ORI GIN AL PA PER

Religiosity is an Important Part of Coping with Griefin Pregnancy After a Traumatic Second Trimester Loss

F. S. Cowchock • S. E. Ellestad • K. G. Meador • H. G. Koenig •

E. G. Hooten • G. K. Swamy

Published online: 23 August 2011� Springer Science+Business Media, LLC 2011

Abstract Women (n = 15) who were pregnant after a traumatic late pregnancy loss

(termination because of fetal death or serious anomalies) completed psychometric

screening tests and scales, including the Perinatal Grief Scale (PGS), the Impact of Event

Scale (IES), the Duke Depression Inventory (DDI), the Generalized Anxiety Disorder-7

(GAD), and the Hoge Scale for Intrinsic Religiosity (IR). Despite a mean elapsed time

since the prior loss of 27 (range, 7–47) months, half (7/15, 47%) of the combined groups

had high levels of grief on the PGS. Multiple positive scores on psychometric tests were

frequent: Sixty percent (9/15) had high scores on the PGS Active Grief subscale or on the

IES. Forty percent (6/15) had a high score on the DDI, and 17% (3/15) on the GAD. IR

scores significantly and negatively correlated with scores on the Despair subscale of the

PGS. The results from this pilot study suggest that high levels of grief and PTS symptoms

are significant problems for pregnant women who have suffered late loss of a wanted

pregnancy. Religiosity may play an important part in maternal coping during these stressful

pregnancies.

Keywords Religiosity � Intrinsic religiosity � Pregnancy � Pregnancy termination �Grief/ving � Posttraumatic stress � Depression � Anxiety � Perinatal Grief Scale

F. S. Cowchock � H. G. Koenig � E. G. HootenCenter for Spirituality, Theology and Health, Duke University Medical Center,Durham, NC 27710, USA

F. S. Cowchock (&)9015 Bahama Woods Lane, Bahama, NC 27503, USAe-mail: fsusan.cowchock@duke.edu; fcowchock@nc.rr.com

S. E. Ellestad � G. K. SwamyDepartment Obstet Gynecol, Duke University Medical Center, Durham, NC 27710, USA

K. G. MeadorAdult Psychiatry, Vanderbilt Psychiatric Hospital, South Nashville, TN 37212, USA

123

J Relig Health (2011) 50:901–910DOI 10.1007/s10943-011-9528-y

Introduction

Women who become pregnant after a traumatic pregnancy loss at any stage are reported to

view gestation as a significant threat and to have more symptoms of pregnancy-related

anxiety and depression (Barr 2006; Bergner et al. 2008; Cote-Arsenault 2007; Hughes et al.

1999; Hunfeld et al. 1996; Tsartsara and Johnson 2006). Women pregnant again after

elective or spontaneous abortion, or after stillbirth (defined in the study cited here as loss

after 18 weeks gestation), are reported to have an increased risk of posttraumatic stress

symptoms (12.6–21%) and depression (16.8–28%) (Hughes et al. 1999; Forray et al. 2009;

Hamama et al. 2010). (Only one study found no increased risk of psychological distress

during pregnancy after a traumatic loss (Scheidt et al. 2008).)

Termination of a wanted pregnancy because of fetal anomaly has been reported to be

associated with the same intensity of grief and depression as a spontaneous loss at the same

gestational age (Salverson et al. 1997; Zeanah et al. 1993). The method of termination in

the second trimester, whether dilation and evacuation or labor induction, did not seem to

affect immediate intensity or resolution of grief and depression at 12-month follow-up

(Burgoine et al. 2005). Termination later in pregnancy is likely to be associated with

greater psychological morbidity: In one study, women with second trimester terminations

for fetal anomalies had a significantly higher level of posttraumatic stress symptoms

6 weeks after termination than a comparison group terminating in the first trimester

(Davies et al. 2005).

Religious beliefs are significant influences on responses to bereavement in general and

thus are expected to influence grief or psychological distress related to pregnancy loss. A

systematic review demonstrated that the vast majority of studies (94%) have demonstrated

a positive effect of religious or spiritual beliefs on adjustment to bereavement in general,

but the heterogeneity of the studies precluded any generalized conclusions (Becker et al.

2007).

