. 33
( 51 .)


and lamb. J. Mol. Endocrinol., 14, 109“16.
Yau, J. L., Noble, J. and Seckl, J. R. (1997a). Site-speci¬c regulation of corticosteroid and sero-
tonin receptor subtype gene expression in the rat hippocampus following 3,4-methylene-
dioxymethamphetamine: role of corticosterone and serotonin. Neuroscience, 78(1), 111“21.
Yau, J. L. W., Noble, J., Widdowson, J. and Seckl, J. R. (1997b). Impact of adrenalectomy on
5-HT6 and 5-HT7 receptor gene expression in the rat hippocampus. Mol. Brain Res., 45, 182“6.
Yuen, B. S., McMillen, I. C., Symonds, M. E. and Owens, P. C. (1999). Abundance of leptin
mRNA in fetal adipose tissue is related to fetal body weight. J. Endocrinol., 163(3), R11“4.

Prenatal stress and stress physiology
in¬‚uences human fetal and infant
Elysia Poggi Davis1, Calvin J. Hobel2, Curt A. Sandman3,
Laura Glynn3 and Pathik D. Wadhwa4
Department of Psychiatry and Human Behavior, University of California, Irvine, California, USA
Department of Obstetrics and Gynecology, Cedars Sinai Medical Center, Los Angeles, California, USA
Department of Psychiatry and Human Behavior, University of California, Irvine, California, USA
Departments of Psychiatry and Human Behavior, and Obstetrics and Gynecology, University of California,
Irvine, California, USA

Prenatal stress has been proposed as a risk factor that may have developmental
consequences persisting throughout the lifespan. Exposing rodents to stress during
pregnancy has consequences for brain development, stress regulation, learning,
emotionality (increased anxiety), and social behavior (increased withdrawal) of
the offspring (Weinstock, 2001; Chapillon et al., 2002). Additionally, non-human
primates who experience stress during pregnancy have offspring with enhanced
behavioral reactivity to stressors later in life (Clarke et al., 1994), lowered levels
of motor behavior (Schneider, 1992), compromised neuromotor responses
(Schneider and Coe, 1993), irritable temperament (Schneider et al., 1992), and
attentional problems (Schneider et al., 1999).
Many researchers have focused on the hypothalamic“pituitary“adrenocortical
(HPA) axis, one of the body™s major stress systems, as a mechanism that may medi-
ate these effects (Ward and Phillips, 2001; Welberg and Seckl, 2001). The HPA axis
activity is regulated by the release of hypothalamic corticotropin-releasing hormone
(CRH) that stimulates the biosynthesis and release of adrenocorticotropin hormone
(ACTH) and -endorphin ( E) from the anterior pituitary. The release of ACTH
triggers the biosynthesis and release of glucocorticoids (cortisol in primates and cor-
ticosterone in rodents) from the adrenal cortex. Glucocorticoids are released into the

Corresponding author: Pathik D. Wadhwa, MD, PhD., Behavioral Perinatology Research Program,
University of California, Irvine, 3117 Gillespie Neuroscience Research Facility, Irvine, CA 92697. Tel: (949)
824-8238, Fax: (949) 824-8218, E-mail: pwadhwa@uci.edu
Preparation of this manuscript was supported, in part, by US PHS (NIH) grants HD-33506 and HD-41696
to P.D.W.

