. 34
( 51 .)


tive design and objective behavioral observations of the child (Huizink et al.,
2002). This study found that the infants of mothers who reported higher levels of
anxiety during pregnancy displayed poorer attention regulation. Owing to the dif-
¬culty of conducting prospective studies, very few exist. There is a need for further
prospective human studies that employ objective assessments of child behavior to
elucidate the independent contribution of postnatal maternal psychological state
on development.
To differentiate the effects of prenatal and postnatal maternal psychological state,
maternal anxiety and depression were assessed prospectively. Infant behavioral
189 In¬‚uence of stress in human fetal and infant development

reactivity was assessed at 4 months using a standardized laboratory-based behav-
ioral assessment protocol (i.e. the Harvard Infant Behavioral Reactivity Protocol,
Kagan and Snidman, 1991). In this paradigm infant motor and cry reactivity to a
series of visual and auditory challenges were assessed. Maternal anxiety and
depression during the third trimester of pregnancy, but not postpartum were asso-
ciated with the development of individual differences in infants™ behavioral regu-
lation. The offspring of mothers who were higher in anxiety and depression during
pregnancy displayed greater behavioral reactivity to novelty. Notably, this associa-
tion remained after controlling for postpartum maternal psychological state indi-
cating that prenatal experiences were responsible for this association (Davis et al.,
2004a). The selective effects of prenatal experiences on behavioral reactivity sup-
ports the hypothesis that the prenatal environment exerts programming effects
on the fetus with consequences for infant behavior (Barker, 2002). These data sup-
port a model that prenatal maternal stress has an independent effect not only on
regulation of length of gestation but also on development of the fetus and thus the

Placental CRH and fetal growth and premature birth
The maternal“HPA axis is one mechanism that has been proposed to mediate the
effects of maternal stress during pregnancy on birth outcome and the development
of the fetus. During pregnancy maternal ACTH and cortisol increases in response to
stress in ways that are similar to the non-pregnant state (Wadhwa et al., 1996). Via
this pathway, maternal stress can modulate placental CRH production. Placental
CRH is involved in the physiology of parturition as well as fetal cellular differ-
entiation, growth, and maturation (Challis et al., 2001; Smith, 2001; Hillhouse and
Grammatopoulos, 2002). We have conducted several studies to examine the role of
CRH in regulation of timing of delivery and fetal growth. The ¬rst study involved a
sample of 63 women with singleton, intrauterine pregnancies. Maternal plasma was
collected at 28“30 weeks gestation, and placental CRH concentrations were deter-
mined by radioimmuno assay. Results indicated that maternal (placental) CRH
levels at 28“30 weeks gestation signi¬cantly and negatively predicted gestational
length after adjusting for antepartum risk. Moreover, subjects who delivered pre-
maturely (prior to 37 weeks gestation) had signi¬cantly higher CRH levels in the
early third trimester than those who delivered at term (Wadhwa et al., 1998).
To explore further the associations between CRH and gestational length and to
examine effects of CRH on fetal growth, 245 women with singleton, intrauterine
pregnancies were recruited. Maternal plasma CRH was assessed at 32“33 weeks
gestation. It was found that elevated CRH was related to both risk of preterm birth
and fetal growth restriction. After adjusting for effects of established risk factors
women with elevated CRH were approximately 3 times more likely to deliver
190 E. P. Davis et al.


CRH at 33 weeks (pg/ml)




Term AGA Preterm AGA Term SGA Preterm SGA
Figure 6.2 Infants born preterm or SGA more likely to be exposed to high levels of CRH during the
third trimester of pregnancy. Adapted from Wadhwa et al. (2004). SGA: small for
their gestational age; AGA: appropriate for gestational age

preterm and/or have an infant that was small for GA (see Figure 6.2; Wadhwa et al.,
2004). The results of this study suggest that placental CRH plays a role in the phys-
iology of parturition as well as in processes related to fetal growth.
These studies suggest that increased activity by the placental“HPA axis during
the third trimester predicts premature labor. The question remains as to whether
increased placental CRH earlier in pregnancy might also predict length of gestation.
To address this issue CRH was assessed in 524 women at 18“20, 28“30, and
35“36 weeks gestation. Eighteen women with spontaneous premature labor
were compared to 18 women who delivered at term. Patients who delivered pre-
maturely had higher levels of CRH at all three measurement time points (Hobel
et al., 1999a). Furthermore, women who delivered prematurely had lower levels of
CRH-binding protein, which inactivates CRH (Hobel et al., 1999a). Thus, maternal
CRH was elevated as early as 18“20 weeks GA in woman who subsequently
delivered prematurely.