Most studies of the effects of religious beliefs on bereavement due to pregnancy loss

have evaluated single parameters such as religious attendance, use of prayer, strength of

religious faith, or self-rated religiosity: An Australian study of parents after a perinatal

death (ranging from stillbirth to sudden infant death/SIDS) found that both these bereaved

parents and normal controls had lower rates of anxiety and depression when they were

weekly church attendees, but the bereaved were not more likely than controls to attend

services frequently (Thearle et al. 1995). Similarly, using a scale developed to measure

spiritual well-being in an American population, Dunn and Shelton found that spiritual well-

being was inversely associated with anxiety and depression in all groups evaluated. These

groups included non-pregnant controls, normal pregnant women, and pregnant women on

bed rest because of obstetric complications (Dunn and Shelton 2007). Mann et al. (2008)

found that religious attendance and self-rated spirituality were significantly and inversely

related to grief, but not depression, following pregnancy loss. In a second study of the

general obstetric population, they did find that religious attendance predicted a lower risk

of postpartum depression (Mann et al. 2008). Kersting et al. (2007) reported that strength

of religious faith and social support were the predictors of lower grief scores on the

Perinatal Grief Scale.

Few reports have evaluated the potential role of specific religious beliefs or practices,

other than prayer, in coping with psychological distress related to pregnancy complications

or loss. Our first study, based on data from the Lehigh Valley Perinatal Loss Project (the

basis for the Perinatal Grief Scale), did evaluate some specific religious beliefs and practices.

No association was found in that study between religious service attendance, agreement with

902 J Relig Health (2011) 50:901–910

123

statements from the intrinsic and extrinsic religiosity scales, positive religious coping, and

grief scores during the 2-year follow-up, after a range of pregnancy losses (Cowchock et al.

2010). Religious struggle, agreement with statements classified as negative religious coping,

and continued attachment to the baby were associated with more severe grief.

Although pregnancy following a perinatal loss has been established as a period of high

risk of psychological distress (Bergner et al. 2008; Cote-Arsenault 2007; Hughes et al.

1999; Tsartsara and Johnson 2006; Forray et al. 2009; Hamama et al. 2010; Scheidt et al.

2008; Rillstone and Hutchinson 2001; Turton et al. 2001), the potential role of persistent or

reemerging grieving for the lost infant in this distress has not been evaluated. Adequate or

conclusive studies of the role of religious or spiritual beliefs or practices in modulating this

distress during pregnancy are lacking: In a study of coping with stress during pregnancy in

a general obstetric population, spiritual coping (questions on the use of prayer and bible-

reading) was used most frequently (Hamilton and Lobel 2008). Greater religiosity and

optimism predicted the use of spiritual coping in that population. Another study of women

pregnant after a prior spontaneous or elective abortion measured the use of prayer to cope

with difficult emotions. In contrast to the first observations from a general obstetric pop-

ulation, in this group at (presumably) increased risk, there was no significant association

between use of prayer and either PTSD or depression during pregnancy (Hamama et al.

2010).

From the studies reviewed above, pregnancies occurring after a late (second trimester)

loss of a wanted pregnancy are at the highest risk of psychological distress and grief. We

wanted to evaluate the frequency of such distress, including depression, anxiety, and PTS

symptoms, in a pilot study of this potentially high-risk group. We hypothesized that grief

for the lost infant would play an important role in the distress suffered by these women

during pregnancy. We also expected that women terminating second trimester pregnancies

because of fetal anomalies would not be different on psychological distress screening

measures compared to women terminating pregnancies in the same period because of

spontaneous fetal death. Our second hypothesis was that psychological distress, including

grieving, would be inversely and significantly related to religiosity and religious attendance

in a group of women at very high risk. We investigated the role of potential predictors such

as maternal age, the interval since the traumatic loss, intrinsic religiosity (as measured by

the Hoge Scale), and history of other pregnancy losses or live-born children as modifiers of

these painful symptoms. We planned to evaluate religious congregational support as well,

but could not because only 4 women in the study (27%) were members of a congregation.

Methods

This study was approved by the Institutional Review Board of Duke University Medical

Center. Women seeking prenatal care at Duke University Medical Center who spoke

English, and had terminated a prior pregnancy after the first trimester, were eligible for

recruitment. In order to minimize any possible psychological trauma associated with study

participation, recruitment did not take place until after all relevant prenatal diagnostic tests

were completed, and normal and each of the women had passed the gestational week of the

prior loss. Eligible women were recruited for study participation by their obstetric care

providers. Twenty eligible women were approached for participation; 15 (75%) partici-

pated. It is possible that those declining participation were more severely affected by

pregnancy loss than the others, but there were no explanatory differences on demographic

or obstetric history between those that refused the study and participants. Statistics were

J Relig Health (2011) 50:901–910 903

123

calculated using SAS v9.1, SAS Institute, Inc., 2006, Cary NC and Microsoft EXCEL,

v2007. Statistical significance was defined as a two-tailed p value less than 0.05. Bor-

derline significance was defined as a two-tailed value[0.05 and\0.15. All P values were

adjusted for multiple comparisons.

