Professional Documents
Culture Documents
Field 2002
Field 2002
Field 2002
3 , 2 6 0 – 2 6 6
OBSTETRICS
One hundred and sixty-six women were classified as neonatal outcome measures than correlations between
experiencing high or low anger during the second trimester of pregnancy depression or anxiety and neonatal outcome
pregnancy. The high-anger women also had high scores on measures (Field et al., 2000). These data highlight the
depression and anxiety scales. In a follow-up across
need to further explore the effects of anger as a primary
pregnancy, the fetuses of the high-anger women were noted
to be more active and to experience growth delays. The high- and/or comorbid pregnancy variable.
anger mothers’ high prenatal cortisol and adrenaline and low The somewhat limited literature on anger would pre-
dopamine and serotonin levels were mimicked by their dict that unexpressed or unacceptable (to the subject)
neonates’ high cortisol and low dopamine levels. The high- anger would be accompanied by depression and physi-
anger mothers and infants were also similar on their relative ological arousal. In one study, individuals who
right frontal EEG activation and their low vagal tone. Finally, considered anger an unacceptable emotion were more
the newborns of high-anger mothers had disorganised sleep likely to have a depressed mood and greater activation
patterns (greater indeterminate sleep and more state changes) of the sympathetic nervous system (Mancuso, 2000). In
For personal use only.
and less optimal performance on the Brazelton Neonatal another study, high anger-out individuals showed the
Behavior Assessment Scale (orientation, motor maturity and
depression). These data highlight the need for prenatal
greatest increases in adrenocorticotrophic hormone
intervention for elevated angry mood during pregnancy. concentrations during acute psychological stress (e.g.
public speaking) (al’Absi et al., 2000). In contrast, in a
study on controlled expression of anger, both depres-
Introduction sion and physiological arousal were reduced (Dahlen
Several recent studies have documented the negative and Deffenbacher, 2000). Similarly, constructive angry
effects of prenatal depression and anxiety on the fetus behaviour has been noted to be a significant predictor of
(Glover et al., 1999; Dieter et al., 2000; Field et al., lower physiological arousal (lower blood pressure)
2000) and the neonate (Abrams et al., 1995; Jones et al., even when controlling for anxiety and depression
1998; Lundy et al., 1999). These have included (Davidson et al., 2000a). Further, matching angry facial
increased fetal activity and delayed fetal growth, and at expressions in a backward masking paradigm has been
the neonatal stage non-optimal performance on the associated with distinct facial muscle reactions that
Brazelton scale, increased indeterminate sleep, reduced correspond with even a very short exposure (30 ms)
vagal tone, relative right-frontal EEG activation and (Dimberg et al., 2000), and with increased regional
neurotransmitter/neurohormone profiles matching the cerebral blood flow in the left and right amygdala as
depressed mothers’ prenatal profiles, including assessed by functional magnetic resonance imaging
increased cortisol and noradrenaline (Lundy et al., (Hariri et al., 2000). In the same study, labelling the
1999) and decreased dopamine and serotonin (Field et angry expression was associated with an increase in
al., 2000). Although less is known about the effects of regional cerebral blood flow in the right prefrontal
elevated prenatal anxiety on the fetus and neonate, cortex.
Glover et al. (1999) have suggested that anxiety effects Further evidence for these associations comes from
would be mediated by elevated noradrenaline levels and cognitive therapy (Dahlen and Deffenbacher, 2000) and
the resulting increase in uterine artery resistance, exercise (at least 2–3 times/week) interventions show-
decrease in blood flow and increased intrauterine ing that anger, depression, anxiety and physiological
growth retardation. Inasmuch as depression and anxiety arousal could be decreased even though, by at least one
are often comorbid (Field et al., 2000), their independent account, anger (anger-in and anger-out), aggression,
effects may be difficult to untangle. depression and anxiety have been notably stable across
Anger is still a third negative emotion that can occur at least an 8-year period (rs = 0·47–0·75) (Magai, 1999).
during pregnancy. Although pregnancy anger has not Intervention models that have been effective with
been the subject of any published papers, a recent study anxiety disorders have also been effectively used with
on pregnancy depression yielded more frequent anger-related difficulties (Brondolo et al., 1997).