Neuroendocrine function during pregnancy and human fetal CNS development
It has been proposed that the HPA“placental axis is a conduit for the effects of
environmental stress on the fetus. Research with animals indicates that during
prenatal development the hormones of the HPA axis have programming effects on
the developing CNS (Matthews, 2000; Welberg and Seckl, 2001). The in¬‚uence of
the maternal and intrauterine environment on the developing human fetal brain
is poorly understood. This is in part, because the assessment and quanti¬cation
191 In¬‚uence of stress in human fetal and infant development

of human fetal brain development presents theoretical and methodological
To quantify and examine the in¬‚uence of the fetal environment on its brain
development we have utilized a habituation“dishabituation paradigm that assesses
the ability of the fetus to learn information. By 32 weeks gestation the fetus habit-
uates and dishabituates to external stimulation (Sandman et al., 1997b). Faster
fetal habituation has been associated with advancing GA (Shalev et al., 1990) con-
sistent with maturation of the CNS. We examined the effect of maternal“placental
CRH on habituation processes. Thirty-three pregnant women were assessed
between 30 and 32 weeks gestation. Fetal heart rate and uterine contractions were
assessed by placing transducers on the maternal abdomen. A total of 41 trials of
vibroacoustic (VA) stimuli were presented over a 45-min period. The ¬rst series of
15 VA (63 dB, 300 Hz) stimuli (S1) was presented on the maternal abdomen. On
the 16th trial S2, the dishabituating VA stimuli (68 dB, 400 Hz, novel in frequency
and intensity) was presented. The original VA stimuli (S1) was then presented for
trials 17“31. As a control the ¬nal 10 stimuli were presented to the mother™s thigh.
The fetuses of mothers with highly elevated CRH levels did not respond signi¬-
cantly to the presence of the novel stimulus (Sandman et al., 1999). These data pro-
vide preliminary evidence that abnormally elevated levels of placental CRH may
play a role in impaired neurodevelopment, as assessed by the degree of dishabituation
(Sandman et al., 1999).
In addition to the effects of placental CRH on fetal CNS development described
above, maternal pituitary and adrenal hormones may also shape fetal develop-
ment. The in¬‚uence of circulating maternal ACTH and E levels with measures of
fetal responses to challenge was determined in a sample of 132 women at 31“32
weeks gestation. Fetal responses were measured by measuring heart rate (HR)
habituation to a series of repeated VA stimuli. Individual differences in habituation
were determined by computing the number of consecutive HR responses that were
greater than the standard deviation of the HR during a control (non-stimulated)
period. There was no signi¬cant relation between absolute levels of ACTH, E and
fetal HR responses to challenge. However an index of POMC disregulation, the
degree of uncoupling between ACTH and E, was signi¬cantly related to fetal
responses such that fetal exposure to relatively high levels of the maternal opiate,
E, relative to ACTH, was associated with a signi¬cantly lower rate of habituation
(see Figure 6.3; Sandman et al., 2003).
Our ¬ndings are consistent with those of longitudinal investigations of the
functional development of the human fetal CNS over the course of gestation, that
have suggested chronic maternal psychologic distress is signi¬cantly related to
measures of fetal neurobehavioral maturation and reactivity (DiPietro et al., 1996;
DiPietro et al., 2000; Monk et al., 2000; DiPietro et al., 2002; Monk et al., 2003).
192 E. P. Davis et al.


Ratio of E to ACTH





1 2 3 4
Rate of habituation in quartiles
Figure 6.3 Fetal exposure to relatively high levels of the maternal opiate, E, relative to ACTH, is
associated with a signi¬cantly lower rate of habituation. Adapted from Sandman
et al. (2003)

The continuity of development from prenatal to postnatal development requires
further exploration.