Informed consent was obtained and demographic and psychological questionnaires were

administered by the first author. Established scales used in this study included the Hoge

Scale for Intrinsic Religiosity(IR) (Hoge 1972), the Brief Perinatal Grief Scale (PGS)

(Potvin et al. 1989; Toedter et al. 2001), the Impact of Event Scale (IES) for symptoms of

posttraumatic stress disorder (PTSD) (Horowitz et al. 1979; Horowitz 1982; Joseph 2000;

Witteveen et al. 2005), the Duke Depression Inventory (DDI) (Koenig et al. 1995, 1997),

and the GAD-7 screen for generalized anxiety disorder (GAD) (Lowe et al. 2008; Spitzer

et al. 2005). Questions from the Hoge Scale were used in a previously reported study

evaluating women after perinatal losses and were validated for use in that group with a

Cronbach’s alpha of 0.82 (Cowchock et al. 2010). The PGS has been used for more than

two decades to evaluate grief associated with pregnancy loss (Potvin et al. 1989; Toedter

et al. 2001). The PGS was demonstrated to be reliable and valid in a group of pregnant

women similar to those in our study: In that series, a severe or lethal fetal malformation

was diagnosed at 24 weeks gestation or later (Hunfeld et al. 1993). Scores from bereaved

men and women on the PGS fall between 78 and 91, 95% of the time. Thus, scores of 91 or

greater represent a high degree (greater than the 95 percentile) of grief (Toedter et al.

2001). Increasing maternal age has been associated with lower PGS scores after pregnancy

loss (Mann et al. 2008; Toedter et al. 1988).

The IES has been used for more than 20 years to screen for symptoms of PTSD and has

been validated in varied populations, including subjects with a history of pregnancy loss

(Cronbach’s alpha [ 0.80) (Horowitz et al. 1979; Joseph 2000; Kelly et al. 2010; Sundin

and Horowitz 2002). The original IES does not screen for hypervigilance and has been

criticized on that ground as a diagnostic tool for PTSD (Joseph 2000). However, the

intrusion and avoidance subscales have shown at least ‘‘moderate’’ correlation with other

diagnostic tests for PTSD, and r values ranging from 0.32 to 0.80. The IES also has

convergent validity with observer-diagnosed PTSD (Sundin and Horowitz 2002). Horo-

witz’ defined a total score on the IES of C19 as a ‘‘high’’ symptom level (Horowitz et al.

1979). Neal et al. (1994) found that a cutoff score for the total IES of 35 produced the

highest positive predictive value (0.88) for PTSD, with the lowest misclassification error

rate. We evaluated our data with respect to both of these cutoff levels. As in similar studies,

our subjects were instructed to apply statements from the IES to their pregnancy loss.

The GAD-7 screen for generalized anxiety disorder is a part of the Patient Health

Questionnaire and has been validated in more than 10,000 primary care outpatients,

including a study of 3,000 women patients from obstetric and gynecology practices

(Hunfeld et al. 1993; Kroenke et al. 2010). Cronbach’s alpha in the original study was 0.92.

A cutoff score of 10 had 89% sensitivity and 85% specificity for diagnosis of generalized

anxiety disorder (Spitzer et al. 2005). The Duke Brief Depression Scale has been validated

for use in medically ill patients, and scores of 3 or higher were 83% sensitive and 77%

specific for major depression (Koenig et al. 1995; Hunfeld et al. 1993).

Results

The mean gestational week for prior losses was 19.3 weeks (range, 13–23), and the women

were interviewed at a mean of 21.9 weeks (range 15–29) of the current pregnancy. These

904 J Relig Health (2011) 50:901–910

123

dates were in concordance with the recruitment restrictions outlined above. Seven women

were Caucasian and the other 8 were African-American. The mean age was 34.2 years

(range 26–38). Prior traumatic losses included 7 terminations for diagnosed fetal anomalies

and 8 for fetal demise. The mean elapsed time since the traumatic loss was 27 months

(range 7–47). The number of living children for the entire group ranged from 0 to 4, and

the number of other perinatal losses ranged from 0 to 2 (all first trimester). All have

delivered normal babies, but two (1 in the fetal anomaly group and 1 in the fetal death

group) were diagnosed with severe postpartum depression requiring psychiatric treatment.