184 E. P. Davis et al.

general circulation and have effects on nearly every organ and tissue in the body
(Munck et al., 1984). Consequences of glucocorticoid release include energy mobi-
lization and immunosuppression (Chrousos and Gold, 1992). Glucocorticoids easily
pass through the blood“brain barrier (Zarrow et al., 1970). There are receptors for
glucocorticoids throughout the central nervous system (CNS) (de Kloet et al., 1998;
Sanchez et al., 2000). Glucocorticoids regulate their own release by negative feedback
actions at the hypothalamus and pituitary inhibiting the release of CRH and ACTH.
Glucocorticoids additionally act on extrahypothalamic sites including the hip-
pocampus and frontal cortex further activating negative feedback regulation of CRH
production in the hypothalamus (Jacobson and Sapolsky, 1991; Sanchez et al., 2000).
In contrast glucocorticoids increase CRH production in extrahypothalamic brain
regions, such as the central nucleus of the amygdala (Swanson and Simmons, 1989;
Makino et al., 1994; Watts and Sanchez-Watts, 1995).
The HPA axis is particularly sensitive to early experiences. In rat pups, manipula-
tions, such as daily handling or maternal deprivation produce lifelong changes in
stress reactivity, fearful behavior, and cognitive functioning (Levine, 1957; Meaney
et al., 1988; Liu et al., 1997). Rodent models further suggest that prenatal stress has an
impact that persists through adulthood. The offspring of stressed dams display pro-
longed glucocorticoid responses to stress indicating that exposure to stress in utero
may impair negative feedback mechanisms (Weinstock et al., 1992; Henry et al., 1994;
Herman and Cullinan, 1997). The offspring of rodents stressed during pregnancy also
display an increase in behavioral signs of anxiety (Takahashi et al., 1992; Vallee et al.,
1997). Alterations of CRH regulation in the amygdala is a proposed mechanism for
this effect. The amygdala is considered to be the structure where fear-inducing sen-
sory and autonomic input and behavioral output converge. Prenatally stressed rats
display an increase in amygdala CRH (Cratty et al., 1995). Thus, elevations in amyg-
dala CRH, resulting from prenatal stress, may contribute to the increase in HPA axis
reactivity and anxiety seen in these animals. Cognitive functions are also impaired
in prenatally stressed animals. This may be particularly true for functions that are
dependant on the hippocampus, a structure that is vulnerable to elevations in gluco-
corticoids (Takahashi, 1998; McEwen, 1999). In sum, these studies suggest that prena-
tal stress has lasting implications for CNS development and function.
Animal studies have offered valuable insights into physiological mechanisms
that may be involved in mediating the effects of stressful maternal and intrauterine
environments on the developing organism. However, the generalizability of these
¬ndings from animals to humans is limited by the existence of inter-species differ-
ences in physiology and the developmental time-line. The timing of maturation of
the HPA axis relative to birth is highly species-speci¬c and is closely linked to land-
marks of brain development (Dobbing and Sands, 1979). In animals that give birth
to precocious offspring (sheep, guinea pigs, primates), maximal brain growth and
185 In¬‚uence of stress in human fetal and infant development

a large proportion of neuroendocrine maturation takes place in utero. By contrast,
in species that give birth to non-precocious offspring (rats, rabbits, mice), much of
neuroendocrine development occurs in the postnatal period (Dent et al., 2000).
A second major difference is that anthropoid primates are the only species known
to produce placental CRH during pregnancy.
The placenta expresses the genes for CRH (hCRHmRNA) and the preprotein for
ACTH and E (pro-opiomelanocortin, POMC). Placental CRH is identical to
hypothalamic CRH in structure, immunoreactivity, and bioactivity (Petraglia
et al., 1996). There is, however, one crucial difference in the regulation of hypo-
thalamic and placental CRH. In contrast to the negative control on hypothalamic
CRH, glucocorticoids stimulate the expression of hCRHmRNA in the placenta
creating a positive feedback loop that is similar to the central nucleus of the amyg-
dala (Schulkin, 1999). Placental CRH is released into the maternal and fetal circu-
lation, establishing a positive feedback loop that allows for the simultaneous
increase of CRH, ACTH, and cortisol in the maternal and fetal compartments over
the course of gestation (Petraglia et al., 1996; King et al., 2001).
The HPA“placental axis is a mechanism by which the environment shapes fetal
development. The activity of the HPA“placental axis is regulated by characteristics
of the maternal and intrauterine environment. Maternal cortisol, which crosses the
placenta, increases with maternal stress (Wadhwa et al., 1996). Furthermore, in
vitro and in vivo studies have demonstrated that placental CRH output is modu-
lated in a positive, dose“response manner by the major biological effectors of
stress, including cortisol (Korebrits et al., 1998; Marinoni et al., 1998). During
pregnancy CRH levels were positively correlated with ACTH and E (Wadhwa
et al., 1997). These ¬ndings support the premise that in human pregnancy placen-
tal CRH activity is modulated by maternal pituitary adrenal hormones. Both pla-
cental CRH and cortisol in turn may in¬‚uence fetal development. Placental CRH
is involved in the physiology of normal parturition and elevated CRH concen-
trations are associated with an increased risk for spontaneous preterm birth
(McLean et al., 1995; Hobel et al., 1999a; Erickson et al., 2001; Holzman et al., 2001;
Inder et al., 2001; Moawad et al., 2002). It has been proposed that the activity of
the maternal“HPA“placental axis during pregnancy programs the development of
the offspring™s HPA axis (Ward and Phillips, 2001; Matthews, 2002). Additionally,
placental CRH and cortisol may contribute to the organization of the fetal CNS
(Sandman et al., 1997a; Florio and Petraglia, 2001). Few studies have considered
the consequences of prenatal stress on human fetal behavior and fewer still have
assessed the effects of maternal stress on the continuum between the fetus and the
infant. We will discuss a neurobiological model of prenatal stress that proposes
the developmental consequences of maternal psychosocial stress are mediated,
in part, via maternal“placental“fetal neuroendocrine mechanisms.
186 E. P. Davis et al.

