Acta Obstetricia et Gynecologica.

2009; 88: 818
823

ORIGINAL ARTICLE

Neuroendocrine response to violence during pregnancy

on duration of pregnancy and fetal growth

impact

A2, MARY ELLSBERG4,

ELIETTE VALLADARES1,2,3, RODOLFO PEN
KE PERSSON5 & ULF HO
GBERG3,6
LARS A
1

Department of Obstetrics and Gynecology, National Autonomous University of Leon, Leon, Nicaragua, 2Center for Research
on Health and Demography, Faculty of Medical Sciences, National Autonomous University of Leon, Leon, Nicaragua,
3
Epidemiology, Department of Public Health and Clinical Medicine, Umea University, Umea, Sweden, 4Program for
Appropriate Technology in Health, Gender, Violence, and Human Rights, Washington, DC, USA, 5International Maternal
and Child Health (IMCH), Department of Womens and Childrens Health, Uppsala University, Uppsala, Sweden, and
6
Obstetrics and Gynecology, Department of Clinical Sciences, Umea University, Umea, Sweden

Abstract
Objective. To study the neuroendocrine release of cortisol in response to perceived stress among pregnant women exposed to
partner violence and how this affects the duration of pregnancy and the intrauterine growth of the infant. Design. Crosssectional community-based study. Setting. Health and Demographic Surveillance System of Leon, Nicaragua. Population.
One-hundred and forty-seven pregnant women. Methods. Standardized scales to measure intimate partner violence, social
resources, perceived stress, and socio-economic conditions were applied. Two salivary samples for cortisol were collected in
the morning and afternoon on the same day. Linear regression and path analysis were used. Main outcome measures. Cortisol
levels, gestational age, and weight at delivery. Results. Partner violence during the pregnancy, low social resources, and
perceived maternal stress were associated with high level of salivary cortisol. Pregnant women with high cortisol levels were
significantly more likely to give birth to small-for-gestational age babies, but not to deliver preterm. A substantial decrease of
birthweight, 121
186 g, was associated with an increase in cortisol in association with violence exposure. Conclusion. Partner
violence during pregnancy is a stressor that provokes high levels of cortisol, which is associated with reduction of
birthweight.

Key words: Violence, pregnancy, birthweight, small-for-gestational age, cortisol

Introduction
Violence during pregnancy increases the risk of severe
physical and psychological harm for both women and
their offspring (1
4). These adverse effects can be due
either to direct trauma or a consequence of stress
produced by the abuse (2). In particular, physical and
sexual violence by intimate partners has been associated with increased levels of emotional distress,
depression, anxiety disorders, suicidal ideas, and
behavioral problems (5).
According to a recent study, maternal exposure to
stress can influence the hypothalamic pituitary axis
hormones, although absolute differences in hormonal

blood levels have not been found (6). Evidence of

a connection between maternal stress and negative
outcomes, such as preterm delivery and fetal growth
reduction, was observed in other studies (7
9).
Austin et al. found that the consistent association of
socio-demographic status and preterm birth could be
mediated through maternal stress, but they also
concluded that low social support, heavy workload,
and abuse can be intermediates as well (10).
As indicated by several animal studies, maternal
exposure to stress during pregnancy results in increased sensitivity to induced hypothalamic-pituitaryadrenal (HPA) hormone secretion in the mother as

Correspondence: Eliette Valladares, Departamento de Ginecologa y Obstetricia, Hospital Escuela Oscar Danilo Rosales Arguello, Apartado postal 297, Leon,
Nicaragua. E-mail: eliette.valladares@epiph.umu.se

(Received 11 September 2008; accepted 29 April 2009)

ISSN 0001-6349 print/ISSN 1600-0412 online # 2009 Informa UK Ltd. (Informa Healthcare, Taylor & Francis AS)
DOI: 10.1080/00016340903015321