Only one of the women scored high for depression on the DDI, but both had positive

screens on all the other three tests (PGS, IES, and GAD).

Table 1 Demographic and psychometric scale data for fetal anomaly (FA), fetal death (FD), and combinedgroups

FA group FD group Combined

Maternal age at delivery (mean) 38 years 32 years 35 years

One or more living children 43% (3/7) 88% (7/8) 67% (10/15)

Other pregnancy losses 43% (3/7) 63% (5/8) 53% (8/15)

PGS* mean 75.6 76.4 75.9

No. ‘‘grieving’’ 43% (3/7) 50% (4/7) 47% (7/15)

No. ‘‘high’’ 43% (3/7) 13% (1/8) 27% (5/15)

Subscales

Active grief mean 35 38 36.3

No. ‘‘high’’ 57% (4/7) 63% (5/8) 60% (9/15)

Difficulty coping mean 17.3 17.9 17.5

No. ‘‘high’’ 0% 13% (1/8) 7% (1/15)

Despair mean 23.6 17.3 20.2

No. ‘‘high’’ 43% (3/7) 13% (1/8) 27% (4/15)

IES** mean 22.9 28.2 25.9

No. positive 57% (4/7) 63% (5/8) 60% (9/15)

High levels § 27% (4/15)

Subscales

Intrusion mean 12.1 15.4 13.9

Avoidance mean 10.7 12.8 11.9

DDI*** mean 2.6 2.6 2.6

No. positive*** 43% (3/7) 38% (3/8) 40% (6/15)

GAD**** mean 6.1 6.5 6.3

No. positive**** 29% (2/7) 25% (2/8) 27% (4/15)

IR***** mean (range) 28 (19–36) 34 (14–48) 31 (14–48)

* Perinatal grief score––‘‘grieving’’ 78–91, ‘‘high’’[95th percentile = r [ 91, cutoff for subscales variable

** Impact of event scale—cutoff 19, no cutoff for subscales

*** Duke Depression Inventory—cutoff 3

**** Generalized Anxiety Disorder-7—cutoff 10

***** IR (Intrinsic religiosity/Hoge Scale)––no cutoff. Bold numbers indicate mean values above the cutofflevel for that scale

§ Cutoff of [35

J Relig Health (2011) 50:901–910 905

123

Table 1 presents a comparison of the women whose traumatic loss was due to termi-

nation for fetal anomaly (n = 7, the FA Group) to those whose loss was due to fetal death

(n = 8, the FD Group), as well as results from the combined groups. As expected, maternal

age was higher in the FA group, but the difference in proportion of those C35 years at

delivery between groups was not statistically significant (P = 0.27, Fisher’s exact test).

There were no apparent differences on scores between the FA and FD groups on the

psychometric measures administered in the study. The mean scores on the IES and the

Active Grief subscale of the PGS were above high cutoff levels in both groups.

Results from the psychosomatic measures in the combined groups indicated a high level

of psychological distress among the women. We found high scores on the IES (60%, 95%

CI = 36–80%), and on the Active Grief subscale of the PGS (60%, 95% CI = 36–80%),

followed by those in the grieving range on the total PGS (47%, 95% CI = 25–70%),

positive screens for depression on the DDI (40%, 95% CI = 25–70%), and lastly, high

scores for anxiety on the GAD (27%, 95% CI = 15–58%). Multiple positive screening

tests were frequent, most often high scores on both the Active Grief PGS subscale and on

the IES (53%, 8/15, 95% CI = 30–75%)).

Table 2 shows Pearson’s r values for correlations between the psychometric scales used

to measure grief and distress among these women. It is apparent that results from all the

Table 2 Correlations between scores on psychometric scales used

Perinatal GriefScale

Impact of EventScale

Duke DepressionInventory

Generalized AnxietyDisorder-7

Perinatal GriefScale

XXXXXXXXXX r = 0.81,P \ 0.001

r = 0.49,P = 0.05

r = 0.71, P = 0.003

Impact of EventScale

XXXXXXXXXX XXXXXXXXXX r = 0.55,P = 0.04

r = 0.79, P = \0.001

Duke DepressionInventory

XXXXXXXXXX XXXXXXXXXX XXXXXXXXXX r = 0.62, P = 0.01

Table 3 Correlations (Pearson’s r values and statistical significance) between possible predictors andscores on psychometric scales for combined groups

Maternal age Monthssince loss

Livingchildren

Other perinatallosses

Intrinsic religiosityscores

PGS* total -0.3 (n.s.) -0.06 (n.s.) 0.33 (n.s.) -0.11 (n.s.) -0.5 (P = 0.05)*

Subscale

Active grief -0.44 (P = 0.10) 0.02 (n.s.) 0.22 (n.s.) -0.17 (n.s.) -0.33 (n.s.)