Correspondence to: Tiffany Field PhD, Touch Research Institutes, University of Miami School of Medicine, Department of
Pediatrics (D-820), PO Box 016820, Miami, FL 33101 USA. Email: tfield@med.miami.ed u
ISSN 0144– 3615 print/ISSN 1364– 6893 online/02/030260– 07 Taylor & Francis Limited, 2002
DOI: 10.1080/01443610220130526
Prenatal anger effects 261
Further, data from other literature suggest that these the Center for Epidemiological Studies depression scale
three emotions are comorbid with depression, anger and (CES-D) (Radloff, 1977), the Spielberger trait anxiety
anxiety sharing many clinical features. For example, inventory (Spielberger et al., 1970), and the behavioral
depression and anger comorbidity were noted in a study inhibition system/behavioral approach system Scale
showing greater serotonergic dysfunction in depressed (BIS/BAS) (Carver and White, 1994). The ultrasound
individuals having anger attacks (Fava et al., 2000). sessions were coded for fetal activity, and the ultra-
The purpose of the present study was to determine sonographers’ estimates of weight and growth were
the effects of high anger as self-reported on the profile recorded. First-morning urine samples were obtained
of mood states scale (POMS) during pregnancy on fetal from the mothers during the prenatal visit and from
development and neonatal outcomes. In addition, an them and their newborns within 24 hours of delivery.
attempt was made to establish the relative amounts of The CES-D and POMS were given again shortly after
the outcome variance that could be explained by prena- the delivery, and the medical records were used to score
tal anger and the comorbid emotions, depression and the obstetric complications scale (OCS) and the
anxiety. postnatal complications scale (PNS) (Littman and
Parmelee, 1978). The Brazelton Neonatal Behavior
Assessment Scale (Brazelton, 1984) was also measured
Methods within the first 2 days after birth. In addition, at that
time the mothers’ and newborns’ EKGs and EEGs were
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
Participants
The women were recruited during the second trimester recorded and the neonate’s sleep states and activity
(mean = 20·1 weeks; range = 18–22 weeks) from were coded.
obstetrician–gynaecologists’ offices. The women’s
medical charts were reviewed to screen out any women
who showed recreational drug use during pregnancy Measures
based on routine drug screens. The final sample Profile of mood states (POMS). (McNair et al.,
included 166 pregnant women of lower–middle socio- 1971). The POMS consists of 15 items on rating
economic status (mean = 2·9 on the Hollingshead index, depressed mood, nine items on anxiety and 12 items on
1975). The women averaged 27·7 years of age, 37% anger. They are rated on five-point scales ranging from
were single, and their ethnicity was distributed: 35% (0) ‘not at all’ to (4) ‘extremely’ using words such as
African-American, 28% Hispanic, 28% white Non- ‘blue’ and ‘sad’. The scale has adequate concurrent
Hispanic and 9% other. Eighty-four per cent of the validity, good internal consistency (r= 0·95; McNair and
For personal use only.
infants were full term (mean = 37 weeks) and healthy Lorr, 1964) and it is an adequate measure of intervention
and 55% were female. Based on the profile of moods effectiveness (Pugatch et al., 1969).