Endocrine hormones during pregnancy and human infant development
Demonstration of an impact of maternal and placental hormones on fetal CNS
functioning illustrates the importance of exploring the implications of fetal expe-
riences on development in infancy and childhood. In rodents, primates and other
species it has been shown that stimulation of the HPA axis or exposure to elevated
glucocorticoids impairs brain development and HPA axis functioning in the off-
spring. There is transplacental passage of glucocorticoids to the fetus (Matthews
et al., 2002). Animals exposed to prenatal elevations in glucocorticoids display
impairments in brain development and increased reactivity to stress (Takahashi,
1998; Matthews et al., 2002; Antonow-Schlorke et al., 2003).
One method for examining the effects of HPA axis hormone disregulation on
human infant development involves examination of the effects of administration
of synthetic glucocorticoids to women during pregnancy. Antenatal glucocorticoid
administration is a standard of care for women at risk of premature delivery and
has been shown to reduce mortality and respiratory distress among preterm
infants born at less than 34 weeks gestation. However, studies with humans have
demonstrated that antenatal glucocorticoid exposure is associated with reduced
birth weight (Banks et al., 1999; French et al., 1999) and head circumference
193 In¬‚uence of stress in human fetal and infant development

(French et al., 1999; Abbasi et al., 2000). Additionally prenatal glucocorticoid treat-
ment effects postnatal HPA axis regulation in the offspring. Baseline cortisol levels
are suppressed for 2“7 days after prenatal corticosteroid treatment and subse-
quently return to normal levels (Wittekind et al., 1993; Parker et al., 1996; Kauppila
et al., 1978; Ballard et al., 1980; Dorr et al., 1989). These data suggest that prenatal
exposure to elevated levels of glucocorticoids may have implications for infant
development. The effect on the HPA axis response to stress has not, however, been
One of the sequelae of prenatal exposure to elevated glucocorticoids noted in
the animal literature is disregulation of the HPA axis response to stress (Matthews,
2002). We thus examined the effects of prenatal glucocorticoid treatment on the
cortisol response to stress during the ¬rst postnatal week in human infants born at
33“34 weeks with and without prenatal glucocorticoid treatment. Infants in the
glucocorticoid group were on average 12 days post antenatal glucocorticoid treat-
ment. Consistent with previous research demonstrating that baseline cortisol is
suppressed only for the ¬rst 2“7 days after prenatal treatment, these two groups of
infants did not differ in their resting baseline cortisol levels. Infants who were
exposed to antenatal glucocorticoid, however, failed to mount a cortisol response
to a painful stimulus, a heel-stick blood draw (see Figure 6.4). In contrast, prema-
ture infants who did not receive prenatal glucocorticoid treatment displayed an
increase in cortisol in response to the heel-stick stressor (Davis et al., 2004b). This
cortisol response, also displayed by full term infants, is considered appropriate and



Cortisol (ug/dl)






Baseline Response Recovery

No treatment Glucocorticoid treatment
Figure 6.4 The cortisol response to a painful event is suppressed in infants with antenatal
betamethasone treatment. Adapted from Davis et al. (2004b)
194 E. P. Davis et al.

supports adaptation to challenge (Gunnar, 1992). This study suggests that even
after baseline cortisol levels have returned to normal levels the ability to respond to
stress appears disregulated. This ¬nding is consistent with data indicating that
infants exposed to antenatal glucocorticoids displayed a suppressed response to the
CRH stimulation test (Ng et al., 2002).
Prenatal exposure to glucocorticoids seems to have a lasting effect on regulation
of physiologic stress responses in the newborn. Furthermore, as groups were simi-
lar in GA at birth and prenatal history, prenatal glucocorticoid exposure appears to
have a direct effect on postnatal stress physiology. We are currently conducting
longitudinal studies to examine whether this disregulation of the HPA axis
response to stress persists throughout infancy and early childhood.

Conclusions and future directions

Development is an epigenetic process by which, each developing organism plays an
active role in its own construction. This dynamic process is affected by systems that
are present during embryonic and fetal life to acquire information about the
nature of the environment, and to use this information to guide development. Due
to the rapid development that takes place during the prenatal period the fetus may


. 34
( 51 .)