Difficultycoping

-0.40 (P = 0.12) 0.12 (n.s.) 0.28 (n.s.) -0.02 (n.s.) -0.31 (n.s.)

Despair 0.10 (n.s.) -0.33 (n.s.) 0.37 (n.s.) -0.23 (n.s.) 20.73 (P = 0.002)

IES* total -0.15 (n.s.) 0.2 (n.s.) 0.3 (n.s.) 0.02 (n.s.) -0.12 (n.s.)

Subscale

Intrusion -0.28 (n.s.) 0.01 (n.s.) 0.25 (n.s.) -0.14 (n.s.) -0.2 (n.s.)

Avoidance -0.31 (n.s.) 0.14 (n.s.) 0.39 (n.s.) 0.15 (n.s.) -0.08 (n.s.)

DDI -0.36 (n.s.) 0.01 (n.s.) 0.4 (P = 0.12) -0.14 (n.s.) 0.01 (n.s.)

GAD 20.59 (P = 0.02)* 0.2 (n.s.) 0.05 (n.s.) -0.23 (n.s.) 0.06 (n.s.)

* Borderline/not significant—significance level P B 0.01 after Bonferonni correction for multiplecomparisons

906 J Relig Health (2011) 50:901–910

123

scales are significantly correlated. The correlation is highest between the PGS and the IES

(r = 0.81, P \ 0.001) or between the GAD and the IES (r = 0.79, P \ 0.001). Correlation

with the two other scales is least with the DDI, but still significant.

Table 3 displays correlations between the potential demographic predictors (Table 1)

and Intrinsic Religiosity scores (IR) on the Hoge Scale, with results from the psychometric

measures. The correlation between maternal age at delivery and GAD anxiety scores was

borderline significant after Bonferonni correction for multiple comparisons (n = 5, cor-

rected significance level 0.01). Using this corrected level, the negative correlations

between PGS subscales of Active Grief and Difficulty Coping and maternal age and the

positive correlation between number of living children and DDI scores were not signifi-

cant. All correlations between IR scores and PGS scores were negative, but the negative

correlation was only significant after Bongiovanni correction for the Despair subscale.

There were no important or statistically significant correlations between the months

elapsed since the loss and results from any of these psychometric measures, despite a broad

range of 7–47 months. There also was no important correlation with number of other

perinatal losses, possibly because these others were all first trimester losses.

Discussion

We expected that the next pregnancy after a traumatic second trimester loss would be a

period of very high risk of psychological distress, ranging from recurrent grief to symp-

toms of posttraumatic stress, depression, or anxiety. We found that 60% (36–80% CI) of

the women we interviewed had symptoms levels indicating substantial grief on the Active

Grief subscale of the PGS (greater than the 95th percentile of pregnancy loss populations).

Sixty percent of our participants also had high levels of posttraumatic stress symptoms on

the IES. More than half scored high on more than one psychometric measure, most often

(53, 30–75% CI) a combination of the Active Grief subscale of the PGS and a positive

screen on the IES. High scores for anxiety (GAD) or depression (DDI) were less frequent,

although two of these women (one in the FD group and one in the FA group) suffered

clinically severe postpartum depression. One of the two had a positive DDI score, but both

had positive scores on all the other 3 psychometric screening tests. As we had expected,

women who terminated a prior second trimester pregnancy because of fetal death did not

differ in these respects from those who terminated because of serious fetal anomalies. The

small numbers in our study, however, preclude any definite conclusion on this point. These

results confirm our first hypothesis that women pregnant after such a traumatic loss often

suffer from high levels of grief, in addition to other forms of psychological distress. These

high levels of distress were not related to the length of time since that loss. This was true

even though more than 2 years, on average, had passed (mean, 27; range, 7–47 weeks).

These preliminary findings suggest that such grief in pregnancy might be reemergent rather

than persisting since the loss. This possibility needs to be evaluated in future studies.