Neuroendocrine response to violence 819

well as in the offspring (11,12). High levels of
maternal hypothalamic, pituitary, and placental hormones could trigger labor (13,14) as well as decrease
the utero-placental perfusion due to the vasoconstriction and hypoxia in response to sympathetic adrenal
pituitary activation (15). Although levels of salivary
cortisol are increased two-fold during pregnancy, the
cortisol response to awakening and stress are similar
for non-pregnant subjects (16).
To date, the neuroendocrine release of cortisol
in response to perceived stress in abused pregnant
women has not been examined. The objective of
this study was to analyze this phenomenon as part of
the mechanism through which violence may affect
the length of the pregnancy and the growth of the
unborn child.
Material and methods
The study sample was composed of 147 pregnant
women 14
40 years old and 18
39 weeks pregnant at
the time of the cortisol sampling and was consecutively recruited within a larger sample of 478 total
pregnant women identified from November 2002
to January 2004 through an ongoing community
demographic surveillance conducted in the municipality of Leo n, the second largest city in Nicaragua
(17). Using a questionnaire and ensuring complete
privacy, two trained field workers conducted the
interviews in the womens houses to collect information that addressed intimate partner violence
(18,19), emotional distress, socio-economical aspects, and social resources (17).
One trained field worker collected two saliva
samples from each subject. Samples were taken the
same day, the first between 7:00 and 8:00 AM and the
second between 14:00 and 15:00 PM. Both samples
were collected in the womens houses. Samples were
returned to the laboratory on the same day in
temperature-control containers.
Salivary cortisol was measured as a dynamic test
of adrenal function since it has been shown to reflect
the biologically active serum concentration of the
hormone (20). In addition, saliva samples are less
stressful and not invasive as compared to blood
samples. It was measured using the Salimetric HSCortisol Kit, an Elisa Immunoassay (product no. 11102; Salimetrics LLC, State College, PA 16803,
USA) specifically designed and validated for quantitative measurement of salivary cortisol (ng/ml)
(21). The test was performed by a certified laboratory in Leo n; 30% of the samples were randomly
chosen and tested by the Microbiology Department
of Leo n University to validate and confirm the

accuracy of the tests. Both laboratories used the

same kit and followed the same protocols and
instructions without modification. Both laboratories
assayed statistically significant number of controls to
establish their own mean values and ranges to assure
validity. The intraclass correlation coefficient for the
morning values between the two labs was 0.79 and
for the afternoon values 0.73. The differences
between the results of the labs were within 92 SD
for 88 and 87% of the morning and afternoon
samples, respectively.
According to WHO, low birthweight (LBW) is
defined as a birthweight below 2,500 g, and preterm
birth is defined as less than 37 completed weeks
of gestation (22). Gestational age at birth was
calculated from the date of the last menstruation
period (LMP) and the Capurro method was used for
mothers who did not remember the date (23). Both
methods were correlated in the same setting showing
a good k agreement (0.80) (3). Intrauterine growth
was calculated by relating birthweight to the gestational age at birth against reference curves. A birthweight below the tenth percentile of the curves was
considered small-for-gestational age (SGA) (22).
In theory, the release of cortisol as a response to
stress increases vasoactive substances that provoke
vasoconstriction in the placental vessels and thus
could lead to fetal growth reduction or preterm
labor. Release of cortisol can also be the result of
a direct hit to the abdomen, a trauma that can result
in the production of prostaglandins, uterine contractions, and preterm birth. However, to estimate
properly a possible effect of violence, other factors
such as socio-economic status and social resources
should be considered (Figure 1).
For the analysis of the cortisol results, the mean and
standard deviation of the morning and afternoon
values were calculated. Since an increase in the levels
of cortisol is related to increased gestational age,
mainly during the third trimester, unstandardized
residuals of salivary cortisol were calculated using
Growth
retardation

Violence
during
pregnancy

Reduced
birth weight

Maternal
stress

Preterm
birth
Low socioeconomic position
Low social support
Drug abuse
STD

Figure 1. Conceptual framework of mechanisms of violence

during pregnancy affecting birth weight.

E. Valladares et al.

linear regression analysis including AM and PM

salivary cortisol, delta-values, and gestational week
at sampling. These data were used for further analysis.
For the final model, only PM salivary cortisol was kept
for analysis because that was strongly associated with
both abuse in pregnancy and reduction in birthweight. t-Tests were calculated to examine the relation
of the explanatory variables and outcomes of interest.
Finally, based on the theoretical model and the
findings from the linear regressions and t-tests performed, a hypothetical Path Model was built using
Amos, a structural equation modeling software. The
model was used to calculate the predictive values of
the different pathways through which abuse during
pregnancy can affect birthweight. The chi-squared
test was used to measure the goodness of fit (x2/df) of
the model.
The WHO ethical guidelines for research on
domestic violence were followed (24). Informed
and signed consent was obtained from all participants during every visit. Ethical approval was obtained from the Ethical Committee of the Faculty of
Medicine at the University of Leo n, and from the
Research Review Committee at the Medical Faculty,
Umea University, Sweden. Services for counseling
and treatment (psychological, medical, or legal) were
offered to the participating women; 20% of the
participants requested some sort of aid.
Results
Forty percent of the women were primiparous, 33%
were adolescents, 71% had low socio-economic
conditions, and 61% were living in rural areas.
Emotional partner violence during the index pregnancy were reported by 44% of the participants,
physical abuse by 17%, and sexual aggression by
8%. Scale measuring perceived stress was positive
in 43%, 35% had low social resources, and 7%
had attempted suicide during the index pregnancy.
Thirteen percent gave birth to a LBW baby, 6% were
term and SGA, 4% were preterm, and 3% were both
preterm and SGA.
The mean of the salivary cortisol values measured
were 10.1 ng/ml91.9 SD in the morning and 3.0 ng/
ml92.8 SD in the afternoon. A trend of increasing
salivary cortisol related to the increase in gestational
age was found (Figure 2). The mean of the unstandardized residuals of PM cortisol for abused subjects
(65) was 1.292.9; the corresponding number for
non-abused subjects (82) 0.992.0 (p 0.000). Women who reported violence during pregnancy were
also more likely to give birth to LBW babies (p 
0.001) than their non-abused peers. A significant
association was found between violence during