anger scale scores the mothers were divided by a
median split into high-and low-anger groups. The two
groups differed on several demographic variables (see Center for epidemiological studies depression
Table I) including that the higher anger group was scale (CES-D; Radloff, 1977). TheCES-Disa20-
comprised of (1) younger mothers, P<0·05; (2) a greater item self-report scale designed to measure depressive
percentage of black mothers, P<0·05; (3) a greater per- symptoms including depressed mood, feelings of guilt,
centage who were single; and (4) a greater percentage worthlessness, helplessness and hopelessness, loss of
of males born to high-anger mothers. energy and sleep and appetite disturbances (Radloff and
Teri, 1986). The 20 symptoms are rated for frequency
(over the past week) from ‘rarely or none of the time’ to
Procedure ‘most or all of the time’. A summary score ranges from
At their prescribed ultrasound session (mean gesta- 0 to 60 by summing all items. Reliability and validity
tional age = 20·1 weeks) the mothers were given the have been acceptable across a variety of demographic
profile of mood states (POMS) (McNair et al., 1971), characteristics including age, education, geographic
Groups
Variables Low anger High anger F P
Mother age 2 8·47 2 6·92 3·8 4 0·05
Mother SES 2·81 3·0 4 2·2 8 ns
Mother ethnicity(%) c2 P
Anglo 3 3·1 2 2·6 9·7 4 0·05
Hispanic 27·1 2 7·8
Black 2 8·8 4 1·7
Other 1 1·0 7·9
Marital status (%) c2 P
Single 18·9 4 4·1 6·7 9 0·03
Married 8 1·1 5 2·9
Divorced 2·9
Infant sex (%) c2 P
Male 3 5·2 5 8·1 5·0 9 0·02
Female 64 ·8 4 1·9
26 2 T. Field et al.
area and racial, ethnic and language groups (Radloff First-morning urine samples. First-morning urine
and Teri, 1986). The CES-D was given to the mother at samples were collected from the women at their ultra-
the time of prenatal recruitment and again within 24 sound visit and from both the women and their
hours following delivery. newborns within 24 hours following delivery. Newborn
first-morning urine samples were collected in the
nursery. The samples were frozen and sent to Duke
The state trait anxiety inventory. (Spielberger et University to be analysed by technicians who were
al., 1970). This was used to assess levels of anxiety.
‘blind’ to group assignment. The urines were assayed
This 20-question four-point Likert scale forms a
for nor-adrenaline, adrenaline, dopamine, serotonin and
summed score from 20 to 80 for the trait subscale.
cortisol levels after correcting for creatinine levels.
Higher scores indicate greater anxiety. The state
Nor-adrenaline, adrenaline and cortisol levels were
anxiety scale is comprised of 20 items and assesses how assessed because of their previously reported positive
the subject feels at that moment in terms of severity association with prenatal anger (al’Absi et al., 2000;
(‘not at all’ to ‘very much so’). The trait anxiety scale is Field et al., 2000), dopamine because of a recent animal
comprised how the subject ‘typically feels’. Character- model implicating a negative association between
istic items include: ‘I feel nervous’ and ‘I feel calm’. dopamine levels and depression (Weiss et al., 1996)
The scores increase in response to stress and decrease and serotonin because of its association with anger
under relaxing conditions (Spielberger et al., 1970). (Fava et al., 2000).
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
EEG for a baseline measure of relative right or the natural log power data for both hemispheres in the
left frontal EEG activation. At the neonatal stage frontal and parietal regions. Frontal alpha laterality
the EEG was recorded while the infant was in a quiet ratios (FLR) were computed by obtaining the difference
alert state in a bassinet. The researcher was blind to the of the mean log power density scores of a right hemi-
mother’s depression status. EEG and behaviour were sphere site and its homologous left hemisphere site
recorded for 3 minutes while the infant was awake and (LnRight–LnLeft). A score of zero represents hemi-
alert. We have reported significant 1–3-month stability spheric symmetry, a negative score represents relative
of EEG asymmetry (r= 0·45) (Jones et al., 1996) as well right frontal activation and a positive score represents
as significant 3-month to 3-year stability (Jones et al., relative left frontal activation.