We can compare the level of psychological distress in this group with other reports,

arranged by approximate gestation of the loss: In a large group of women who had a prior

elective or spontaneous abortion, the rate of PTSD was 12.6% (assessed by a telephone

interview) (Hamama et al. 2010). The frequency of depression was 12.6% in that study and

5.4% met criteria for both disorders. The frequency of ‘‘partial PTSD’’ (assessed by patient

completion of the Clinician Administered PTSD Scale) in a group of 56 pregnant women,

most of whom had suffered miscarriage of a prior pregnancy, was 28.6% (Forray et al.

2009). In a group at theoretical highest risk, whose prior pregnancy ended in ‘‘stillbirth’’

J Relig Health (2011) 50:901–910 907

123

after 18 weeks gestation, 21% had case-level symptoms of PTSD (assessed by PTSD-I

interview) (Turton et al. 2001). A strong relationship between PTSD symptoms, depres-

sion, and anxiety was also observed (Hughes et al. 1999). All these studies used different

criteria and assessment methods for PTSD symptoms, but the levels of symptoms in our

group are comparable to those reported, even when using Neal’s higher cutoff level for the

IES (27% were still elevated) (Neal et al. 1994).

Unique to our study is an evaluation of grief during a next normal pregnancy using the

PGS. One other study used the PGS to evaluate women in late pregnancy who were

recently informed of the diagnosis of a severe or lethal fetal malformation affecting that

pregnancy (Hunfeld et al. 1993). That study also noted high levels of PTSD symptoms and

a significant correlation between scores on the PGS and IES Scales.

This pilot study is also the first to use an established scale to evaluate intrinsic religi-

osity (IR), rather than single surrogate parameters, in order to evaluate any correlation with

psychometric test scores. In our previous report based on results from a lower-risk group of

women who had suffered perinatal loss at varied periods of gestation (primarily first

trimester miscarriage), and who were not then pregnant, no significant correlation between

statements from the Hoge IR Scale and PGS scores, including the Despair subscale, was

found.

Our second hypothesis was that any influence from religious factors would be most

apparent in pregnant women at highest risk, and we did find a significant negative cor-

relation between IR scores and scores on the PGS Despair subscale. Our observation of this

correlation only with Despair scores, even though high scores on the Active Grief subscale

were more frequent in our population, suggests that these two subscales are indeed mea-

suring different aspects of perinatal grief.

The significant correlations between all of the psychometric screening tests reported

here make it difficult to interpret multiple high scores on such tests as evidence of coex-

istent disorders. We are not the first to make this observation: Hunfeld et al. (1993) also

reported significant correlations between the PGS Scales and subscales and the PEL (IES)

Scale and subscales. Turton and others also noted highly significant relationships between

the PTSD category and Edinburgh depression scores and between the state and trait scores

on anxiety screening. These observations may simply demonstrate that depression and

posttraumatic stress symptoms are both part of grieving. Indeed some of the statements on

these scales are very similar. This degree of correlation, or colinearity, does mean that

results from such screening questionnaires could not be used as predictor variables in a

multiple regression analysis.

In conclusion, the results of this pilot study call attention to the high levels of psy-

chological distress in a next pregnancy after late loss of a wanted pregnancy, including

termination for fetal anomalies. Religiosity and religious beliefs may play an important

part in maternal coping in these stressful pregnancies. Future studies of pregnant women

who have suffered prior losses should include screening for abnormal levels of grief, as

well as symptoms of PTS, depression, and anxiety. Active grieving and PTS symptoms

may be more significant problems for this group of women than the postpartum depression

that is now a part of routine obstetric screening. Confirmation of our findings should lead to

proposals for effective support, including appropriate religious and spiritual modalities, of

pregnant women at high risk for such suffering.

Acknowledgments This work was supported in part by a grant from the John Templeton Foundation to theCenter for Spirituality, Theology and Health of Duke University Medical Center. Thanks are due to AmandaAnderson, RN of the Maternal Fetal Medicine Division for identification of patients eligible for the study

908 J Relig Health (2011) 50:901–910

123

and to Kathleen Peindl, PhD from the Dept. of Psychiatry and Human Behavior for editorial assistance—bothfrom Duke University Medical Center.

References

Barr, P. (2006). Relation between grief and subsequent pregnancy status 13 months after perinatalbereavement. Journal of Perinatal Medicine, 34(3), 207–211.

Becker, G., Xander, C. J., Blum, H. E., Lutterbach, J., Momm, F., Gysels, M., et al. (2007). Do religious orspiritual beliefs influence bereavement? A systematic review. Palliative Medicine, 21, 207–217.

