*Correspondence address. Recurrent Pregnancy Loss Unit, Fertility Clinic 4071, University Hospital Copenhagen, Rigshospitalet,
Blegdamsvej 9, 2100 Kbenhavn , Copenhagen, Denmark. Tel: +45-3545-4951. E-mail: astrid.marie.kolte@regionh.dk
Submitted on August 19, 2014; resubmitted on November 5, 2014; accepted on January 13, 2015
study question: Is the prevalence of psychological stress and moderate/severe depression higher for women with recurrent pregnancy
loss (RPL) than pregnancy planners trying to conceive naturally?
summary answer: Both psychological stress and major depression are signicantly more common among women with RPL than in those
trying to conceive naturally.
what is known already: RPL has a signicant emotional impact on couples, especially the woman. Previous studies have shown in-
conclusive results.
study design, size, duration: In this cross-sectional study, we compared the prevalence of stress and depression among 301
women with RPL and 1813 women attempting to conceive naturally. We dened RPL as three or more pregnancy losses before 12 weeks ges-
tation. RPL patients were enrolled from 2010 to 2013 and the comparison group from 2011 to 2014.
participants/materials, setting, methods: RPL patients completed an online questionnaire before their rst consultation
at the Danish RPL Unit. In addition, we included data from a comparison group of 1813 women who participated in the Soon Parents Study (www.
SnartForldre.dk). The Major Depression Index (MDI) was used to assess symptoms of depression, and Cohens Perceived Stress Scale (PSS) was
used to measure stress. Relevant demographic data were also retrieved.
main results and the role of chance: Of the RPL patients, 26 (8.6%) had a score on the MDI corresponding to moderate/
severe depression, as did 40 (2.2%) of the women in Soon Parents Study (adjusted odds ratio (OR) 5.53 (95% condence interval (CI): 2.09;
14.61)). A high stress level, dened as 19 on the PSS scale, was reported by 124 (41.2%) of the patients and 420 (23.2%) in the comparison
group (adjusted OR 1.59 (95% CI 1.03; 2.44)).
limitations, reasons for caution: We used online questionnaires, and have no interview data. We were unaware if any of the
women in the comparison group suffer from RPL.
wider implications of the findings: This study should entail a heightened awareness of mental distress among care providers for
women with RPL.
study funding/competing interest(s): No specic funding was sought for this study. The Soon Parents Study is funded by
National Institute of Child Health and Human Development (R01 HD060680-01A4). No authors have competing interests to declare.
trial registration number: N/A.
Key words: recurrent pregnancy loss / depression / stress / cross-sectional study
& The Author 2015. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
RPL patients gave information on number of pregnancy losses (early preg- income, number of prior pregnancies and number of live born children), we
nancy losses, conrmed ectopic pregnancies, late miscarriages and stillbirths) chose to keep all covariates in the model. We tested for interaction between
before referral; number of months (,3; 3 6; or .6) since their last preg- all variables. All interaction terms were insignicant and were excluded.
nancy loss; perception of own health in the last year; present medication; We explored a possible correlation between scores on the PSS scale and
daily smoking (non-smokers and smokers); alcohol intake (none; 1 7; the MDI by Spearmans correlation coefcient test.
8 14; and .14 units) and country of origin. All statistical analyses were performed using Statistical Package for Social
Sciences version 19 (SPSS, USA). A P-value ,0.05 was considered signicant.
Statistical analysis
As age, number of previous pregnancies and number of live born children Results
were not normally distributed in all groups (participating RPL patients, non-
participating RPL patients and Soon Parents Study participants), we used Demographics
Mann Whitney U-test. Education and annual household income were com-
pared by the x 2 test for trend. As shown in Table I, patients were signicantly older and had signicantly
We analysed results from the MDI in two ways. Firstly, we compared the higher education levels and household income compared with partici-
median scores using Mann Whitney U-test. Secondly, we graded the results pants in the Soon Parents Study. The number of pregnancies in the
according to the ICD-10 categories as described (Olsen et al., 2003). We patient group was also higher. The women in the Soon Parents Study
then dichotomized the results on the MDI scale into no/ mild depression comparison group had had slightly more live born children (Table I).
and moderate/severe depression for logistic modelling. We tested a pos- Median number of pregnancy losses among the patients was 4 (Range:
sible correlation between time passed since last pregnancy loss in the RPL 3; 12) and 44.4% of the patients had experienced their last pregnancy loss
group by the x 2 test for trend. within the last 6 months. The patients were mostly non-smokers, and all
Absolute scores on the PSS in the two groups were compared using Stu- reported a weekly alcohol consumption of 14 units per week (Table II).
dents t-test, presented as mean scores and SD. For the logistic regression
Use of antidepressant medication was reported by four patients, of
analyses, we chose a score of 19 to denote high stress, as 18 was the
which one had an MDI score signifying moderate depression and one a
75 percentile in the comparison group.