1997). At a separate time during the session the
mother’s EEG was recorded while she sat quietly with
her eyes open for 3 minutes. Vagal tone. For this measure heart rate was
recorded for a 15-minute period while the infant was in
an infant seat and the mother seated on a chair. Vagal
EEG recording. The EEG for both the mother and tone was assessed by placing 3 EKG electrodes on the
infant were recorded using Lycra stretchable caps infant’s chest and back regions. The electrodes were
(manufactured by Electro-Cap, Inc.) that were posi- connected to a Grass Model 12 Neurodata Acquisition
tioned on the subject’s head using anatomical system preamplifier with bandpass frequencies set at
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
landmarks (Blom and Anneveldt, 1982). Electrode gel 1.0 and 100 Hz and a gain of 2000. The output was
was injected into the electrodes at the following sites: directed to a Dell 325 D PC fitted with Analog Devices
F3, F4, P3, P4, T3, T4, O1, O2 and Cz (used as the RTI-815 A/D board. The data were streamed across the
reference during recording for the majority of subjects), computer screen at a sampling rate of 512 samples per
and impedances were brought below 5000 ohms. Addi- second and then saved to a hard disk using data
tional electrodes were positioned on the external acquisition software (Snapstream, v. 3·21, HEM Data
canthus and above the supra-orbit of the right eye to Corp. 1991). After scoring for artefact, EKG data were
record the subject’s EOG, which was used to determine converted to inter-beat intervals (IBI) using an EKG
horizontal and vertical eye movement artefacts. analysis program (James Long Company, Caroga Lake,
The EEG signals were obtained using a Grass Model NY, USA) and then to vagal tone using Delta-Biomet-
12 Neurodata Acquisition System with filters set at 1 rics, Inc., Mxedit software, which utilizes an algorithm
Hz high-pass and 100 Hz low-pass and a gain of 20 000. developed by Porges (1991).
Before data collection, the signal for each channel was
For personal use only.
Groups
Variables Low anger High anger F P
POMS– anger 2·7 2(2·17 ) 17 ·8 8(7·7 0) 8 6·25 0·00 1
Prenatal CES-D 10·7 7(7·49 ) 19 ·8 2(8·9 6) 2 8·09 0·00 1
Postnatal CES-D 7·3 0(6·68 ) 14 ·4 8(11 ·48) 1 4·31 0·00 1
STAI– prenatal 31·3 8(8·77 ) 44 ·35 (10·16) 43·51 0 ·001
STA I– postnatal 31·1 5(8·51 ) 41 ·5 5(9·5 2) 3 0·72 0·00 1
BAS-BIS difference 0·0 5(0·75 ) 0 ·0 2(0·9 1) 0·2 1 ns
26 4 T. Field et al.
Table III. Means for prenatal and postnatal neurotransmitters and neurohormones (ng/mg) (SDs in
parentheses)
Groups
Variable Low anger High anger F P
Biochemical
Prenatal (mother)
Cortisol 2 53·4 4(16 4·05 ) 32 4·48 (178 ·13) 2·95 0·05
Noradrenaline 53·8 0(26 ·1 1) 54 ·16(24 ·38) 0·07 ns
Adrenaline 5·0 4(3·30 ) 6 ·55(4·7 4) 2·76 0·05
Dopamine 3 35·7 7(10 7·83 ) 26 1·32 (123 ·21) 3·02 0·00 5
Serotonin 49 56·2 4(25 45·3 8) 3 984 ·92(21 00 ·7 5) 2·71 0·05
Postnatal (mother)
Cortisol 1 94·1 6(92 ·2 4) 27 0·79 (211 ·32) 2·58 0·05
Noradrenaline 31·0 5(15 ·8 3) 45·92(24 ·23) 3·83 0·01
Adrenaline 4·2 7(2·90 ) 5 ·19(1·9 9) 1·37 ns
Dopamine 2 59·7 7(11 4·98 ) 25 3·47 (94·9 3) 0·23 ns
Serotonin 45 60·2 8(33 04·0 2) 4 495 ·60(21 91 ·8 6) 0·08 ns
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
Infant
Cortisol 4 41·0 5(17 7·54 ) 52 6·50 (199 ·92) 2·60 0·05
Noradrenaline 56·5 5(38 ·5 8) 63 ·47(33 ·95) 0·67 ns
Adrenaline 5·8 3(3·31 ) 4 ·91(3·0 8) 1·01 ns
Dopamine 5 00·9 0(24 2·72 ) 35 5·62 (253 ·64) 3·07 0·05
Serotonin 94 06·2 7(75 02·2 8) 6 635 ·74(53 56 ·2 6) 1·21 ns
Table IV. Means for prenatal complications and physiological variables (SDs in parentheses) (high values are
optimal except for fetal activity)
Groups
Variables Low anger High anger F P
For personal use only.