16
14
12
10

Mean

820

8
6
4
Cortisol am

Cortisol pm

0
17

21
19

25
23

29
27

33
31

37
35

41
39

Gestational week at the cortisol test

Figure 2. AM and PM values of salivary cortisol per gestational
week among the pregnant population (n 147).

pregnancy and SGA births (p 0.000), but not

with preterm deliveries (p 0.298).
Low social resources and high-perceived maternal
stress during pregnancy were related to an increase
of the unstandardized residuals of salivary PM
cortisol (p0.000 and 0.001, respectively). In contrast, maternal age, parity, place of residence, and
poverty did not show any relation with cortisol levels.
Increased unstandardized residuals of the afternoon
levels of salivary cortisol were significantly associated
to LBW (p 0.000) and SGA infants (p 0.000),
although the association in relation to preterm
delivery did not reach significance (p 0.082).
Taking into account the three hypothesized pathways, structural equation estimations were used
to relate violence during pregnancy to birthweight
(Figures 1 and 3). Social resources were included in
the model, but not poverty as it was not significantly
associated to either abuse during pregnancy or
reduction of birthweight in this setting (Table I and
Figure 3).
The estimations of the first pathway showed that
violence during pregnancy was associated with increased PM cortisol level and reduced birthweight
by an average of 79 g. In the second, violence during
pregnancy was associated with increased PM cortisol levels, decreased gestational age at delivery and
reduced birthweight by 42 g. The third showed that
violence during pregnancy through direct abdominal
trauma was associated to reduced gestational age at
delivery and decreased birthweight by an average of
65 g (Table I and Figure 3).
Low social resources were found to be associated
with increased PM cortisol levels (p 0.000) and a
reduction of the birthweight of 43 g. Although not
significant, low social resources were also associated

Neuroendocrine response to violence 821

Figure 3. Estimates of potential influence of exposure to violence

during pregnancy on reduction of birth weight (g) by different
hypothetical paths, by increasing risk of preterm birth, and by a
maternal stress response with both increasing risk of preterm birth
and growth retardation.

to reduced gestational age at delivery (p 0.453) and

a decrease in the birthweight of 33 g.
The goodness of fit (x2/df) of the model was of
0.047 with p-value of 0.089.
Discussion
As cortisol increased over the course of the pregnancy and the cortisol samples were taken at different
pregnancy weeks, gestational week was a major
confounding factor in this study. Therefore, unstandardized residuals of cortisol after adjustment for
gestational weeks at time of sampling were used for
the analysis.
Increased residuals of PM salivary cortisol were
associated with physical or sexual violence during
pregnancy by an intimate partner, low availability of
social resources, and emotional distress. Pregnant
women with increased residuals of PM cortisol were
significantly more likely to give birth to SGA babies.
A trend was also seen with regard to cortisol and
preterm babies. A substantial decrease of birthweight,

121 g (7942), was associated with an increase in

cortisol due to violence exposure. Overall, violence
during pregnancy, including both increased cortisol
and decreased gestational length, reduced the birthweight by 186 g (794265). When we add the
effects of both violence and low social resources on
birthweight, we found a total reduction of 262 g
(Figure 3).
The community-based design increased the external validity of the results. Skilled interviewers and a
private environment for the interviews enabled the
disclosure of violence and probably minimized underreporting. The second visit at the end of pregnancy increased the reported prevalence of violence
by three times because the incidence of violent
acts between the two visits, the increase in subjects
confidence and the repeated opportunity to disclose
abuse. Recall bias should have been minor with the
prospective design and standardized and pre-tested
questionnaires should have minimized information
bias. The quality of the cortisol analysis, interviews,
and data entry were secured by internal and external
validity procedures.
Few researchers have explored the normal patterns of increased cortisol during pregnancy. In the
present study, the levels of salivary cortisol showed
an increased trend as gestational length advances,
which is consistent with previous findings (16,20).
Perceived maternal stress and low social resources
were common among the study population. As
described before, low social resources were significantly related to high-perceived maternal stress and
to adverse effects on the birthweight (25,26).
A consistent association between low socioeconomic status and preterm births has been reported
from high-income countries (10). In high-income
settings only 1/3 of the LBW children have intrauterine growth reduction, but in low-income countries, as
in this study, 2/3 of low birth infants have that
condition (3). This may explain why in this study
there was a stronger association between maternal
stress and growth reduction than to preterm births.