Methodological approaches

We have assessed the consequences of maternal stress during pregnancy on neuro-
endocrine processes and fetal and infant development using a range of techniques.
Primarily, we have employed longitudinal population-based cohort studies with a
combined sample of approximately 750 women with singleton, intrauterine preg-
nancies. Women were recruited at various time points in pregnancy starting in the
late ¬rst or second trimester of gestation and followed through delivery into the early
postpartum period. Participants were heterogeneous in terms of sociodemographic
and ethnic characteristics. Furthermore, based on conventional measures of obstetric
risk we have included approximately equal numbers of subjects at low- and high-risk
for adverse perinatal outcomes. In these studies standardized and validated interviews
and questionnaires were administered at multiple time points over gestation to assess:
(a) maternal psychosocial constructs including various forms of prenatal stress,
social support, personality characteristics, and attitudes towards pregnancy;
(b) maternal behaviors including diet and nutrition, physical activity, and smoking,
alcohol, and drug use;
(c) sociodemographic characteristics including age, marital status, various indica-
tors of socioeconomic status, and race/ethnicity.
Maternal and cord blood samples were collected during gestation and at delivery
for bioassays of stress hormones, including ACTH, E, cortisol and placental CRH.
Obstetric and birth outcomes were abstracted from the medical records. All preg-
nancies are dated by best obstetric estimate using last menstrual period and early
ultrasonographic con¬rmation. In a sub sample of 156 pregnancies, we have per-
formed fetal assessments in the early third trimester of gestation, including fetal
biometry, doppler ¬‚ow velocimetry of the uteroplacental circulation, and an
experimental challenge paradigm to quantify indices of fetal arousal, reactivity,
learning and habituation, assessed by fetal heart rate (FHR) responses to a series of
vibroacoustic (VA) stimuli. To examine direct effects of glucocorticoids on fetal
and infant HPA axis development, a sample of infants whose mother did or did not
receive synthetic glucocorticoids during their pregnancy were recruited. In this
population cortisol levels at baseline and in response to stress were assessed.