Groups
Variables Low anger High anger F P
Sleep scale
Deep sleep 25 ·74 (23 ·75) 3 6·91 (2 5·65 ) 2·11 ns
Active sleep 1 2·68 (13·60 ) 13 ·61 (1 1·63 ) 0·06 ns
REM sleep 14 ·37 (20·00) 1 2·78 (1 3·05 ) 0·10 ns
Drowsy 9 ·16 (11 ·63) 3·48 (8 ·60) 1·48 ns
Alert inactive 5 ·21 (8·44) 2·65 (5 ·01) 3·30 ns
Awake active 4·26 (7·26) 3·48 (8 ·06) 0·11 ns
Fussing/crying 4 ·00 (7·4 1) 3·13 (5 ·35) 0·19 ns
Indeterminate sleep 39 ·84 (27 ·23) 4 8·91 (1 9·58 ) 1·57 ns
Number of state changes 34 ·21 (20 ·43) 5 1·85 (2 8·73 ) 5·05 0·0 5
Postnatal total movement 0 ·61 (0·1 0) 0·65 (0 ·17) 0·93 ns
Table VI. Means for Brazelton neonatal behavior assessment scale scores (SDs in parentheses)
Groups
Variable Low High F P
Brazelton
Habituation 5·67 (1·72 ) 5·10 (2 ·18) 1 ·80 ns
Orientation 4·73(1·58 ) 3·82 (1 ·79) 6 ·19 0 ·0 5
Motor organisation 4·53 (0·95 ) 3·94 (1 ·42) 5 ·16 0 ·0 5
Range of state 3·82 (1·17 ) 3·50 (1 ·14) 1 ·64 ns
State regulation 4·18 (1·41 ) 4·06 (1 ·69) 0 ·13 ns
A utonomic stability 6 ·27(1 ·5 7) 5·7 7(2·0 7) 1·5 8 ns
Reflexes 2 ·90(2 ·4 8) 2·5 3(1·7 1) 0 ·6 8 ns
Withdrawal symptoms 2·74 (1·98 ) 3·30 (2 ·11) 1 ·59 ns
Excitability 2·15(1 ·7 0) 2·40 (1 ·56) 0 ·49 ns
Depression 2·03 (1·56 ) 4·20 (3 ·01 17·6 7 0·001
Although this has been reported repeatedly for NIMH merit award (MH no. 46586) to Tiffany Field and
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
depressed mothers and infants (Dawson et al., 1992; funding by Johnson and Johnson.