Bergner, A., Beyer, R., Klapp, B. F., Reed, C. D., & Rauchfuss, M. (2008). Pregnancy after early pregnancyloss: A prospective study of anxiety, depressive symptomatology and coping. Journal of Psychoso-matics Obstetrics and Gynaecology, 29(2), 105–113.

Burgoine, G. A., Van Kirk, S. D., Romm, J., Edelman, A. B., Jacobson, S. L., & Jenson, J. T. (2005).Comparison of perinatal grief after dilation and evacuation or labor induction in second trimesterterminations for fetal anomalies. American Journal of Obstetrics and Gynecology, 194(6), 1743.

Cote-Arsenault, D. (2007). Threat appraisal, coping and emotions across pregnancy subsequent to perinatalloss. Nursing, 56(2), 108–116.

Cowchock, F. S., Lasker, J. N., Toedter, L. J., Skumanich, S. A., & Koenig, H. G. (2010). Religious beliefsaffect grieving after pregnancy loss. Journal of Religion and Health, 49(4), 485–497.

Davies, V., Gledhill, J., McFadyen, A., Whitlow, B., & Economides, D. (2005). Psychological outcome inwomen undergoing TOP for ultrasound-detected fetal anomaly in the first and second trimesters: Apilot study. Ultrasound in Obstetrics and Gynecology 25, 380–392.

Dunn, L. L., & Shelton, M. M. (2007). Spiritual well-being, anxiety and depression in antepartal women onbedrest. Issues in Mental Health Nursing, 28(11), 1235–1246.

Forray, A., Mayes, L. C., Magriples, U., & Epperson, C. N. (2009). Prevalence of post-traumatic stressdisorder in pregnant women with prior pregnancy complications. Journal of Maternal Fetal NeonatalMedicine, 22(6), 522–527.

Hamama, L., Rauch, S. A., Sperlich, M., Defever, E., & Seng, J. S. (2010). Previous experience of spon-taneous or elective abortion and risk for posttraumatic stress and depression during subsequentpregnancy. Depress Anxiety, 27(8), 699–707.

Hamilton, J., & Lobel, M. (2008). Types, patterns, and predictors of coping with stress during pregnancy:Examination of the revised prenatal coping inventory in a diverse sample. Journal of PsychosomaticObstetrics and Gynaecology, 29(2), 97–104.

Hoge, D. R. (1972). A validated intrinsic religious motivation scale. Journal of Scientific Study Religion, 11,369–376.

Horowitz, M. (1982). Stress response syndromes and their treatment. In L. Goldberger & S. Breznitz (Eds.),Handbook of stress: Theoretical and clinical aspects (pp. 711–732). New York, NY: Free Press.

Horowitz, M., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of subjective stress.Psychosomatic Medicine, 41(3), 209–218.

Hughes, P. M., Turton, P., & Evans, C. D. H. (1999). Stillbirth as risk factor for depression and anxiety inthe subsequent pregnancy: Cohort study. BMJ, 318, 1721–1724.

Hunfeld, J. A. M., Agterberg, Wladimiroff, J. W., & Passchier, J. (1996). Quality of life and anxiety inpregnancies after late pregnancy loss: A case-control study. Prenatal Diagnosis, 16, 783–790.

Hunfeld, J. A. M., Wladimiroff, J. W., Passchier, J., Venema-van Uden, M. U., Frets, P. G., & Verhage, F.(1993). Reliability and validity of the perinatal grief scale for women who experienced late pregnancyloss. British Journal of Medical Psychology, 66(3), 295–298.

Joseph, S. (2000). Psychometric evaluation of Horowitz’s impact of event scale: A review. Journal ofTraumatic Stress, 13(1), 101–113.

Kelly, T., Suddes, J., Howell, D., Hewison, J., & Robson, S. (2010). Comparing medical versus surgicaltermination of pregnancy at 13–20 weeks of gestation: A randomised controlled trial. BJOG, 117(12),1515–1520.

Kersting, A., Kroker, K., Steinhard, J., Ludorff, K., Wesselmann, U., Orhmann, P., et al. (2007). Compli-cated grief after traumatic loss: A 14 month follow up. European Archives Psychiatry and ClinicalNeuroscience, 257(8), 437–443.

Koenig, H. G., Blumenthal, J., & Moore, K. (1995). New version of brief depression scale. Journal ofAmerican Geriatrics Society, 43(12), 1447.

Koenig, H. G., George, L. K., & Peterson, B. L. (1997). Religiosity and remission from depression inmedically ill older patients. American Journal of Psychiatry, 155, 536–542.

