from Sloboda et al. (2002a)

been shown by numerous studies to alter growth and HPA activity as well as glu-
cose tolerance (Uno et al., 1990; Weinstock et al., 1992; Lindsay et al., 1996; Welberg
and Seckl, 2001; Figure 4.2). Low birth weight in humans correlates with increased
adult cortisol levels as well as insulin resistance and elevated blood pressure
(Phillips et al., 1998; Levitt et al., 2000; Reynolds et al., 2001).

Prenatal glucocorticoid exposure and fetal programming

The placenta and prenatal stress
The placental enzyme 11 -HSD2 acts as a dehydrogenase enzyme, rapidly converting
active glucocorticoids to inactive metabolites (Edwards et al., 1993; Krozowski, 1999)
and represents a barrier reducing fetal exposure to elevated levels of maternally-
derived glucocorticoids (Brown et al., 1993). As a result placental 11 -HSD2
120 D. M. Sloboda et al.

plays a direct role in fetal programming through its regulation of fetal exposure to
endogenous glucocorticoids. In rats, reduced placental 11 -HSD2 activity is asso-
ciated with increased blood pressure in adult offspring (Edwards et al., 1993), sub-
stantiating the role that glucocorticoids play in programming adult disease and
supporting the importance of the placenta in regulating fetal adaptations. Kajantie
et al. (2003) demonstrated that relative birth weight in small preterm infants is
correlated with placental 11 -HSD2 activity and infants with increased umbilical
artery resistance had lower total placental 11 -HSD2 activity. Treatment of preg-
nant rats with carbenoxolone, a potent inhibitor of 11 -HSD2, results in signi¬-
cant reductions in birth weight, and signi¬cantly higher fasting basal glucose levels,
elevated insulin responses to a glucose challenge and elevated basal corticosterone
levels (Lindsay et al., 1996; Saegusa et al., 1999; Welberg et al., 2000). These results
were abolished by maternal adrenalectomy (Lindsay et al., 1996); therefore these
effects are mediated via fetal exposure to maternally-derived glucocorticoids. Even
more importantly, evidence exists to suggest that synthetic glucocorticoid admin-
istration decreases ovine placental 11 -HSD2 expression and results in a reduction
in fetal weight (Kerzner et al., 2002). These observations suggest that following
synthetic glucocorticoid administration the fetus is not only exposed to exogenous
glucocorticoids, but also may be exposed to increased levels of maternally-derived
endogenous glucocorticoids.
Prenatal maternal stress increases maternal endogenous glucocorticoid levels
potentially resulting in fetal exposure to elevated levels of glucocorticoid during
development. Prenatal stress has been shown to permanently program the pattern
of HPA and metabolic responses, even though these relationships are complex, and
subtle differences in stimuli exert different effects (Seckl, 1997). Most human data
come from retrospective studies on children whose mothers experienced psycho-
logical stress during pregnancy (Weinstock, 1996; Austin and Leader, 2000;
Niederhofer and Reiter, 2000). Some of these children have delayed motor develop-
ment and abnormal behavioral characteristics (Weinstock, 1996). Experimental
evidence suggests that stress increases both maternal and fetal glucocorticoid levels
and that maternally-derived glucocorticoids may program postnatal HPA activity
(Barbazanges et al., 1996; Takahashi, 1998). Stress during pregnancy in the rat has
resulted in offspring with elevated basal plasma ACTH and corticosterone levels
(Takahashi and Kalin, 1991), increased corticosterone and ACTH responses to a
stressor (Takahashi and Kalin, 1991; Weinstock et al., 1992; Barbazanges et al., 1996)
and altered anxiety behavior (Weinstock et al., 1992). It has been shown that post-
natal responses of prenatally stressed offspring can be suppressed by maternal
adrenalectomy, further supporting the observation that maternally-derived
glucocorticoids program postnatal alterations in HPA function (Barbazanges
et al., 1996).
121 Fetal HPA activation, preterm birth and postnatal programming