Jones et al., 1996; Field et al., 2000), this again is the
first observation of this matching of EEG patterns in
angry mothers and their infants. The relative right- References
frontal EEG activation noted in our high-anger moth- Abrams S.M., Field T., Scafidi F. and Prodromidis M. (1995)
ers and infants is consistent with greater blood flow Maternal ‘depression’ effects on infants’ Brazelton Scale
reported for the functional MRIs in the right prefrontal performance. Infant Mental Health Journal, 16, 231–235.
al’Absi M., Bongard S. and Lovallo W.R. (2000) Adrenocor-
cortex (Hariri et al., 2000).
ticotropin responses to interpersonal stress: effects of overt
The frequent report of low vagal tone in depressed anger expression style and defensiveness. International
mothers and infants was also replicated here on the Journal of Psychophysiology, 37, 257–265.
high-anger mothers and infants. Elevated stress bio- Blom J.L. and Anneveldt (1982) An electrode cap tested.
chemistry and lower vagal activity (sympathetic Electroencephalography and Clinical Neurophysiology,
For personal use only.
Field T., Diego M., Dieter J., Hernandez-Reif M., Schanberg Porges S.W. (1991) Vagal tone: a mediator of affect. In: The
S., Kuhn C., Yando R. and Bendell D. (2000) Prenatal Development of Emotion Regulation and Dysregulation.
depression effects on the fetus and the newborn. In review. Edited by Garber J. and Dodgel K.A. Cambridge,
Glover V., Teixeira J., Gitau R. and Fisk N.M. (1999) Cambridge University Press.
Mechanisms by which maternal mood in pregnancy may Pugatch D., Haskell D. and McNair D.M. (1969) Predictors
affect the fetus. Contemporary Reviews in Obstetrics and and patterns of change associated with the course of time
Gynecology, 1–6. limited psychotherapy. Mimeo Report.
Hariri A.R., Bookheimer S.Y. and Mazziotta J.C. (2000) Radloff L.S. (1977) The CES-D Scale: a self-report depres-
Modulating emotional responses: effects of a neocortical sion scale for research in the general population. Journal of
network on the limbic system. Neuroreport , 11, 43–48. Applied Psychological Measures, 1, 385–401.
Hassmen P., Koivula N. and Uutela A. (2000) Physical Radloff L.S. and Teri L. (1986) Use of the Center for Epide-
exercise and psychological well-being: a population study miological Studies Depression Scale with older infants.
in Finland. Preventive Medicine, 30, 17–25. Clinical Gerontologist, 5, 119–135.
Hollingshead A. (1975) Four-factor Index of Social Status. Scafidi F., Field T., Schanberg S., Bauer C., Vega-Lahr N.,
New Haven, CT, Yale University. Garcia R., Poirier J., Nystrom G. and Kuhn C.M. (1986)
Jones N.A., Field T., Fox N.A., Davalos M., Lundy B. and Effects of tactile/kinesthetic stimulation on the clinical
Hart S. (1998) Newborns of mothers with depressive symp- course and sleep/wake behavior of preterm neonates. Infant
toms are physiologically less developed. Infant Behavior Behavior and Development, 9, 91–105.
and Development, 21, 537–541. Sostek A.M. and Anders T.F. (1975) Effects of varying
Lester B., Als H. and Brazelton T.B. (1982) Regional obstet-
J Obstet Gynaecol Downloaded from informahealthcare.com by The University of Manchester on 12/20/14
emotional and physiological reactivity (socialization, emo- Wadhwa P.D., Porto M., Garite T.J., Chicz-DeMet A. and
tional, reactivity). Dissertation Abstracts International, Sandman C.A. (1998) Maternal corticotropin-releasin g
Section B, Sciences and Engineering, 60, 3620. hormone levels in the early third trimester predict length of
McNair D.M. and Lorr M. (1964) An analysis of mood in gestation in human pregnancy. American Journal of
neurotics. Journal of Abnormal Social Psychology, 69, Obstetrics and Gynecology, 179, 1079–1085.
620–627. Weiss J.M., Demetrikopoulos M.K., West C.H.K. and
McNair D.M., Lorr M. and Droppleman L.F. (1971) POMS Bonsall R.W. (1996) Hypothesis linking the noradrenergic
profile—Profile of Mood States. San Diego, CA, and dopaminergic systems in depression. Depression, 3,
Educational and Industrial Testing Services. 225–245.