The prevalence (absolute numbers and percentages) of moderate/severe score corresponding to severe depression. Age, number of live born chil-
depression and high stress in the two groups was compared using multiple dren and pregnancy losses as well as smoking status did not differ signi-
logistic regression analysis with results presented as adjusted OR and 95% cantly between the participating and non-participating RPL patients
condence intervals (95% CI) with the comparison group as the indicator (Supplementary Table SI). Thirty-seven (27%) of the non-respondents
group. As we only had ve demographic variables (age, education, household did not have Denmark as their country of origin versus 20 (7%) of the
Table I Demographic data on recurrent pregnancy loss patients (RPL) and comparison group.
Table IV Prevalence of moderate/severe depression and high stress among RPL patients (n 5 301) and comparison group
(n 5 1813).
RPL patients, n (%) Comparison group, n (%) Unadjusted analysis, Adjusted analysisb,
odds ratio (95% CIa) odds ratio (95% CI)
.............................................................................................................................................................................................
Moderate/severe depression 26 (8.6) 40 (2.2) 4.19 (2.52; 6.98) 5.53 (2.09; 14.61)
c
High stress 124 (41.2) 420 (23.2) 2.34 (1.80; 3.60) 1.59 (1.03; 2.44)
a
CI: condence interval.
b
Adjusted for: Age, education, household income, number of live born children and number of prior pregnancies.
c
High stress: Score of 19 on the Perceived Stress Scale.
Median Range U
........................................................................................
Conclusions
RPL patients 9.0 0; 58 221944 Symptoms of stress and depression are frequent among women referred
Comparison group 7.0 0; 43 P , 0.001a to the Danish RPL Unit. We hope that the present study will result in a
a
heightened awareness of mental distress among care providers for
Mann Whitney U-test.
women with RPL.
had tried to conceive was relatively high (mean 5.4 months). Unfortu-
nately, we did not have information on country of origin in the compari- Supplementary data
son group, but as the questionnaire is only available in Danish, we assume
Supplementary data are available at http://humrep.oxfordjournals.org/.
that the majority have an ethnically Danish background or have at least
lived in Denmark for a substantial period of time. This is supported by
the signicantly higher proportion of non-native Danish women in the
RPL non-participating group. Acknowledgement
It is a key strength in this cross-sectional study that we are able to dir- A.M.K. would like to thank Lisbeth Egestad, former nurse at the RPL Unit,
ectly compare prevalence of stress and depression as both groups used for the idea for the study.
the same psychometric scales. We cannot rule out that some of the
women in the Soon Parents Study may suffer from RPL themselves, as
this information was not available to us. If anything, this would lead to
an underestimation of the difference between women with RPL and
Authors roles
other women attempting to have a child. A.M.K. initiated the study of stress and depression among women with
As the RPL patients are selected a priori by having had at least three RPL, performed all statistical analyses and wrote the paper. L.R.O. pro-
pregnancy losses in their history, it is not surprising that they are older vided the questionnaires used in the RPL group, contributed to data in-
than the women in the comparison group, of which more than half had terpretation and critically revised the manuscript. E.M.M. is the principal
not previously been pregnant. This selection can also reasonably investigator in the Soon Parents Study, provided the base line data on the
account for higher household income and higher educational levels. women in this cohort, contributed to data interpretation and critically
The women in the two groups may differ on variables that we have not revised the manuscript. O.B.C. contributed to data interpretation and
adjusted for, because they are unknown to us, but they are alike in the critically revised the manuscript. H.S.N. supervised A.M.K. in data
desire to have a child. It is our opinion that the comparison between analysis, participated in data interpretation and critically revised the
the two groups is valid and relevant. manuscript.