J Relig Health (2011) 50:901–910 909

123

Kroenke, K., Spitzer, R. L., Williams, J. B. W., & Lowe, B. (2010). The patient health questionnaire somaticanxiety, and depressive symptom scales: A systematic review. General Hospital Psychiatry, 32(4),1–21.

Lowe, B., Decker, O., Muller, S., Brahler, E., Schellberg, D., Herzog, W., et al. (2008). Validation andstandardization of the generalized anxiety disorder screener (GAD-7) in the general population.Medical Care, 46(3), 266–274.

Mann, J. R., McKeown, R. E., Bacon, J., Vesselinov, R., & Bush, F. (2008). Predicting depressive symptomsand grief after pregnancy loss. Journal of Psychosomatic Obstetrics and Gynaecology 29, 1–6.

Mann, J. R., McKeown, R. E., Bacon, J., Vesselinov, R., & Bush, F. (2008). Do antenatal religious andspiritual factors impact the risk of postpartum depressive symptoms? Journal of Women’s Health,17(5), 745–755.

Neal, L. A., Busuttil, W., Rollins, J., Herepath, R., Strike, P., & Turnbull, G. (1994). Convergent validity ofmeasures of post-traumatic stress disorder in a mixed military and civilian population. Journal ofTraumatic Stress, 7, 447–455.

Potvin, L., Lasker, J., & Toedter, L. (1989). Measuring grief: A short version of the perinatal grief scale.Journal of Psychopathology and Behavioral Assessment, 11(1), 29–42.

Rillstone, P., & Hutchinson, S. A. (2001). Managing the reemergence of anguish: Pregnancy after a loss dueto anomalies. Journal of Obstetrics and Gynecology Neonatal Nursing, 30(3), 291–298.

Salverson, K. A., Oyen, L., Schmidt, N., Malt, U. F., & Eik-Nes, S. H. (1997). Comparison of long-termpsychological responses of women after pregnancy termination due to fetal anomalies and afterperinatal loss. Ultrasound in Obstetrics Gynecology, 9, 80–85.

Scheidt, C. E., Kunze, M., Wangler, J., Klar, M., Walter, N., & Hasenburg, A. (2008). Psychologicalconsequences of perinatal loss in subsequent pregnancies-a comparative study. Psychotherapie Psy-chosomatic Medicine Psychology, 58(12), 475–478.

Spitzer, R., Kroenke, K., Williams, J. B. W., & Lowe, B. (2005). A brief measure for assessing generalizedanxiety disorder. Archives of Internal Medicine, 166(10), 1092–1097.

Sundin, E. C., & Horowitz, M. J. (2002). Impact of event scale: Psychometric properties. The British Journalof Psychiatry, 180, 205–209.

Thearle, M. J., Vance, J. C., & Najman, J. M. (1995). Church attendance, religious affiliation and parentalresponses to sudden infant death, neonatal death and stillbirth. Omega, 31(1), 51–58.

Toedter, L. J., Lasker, J. N., & Alhadeff, J. M. (1988). The perinatal grief scale: Development and initialvalidation. American Journal of Orthopsychiatry, 58, 435–449.

Toedter, L. J., Lasker, J. N., & Janssen, H. J. E. M. (2001). International comparison of studies using theperinatal grief scale: A decade of research on pregnancy loss. Death Studies, 25, 225–228.

Tsartsara, E., & Johnson, M. P. (2006). The impact of miscarriage on women’s pregnancy-specific anxietyand feelings of prenatal maternal-fetal attachment during the course of a subsequent pregnancy: Anexploratory follow-up study. Journal of Psychosomatic Obstetrics and Gynaecology, 27(3), 173–182.

Turton, P., Huges, P., Evans, C. D. H., & Fainman, D. (2001). Incidence, correlates and predictors of post-traumatic stress disorder in the pregnancy after stillbirth. British Journal of Psychiatry, 178, 556–560.

Witteveen, A. B., Bramsen, I., Hovens, J. E., & van der Ploeg, H. M. (2005). Utility of the impact of eventscale in screening for posttraumatic stress disorder. Psychological Report, 97(1), 297–308.

Zeanah, C. H., Dailey, J. V., Rosenblatt, M., & Saller, D. N. (1993). Do women grieve after terminatingpregnancies because of fetal anomalies? A controlled investigation. Obstetrics and Gynecology, 82,270–275.

910 J Relig Health (2011) 50:901–910

123