Table I. Estimations, coefficients, and p-values of the Structural Equations in the hypothesized Path Model relating violence during
pregnancy and childs weight at birth.

Residual PM cortisol
Residual PM cortisol
Social resources
Gestational age at birth
Gestational age at birth
Gestational age at birth
Childs weight at birth
Childs weight at birth

B{{{Fx
B{{{Fx
B{{{Fx
B{{{Fx
B{{{Fx
B{{{Fx
B{{{Fx
B{{{Fx

Violence in pregnancy
Social resources
Violence in pregnancy
Residual PM cortisol
Violence in pregnancy
Social resources
Residual PM cortisol
Gestational age at birth

Estimate

p-Value

1.819
0.922
0.314
141
397
204
43.628
163.437

B0.001
B0.001
0.005
0.144
0.453
B0.001
B0.001

822

E. Valladares et al.

Previous research has indicated maternal stress

as an intermediary in the association between
socio-economic position and preterm birth (10). In
contrast, our results showed maternal stress as an
intermediary between exposure to violence on the one
hand, and preterm birth and growth reduction on the
other, irrespective of social resources and poverty.
Suicide attempts during pregnancy were highly
prevalent compared to what has been reported from
a hospital surveillance in the same setting (27).
Pregnant women with high-perceived stress were
significantly more likely to attempt suicide during
the index pregnancy compared to non-abused women having high-perceived stress. Moreover, the
women attempting suicide had significantly higher
levels of cortisol, and were probably most emotionally affected by stress.
The proportion of LBW, SGA, and preterm
infants in this study was comparable with national
statistics. As in previous results from the same
setting, poverty was not associated with abuse during
pregnancy (28). This can be partially explained by
the widespread distribution of poverty among the
study population and that violence against women
crosses all social frontiers.
Our results showed that maternal stress, measured
as increased cortisol, could be viewed as the mechanism through which violence during pregnancy
decreases the infants weight at birth, comparable to
the reduction produced by smoking (29).
As in other studies, we found that the association
between psychosocial factors, neuroendocrine response and birth outcomes is complex and may
depend on the time of occurrence, the number of
psychosocial stressors, and the nature of these stressors (10,30).
The p-value of the entire path model applied
in this study (p 0.089) indicates a good fit because
the null hypothesis (p B0.05)
that the model fits
the data
cannot be rejected.
Declaration of interest: The authors report no
conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