A substantial body of evidence exists describing HPA function after postnatal
manipulations in the neonatal rat (Meaney et al., 1985; 1989; Liu et al., 1997;
Avishai-Eliner et al., 2001). The rodent gives birth to immature offspring and the
period of rapid brain growth associated with HPA development occurs in the ¬rst
2 weeks of postnatal life (Rosenfeld et al., 1992). Therefore, the rat HPA axis is sus-
ceptible to programming in the early postnatal period. Early postnatal events such
as maternal separation or neonatal handling, result in signi¬cant elevations in hip-
pocampal GR-binding capacity and number (Meaney et al., 1985; 1989), reduced
plasma ACTH and corticosterone responses to stress and enhanced glucocorticoid
feedback sensitivity (Liu et al., 1997). These occur at a time in which the HPA axis
is relatively quiescent in the developing neonate (Rosenfeld et al., 1992). Therefore,
neonatal handling during a critical developmental window (1“3 weeks postnatally)
in the rat results in permanent alterations in HPA function as a result of alterations
in hippocampal corticosteroid receptors. Several studies have shown that the pre-
natal effects of stress are reversible by early postnatal manipulations. Prenatally
stressed rats exposed to postnatal handling exhibited signi¬cantly lower corticos-
terone responses to stress as adults (Vallee et al., 1996). Postnatal adoption that
encourages maternal interaction with pups also reverses the effects of prenatal stress,
decreasing stress-induced corticosterone peak levels in adult offspring (Maccari
et al., 1995). These observations highlight the importance of different develop-
mental windows, during which exposure to elevated glucocorticoid levels may
produce permanent effects. It has been suggested that in the rat prenatal stress
may have to occur several days beyond birth in order to cause permanent effects
(Takahashi, 1998). This concept is somewhat different in mammals that exhibit
HPA axis development and brain growth in the prenatal or perinatal period such
as in the primate or sheep.

Antenatal administration of glucocorticoids
Over 30 years ago, Liggins (1969) demonstrated that lambs delivered prematurely
(118“123 days of gestation) after fetal infusions of ACTH, cortisol or dexametha-
sone exhibited advanced alveolar stability in their lungs and suggested that the
maturational properties of glucocorticoids caused premature pulmonary develop-
ment and maturation. Subsequently, Liggins and Howie (1972) were the ¬rst to
demonstrate that the administration of maternal glucocorticoids to women at risk
of preterm delivery signi¬cantly enhanced fetal lung maturation and reduced
neonatal morbidity and mortality. In this study, women in premature labor at
24“34 weeks of gestation were admitted into the ¬rst controlled trial of antepar-
tum glucocorticoid treatment for the prevention of respiratory distress syndrome
(RDS) in premature infants (Liggins and Howie, 1972). The administration proto-
col consisted of an intramuscular injection of a mixture of 6 mg of betamethasone
122 D. M. Sloboda et al.

phosphate and 6 mg of betamethasone acetate or a control injection, followed by a
second injection 24 hours later. The incidence of RDS in preterm infants was
reduced by 50% and neonatal death in the ¬rst 7 days of life was signi¬cantly less
frequent, although the maximum effects were seen if delivery occurred more than
24 hours and less than 7 days after treatment (Liggins and Howie, 1972). Synthetic
glucocorticoids such as betamethasone and dexamethasone are 25“30 times more
potent glucocorticoids than cortisol with insigni¬cant mineralocorticoid action
(Speight, 1987). Furthermore, synthetic glucocorticoids do not bind to circulating
binding proteins (corticosteroid binding protein/corticosteroid binding globulin,
CBG) (Pugeat et al., 1981) and are poor substrates for metabolism by placental
11 -HSD2 (Siebe et al., 1993) making synthetic glucocorticoids prime candidates
for clinical management of women at risk of preterm delivery.
Since the ¬rst report by Liggins and Howie (1972), multiple trials have demon-
strated a decrease in the number of cases of RDS and mortality among treated
infants (Kari et al., 1994; Ballard and Ballard, 1996; Anyaegbunam et al., 1997; Ee
et al., 1998). In 1995, The National Institutes of Health (NIH) Consensus Develop-
mental Conference on the Effects of Corticosteroid for Fetal Maturation con-
cluded that antenatal corticosteroid therapy for fetal lung maturation reduced
mortality, RDS and intraventricular hemorrhage in preterm infants (NIH Consensus,
1995). According to the panel, corticosteroids should be administered to women at
risk of preterm birth between 24 and 34 weeks of gestation and in a treatment win-
dow of 24 hours to 7 days prior to delivery. Since 1972, the administration of syn-
thetic glucocorticoids to women threatened with preterm delivery has become
routine practice. Until recently many medical practitioners assumed that more
may be better. By the late 1990s surveys demonstrated that a high percentage of
obstetricians prescribed repeat doses in cases of pregnant women who had a per-
sisting risk of preterm delivery (Quinlivan et al., 1998; Brocklehurst et al., 1999).
However, the mechanisms regulating the onset of preterm labor are poorly under-
stood and as a result preterm labor is dif¬cult to diagnose accurately. Given the
increasing evidence suggesting that excessive fetal glucocorticoid exposure has
long term consequences, it is worrying that women who are not in preterm labor
may be receiving unnecessary corticosteroid administration.
There is substantial evidence from animal studies demonstrating that fetal
exposure to elevated levels of glucocorticoids alters fetal growth and has long-term
effects on cardiovascular, HPA and metabolic function. Early studies with rhesus
monkeys demonstrated that maternal intramuscular betamethasone administra-
tion at 120“133 days of gestation (term 167 days) resulted in signi¬cant reductions
in fetal body weight of 23% at 133 and 167 days of gestation. In addition, brain,
cerebellar, pancreatic, adrenal and pituitary weights were all signi¬cantly reduced
with treatment (Johnson et al., 1981). Signi¬cant growth restriction has also been
123 Fetal HPA activation, preterm birth and postnatal programming