Research ¬ndings
Maternal stress during pregnancy and birth outcomes
Disruption of reproductive function in mammals is a well-known consequence of
stress. Results from experimental approaches in animal models support a causal
role for prenatal stress as a developmental teratogen (Weinstock, 2001). In humans,
studies examining the in¬‚uence of maternal stress during pregnancy have focused
primarily on length of gestation and fetal growth/size at birth, the two primary
187 In¬‚uence of stress in human fetal and infant development

indicators of newborn health. Using women™s self report of stress during preg-
nancy we have found that maternal psychosocial processes signi¬cantly in¬‚uence
both length of gestation and fetal growth and that this in¬‚uence is independent
of the effects of other established sociodemographic and obstetric risk factors
(Wadhwa et al., 1993; Rini et al., 1999; Feldman et al., 2000). Maternal stress has
differential effects depending on its timing during pregnancy. From a prospective
investigation of stress and stress physiology in pregnancy, 40 pregnant women
were identi¬ed who had experienced a 6.8 magnitude earthquake during preg-
nancy or shortly after delivery. The participants lived, on average, 50 miles from
the epicenter of the earthquake and were physically unaffected by the damage pro-
duced. The effect of exposure to the earthquake was linearly moderated by the stage
in gestation of its occurrence. Women who experienced the earthquake earlier in
their pregnancy had a signi¬cantly shorter gestational length than those who expe-
rienced it later in gestation (see Figure 6.1). This study supports the notion that the
timing of stress in pregnancy may be an important factor in determining its impact
on the length of human gestation (Glynn et al., 2001).
Our results are consistent with several population-based epidemiological studies
that have suggested that high levels of maternal psychosocial stress are independently
associated with a signi¬cant increase in the risk for prematurity and that effects are
observed across the entire range of the outcome distribution (Hedegaard et al., 1993;
Pritchard and Teo, 1994; Copper et al., 1996; Hedegaard et al., 1996; Misra et al.,
2001). Additionally, the effect size of maternal psychosocial processes in pregnancy
on prematurity-related outcomes is comparable to that of most other obstetric risk
factors suggesting that these processes warrant the same degree of consideration.


Gestational age at birth




1st trimester 2nd trimester 3rd trimester Post partum
Gestational age at time of earthquake

Figure 6.1 Stress during the ¬rst trimester of pregnancy signi¬cantly predicts shorter gestational
length. Adapted from Glynn et al. (2001)
188 E. P. Davis et al.

Maternal stress during pregnancy and infant developmental outcomes
Maternal psychological state during pregnancy seems to in¬‚uence birth outcome
in terms of length of gestation and fetal growth. The in¬‚uence of maternal experi-
ences during pregnancy on the development of the fetal CNS and the implications
for infant development has largely been neglected in human research. Existing
research considering the effects of prenatal experience on postnatal development
in humans is often limited by a failure to control for the effects of birth outcome.
For example, infants born prematurely or small for gestational age (GA) are at risk
for a wide variety of developmental problems (Peterson et al., 2003). It is necessary
to consider these factors to examine the independent in¬‚uence of prenatal stress
physiology on postnatal development.
Recent studies suggest that maternal anxiety, stress and depression during preg-
nancy, shape the fetal behavioral patterns (DiPietro et al., 2002; Monk et al., 2003)
and predict higher cortisol and norepinephrine and lower Brazelton scores in the
newborn (Jones et al., 1998; Lundy et al., 1999). To examine whether this in¬‚uence
continued into infancy we conducted preliminary studies to prospectively assess
the relationship between maternal stress during pregnancy and indices of infant
behavioral development. Forty-seven mother“infant pairs were assessed during
pregnancy and at 6 weeks after delivery. All infants in this sample were full term at
birth. Questionnaires were administered to mothers to assess pre- and postnatal
maternal anxiety and infant temperament. Infant fussiness was associated with
higher levels of maternal anxiety during the third trimester even after controlling
for postpartum maternal affect, intrapartum compromise, infant sex and birth
weight (Davis et al., 2003). These data are consistent with the few prospective stud-
ies present in the literature illustrating that maternal stress, anxiety, and depression
during pregnancy is related to emotional disturbances and dif¬cult temperament
in the offspring (Van den Bergh, 1990; Susman et al., 2001; O™Connor et al.,
2002a, b).
Subjective description of child behavior by the parent is confounded by the parent™s
psychological state at the time of reporting. One study identi¬ed an association
between maternal anxiety during pregnancy and child behavior using a prospec-


. 33
( 51 .)