4. Rodrigues T, Rocha L, Barros H. Physical abuse during

pregnancy and preterm delivery. Am J Obstet Gynecol. 2008;
171:e171
6.
5. Ellsberg M, Caldera T, Herrera A, Winkvist A, Kullgren G.
Domestic violence and emotional distress among Nicaraguan
women: results from a population based study. Am Psychol.
1999;54:30
6.
6. Talley P, Heitkemper M, Chicz-Demet A, Sandman C. Male
violence, stress and neuroendocrine parameters in pregnancy:
a pilot study. Biol Res Nurs. 2006;7:222
33.
7. Hedegaard M, Herricksen TB, Secher NJ, Hatch MC, Sabroe
S. Do stressful life events affect duration of gestation and risk
of preterm delivery? Epidemiology. 1996;7:339
45.
8. Aarts MC, Vingerhoets AJ. Psychosocial factors and intrauterine fetal growth: a prospective study. J Psychosom Obstet
Gynaecol. 1993;14:249
58.
9. Cooper RL, Goldenberg RL, Das A, Elder N, Swain M,
Norman G. The preterm prediction study: maternal stress is
associated with spontaneous preterm birth at less than thirty
five weeks gestation. Am J Obstet Gynecol. 1996;175:
1286
92.
10. Austin MP, Leader L. Maternal stress and obstetric and
infant outcomes: epidemiological findings and neuroendocrine mechanisms. Aust N Z J Obstet Gynaecol. 2000;40:
331
7.
11. Henry C, Kabbaj M, Simon H, Le Moal M, Maccari S.
Prenatal stress increases the hypothalamo-pituitary-adrenal
axis response in young and adult rats. J Neuroendocrinol.
1994;6:341
5.
12. Takahashi LK, Turner JG, Kalin NH. Prolonged stressinduced elevation in plasma corticosterone during pregnancy
in the rat: implications for prenatal stress studies. Psychoneuroendocrinology. 1998;23:571
81.
13. Majzoub JA, McGregor JA, Lockwood CJ, Smith R, Taggart
MS, Schulkin J. A central theory of preterm and term labor:
putative role for corticotropin-releasing hormone. Am J
Obstet Gynecol. 1999;180:S232
41.
14. McLean M, Bisits A, Davies J, Woods R, Lowry P, Smith R.
A placental clock controlling the length of human pregnancy.
Nat Med. 1995;1:460
3.
15. Cosmi EV, Luzi G, Gori F, Chiodi A. Response of uteroplacental fetal blood flow to stress situation and drugs. Eur
J Obstet Gynecol Reprod Biol. 1990;36:239.
16. De Weerth C, Gispen-De Wied C, Jansen LMC, Buitelaar
JK. Cardiovascular and cortisol responses to a psychological
stressor during pregnancy. Acta Obstet Gynecol Scand. 2007;
86:1181
92.
17. Valladares E, Pen a R, Persson L, Ho gberg U. Violence against
pregnant women: prevalence and characteristics. A population-based study in Nicaragua. BJOG. 2005;112:1243
8.
18. Garca-Moreno C, Heise L, Ellsberg M. WHO multi-country
study on woman health and life events. Final core questionnaire version 9.9. Geneva: World Health Organisation;
2002.
19. Garcia-Moreno CJH, Ellsberg M, Heise L, Watts CH.
Prevalence of intimate partner violence: findings from the
WHO multi-country study on womens health and domestic
violence. Lancet. 2006;368:1260
9.
20. Vining RF, McGinley RA, Maksvytis JJ, Ho KY. Salivary
cortisol: a better measure of adrenal cortical function than
serum cortisol. Ann Clin Biochem. 1983;20:329
35.
21. Raff H, Homar PJ, Skoner DP. New enzyme immunoassay for
salivary cortisol. Clin Chem. 2003;49:203
4.
22. WHO: Physical status: the use and interpretation of anthropometry. Infant and children. Geneva: World Health Organisation. Tech Rep Ser. 1995;854:161
262.
/

1. Bullock LF, McFarlane J. The birth-weight/battering connection. Am J Nurs. 1989;89:1153
5.

2. Parker B, McFarlane J, Soeken K. Abuse during pregnancy:
effects on maternal complications and birthweight in adult
and teenage women. Obstet Gynecol. 1994;84:323
8.
3. Valladares E, Ellsberg M, Pena R, Hogberg U, Persson LA.
Physical partner abuse during pregnancy: a risk factor for low
birthweight in Nicaragua. Obstet Gynecol. 2002;100:700
5.

Neuroendocrine response to violence 823

23. Capurro H, Konichezky S, Fonseca D, Caldeyro-Barcia R.
A simplified method for diagnosis of gestational age in the
newborn infant. J Pediatr. 1978;93:120
2.
24. WHO: Ethical and safety recommendations for research on
domestic violence against women. Geneva: World Health
Organisation; 1999.
25. Lobel M. Conceptualizations, measurement, and effects of
prenatal maternal stress on birth outcomes. J Behav Med.
1994;17:225
72.
26. Heins HC Jr, Nance NW, Ferguson JE. Social support
in improving perinatal outcome: The Resource Mothers
Program. Obstet Gynecol. 1987;70:263
6.
27. Caldera T, Herrera A, Renberg ES, Kullgren G. Parasuicide
in a low-income country: results from three-year hospital
/

surveillance in Nicaragua. Scand J Public Health. 2004;32:

349
55.
28. Ellsberg M, Pena R, Herrera A, Liljestrand J, Winkvist A.
Candies in hell: womens experiences of violence in Nicaragua. Soc Sci Med. 2000;51:1595
610.
29. Wang X, Zuckerman B, Pearson C, Kaufman G, Chen C,
Wang G, et al. Maternal cigarette smoking, metabolic
gene polymorphism, and infant birthweight. JAMA.
2002;287:195
202
30. Wadhwa PD, Dunkel-Schetter C, Chicz-DeMet A, Porto M,
Sandman CA. Prenatal psychosocial factors and the neuroendocrine axis in human pregnancy. Psychosom Med. 1996;58:
432
46.
/