shown in most animal models studied (Bakker et al., 1995; Levitt et al., 1996;
Nyirenda et al., 1998; Newnham et al., 1999; Thakur et al., 2000; Sloboda et al., 2000).
Some time ago, our research group developed a model to investigate the effects
of maternal synthetic glucocorticoid administration on the developing fetal lung
(Jobe et al., 1993; Ikegami et al., 1997). In this model intramuscular injections of
betamethasone (0.5 mg/kg) are administered to the pregnant sheep, beginning at
104 days of gestation (term 150 days), with repeated injections given again at
111, 118 and 125 days. This dose has been shown to be the minimal dose required
for maximal fetal lung maturation in this model. In our model of maternal admin-
istration of synthetic glucocorticoids, fetal weight is signi¬cantly reduced in a dose
dependant manner (Ikegami et al., 1997, Newnham et al., 1999; Figure 4.3), per-
sisting until 3 months of postnatal age (Moss et al., 2001). These alterations in
body weight have been associated with signi¬cant reductions in whole brain and
cerebellum weights, as well as reductions in the myelination of axons located in the
optic nerve and the corpus callosum (Dunlop et al., 1997; Huang et al., 1999).
Prenatal glucocorticoid exposure has long-term effects on brain growth, reducing
brain weight in adult animals aged 3.5 years (Moss et al., 2005). Such observations
have important implications for the ˜hard-wiring™ of the brain and suggest that
long-term brain function may be quite vulnerable to glucocorticoid administra-
tion. French et al. (1999) demonstrated a dose dependant reduction in neonatal
head circumference with increasing doses of maternal corticosteroids in a geo-
graphical based cohort of preterm infants. Further, re-evaluation of these infants at
3 and 6 years of age demonstrated that children who had received 3 or more






Saline M1 M4
Figure 4.3 Betamethasone administration in ewes signi¬cantly reduces lamb birth weight in a
dose-dependent manner. M1 represents animals that received one single dose of
betamethasone at 104 days of gestation followed by three weekly injections of saline.
M4 represents animals that received 4 weekly doses of maternal betamethasone
beginning at 104 days of gestation. Adapted from Newnham et al. (1999)
124 D. M. Sloboda et al.

courses of antenatal corticosteroids had signi¬cantly higher relative risks of
demonstrating externalizing behavioral disorders (French et al., 2003). This effect
was speci¬c to behavior with no differences in intelligence testing.

Adult disease and programming HPA axis function: human and
animal evidence

Growth restricted babies have elevated levels of cord plasma CRH, ACTH and cor-
tisol (Economides et al., 1991; Goland et al., 1993). In addition, increases in uri-
nary glucocorticoid metabolites in children of 9 years of age were associated with
reduced birth weight (Clark et al., 1996). Recent epidemiological studies have
begun to establish a strong correlation between circulating cortisol levels and the
incidence of hypertension and diabetes. Phillips et al. (1998) have shown that fast-
ing plasma cortisol levels in men aged 64 years were inversely related to birth
weight, independent of body mass index (BMI), and that elevated cortisol levels
were signi¬cantly associated with higher blood pressure, plasma glucose levels,
fasting triglyceride levels and insulin resistance. More recently, low birth weight
has been associated with elevated fasting and stimulated cortisol concentrations in
adult human beings (Levitt et al., 2000; Phillips et al., 2000; Reynolds et al., 2001).
In each case, cortisol levels were positively associated with high-blood pressure and
in some populations, associated with glucose intolerance (Levitt et al., 2000; Reynolds
et al., 2001). These observations support a role for the programming of HPA axis
function in the predisposition to adult disease. Nilsson et al. (2001) found that
men with lower birth weight and a small head circumference at birth scored poorly
on psychological assessment surveys compared to their heavier counterparts. It was
suggested that impaired fetal growth was predictive of suboptimal psychological
functioning and increased stress susceptibility. Although the mechanisms regulat-


. 22
( 51 .)