DOI: 10.1111/j.1471-0528.2008.01943.

x
www.blackwellpublishing.com/bjog
Epidemiology

Does miscarriage in an initial pregnancy lead to

adverse obstetric and perinatal outcomes in the
next continuing pregnancy?
S Bhattacharya,a J Townend,b A Shetty,c D Campbell,a S Bhattacharyac
a Dugald Baird Centre for Research on Women’s Health, Aberdeen Maternity Hospital, Aberdeen, UK b Department of Public Health, University of

Aberdeen, Aberdeen, UK c Department of Obstetrics and Gynaecology, University of Aberdeen, Aberdeen Maternity Hospital, Aberdeen, UK
Correspondence: Dr S Bhattacharya, Dugald Baird Centre for Research on Women’s Health, Aberdeen Maternity Hospital, Cornhill Road, Aberdeen
AB25 2ZL, UK. Email sohinee.bhattacharya@abdn.ac.uk

Accepted 16 August 2008. Published OnlineEarly 8 October 2008.

Objective To explore pregnancy outcomes in women following an
initial miscarriage.
Design Retrospective Cohort Study.
Setting Aberdeen Maternity Hospital, Aberdeen, Scotland.
Population All women living in the Grampian region of Scotland
with a pregnancy recorded in the Aberdeen Maternity and
Neonatal Databank between 1986 and 2000.
Main outcome measures (A) Maternal outcomes: Pre-eclampsia,
antepartum haemorrhage, threatened miscarriage, malpresenation,
induced labour, instrumental delivery, Caesarean delivery,
postpartum haemorrhage and manual removal of placenta. (B)
Perinatal outcomes: preterm delivery, low birth weight, stillbirth,
neonatal death, Apgar score at 5 minutes.
Methods Retrospective cohort study comparing women with
a first pregnancy miscarriage with (a) women with one previous
successful pregnancy and (b) primigravid women. Data were
extracted on perinatal outcomes in all women from the
Aberdeen Maternity and Neonatal Databank between 1986 and
2000.
Results We identified 1561 women who had a first miscarriage
(1404 in the first trimester and 157 in the second trimester),
10 549 who had had a previous live birth (group A) and 21 118
primigravidae (group B). The miscarriage group faced a higher risk
of pre-eclampsia (adj OR 3.3, 99% CI 2.6–4.6), threatened
miscarriage (adj OR 1.7, 99% CI 1.5–2.0), induced labour (adj OR
2.2, 99% CI 1.9–2.5), instrumental delivery (adj OR 5.9, 99% CI
5.0–6.9), preterm delivery (adj OR 2.1, 99% CI 1.6–2.8) and low
birthweight (adj OR 1.6, 99% CI 1.3–2.1) than group A. They were
more likely to have threatened miscarriage (adj OR 1.5, 99% CI
1.4–1.7), induced labour (adj OR 1.3, 99% CI 1.2–1.5), postpartum
haemorrhage (adj OR 1.4, 99% CI 1.2–1.6) and preterm delivery
(adj OR 1.5, 99% CI 1.2–1.8) than group B.
Conclusion An initial miscarriage is associated with a higher risk
of obstetric complications.
Keywords Maternal outcomes, miscarriage, perinatal outcomes.

Please cite this paper as: Bhattacharya S, Townend J, Shetty A, Campbell D, Bhattacharya S. Does miscarriage in an initial pregnancy lead to adverse obstetric and
perinatal outcomes in the next continuing pregnancy? BJOG 2008;115:1623–1629.

three previous miscarriages (recurrent miscarriage). Reginald

Introduction
Miscarriage or spontaneous pregnancy loss before 24 com-
pleted weeks of gestation is estimated to affect 10–15% of
pregnancies.1 While a spontaneous miscarriage is distressing
at any time, it is particularly so if it occurs in the first preg-
nancy. Previous work has focused on the risk of further mis-
carriage in these women,2 but there have been few attempts to
study obstetric and perinatal outcomes in subsequent preg-
nancies that progressed beyond 24 weeks. Four such studies
were identified by searching Ovid MEDLINE, EMBASE and
CINAHL bibliographic databases. All examined obstetric out-
comes in continuing pregnancies in women who had at least
et al.3 found increased rates of preterm delivery, small for date
and perinatal mortality in women with previous miscarriages;
Hughes et al.4 found these to be no higher than in the control
population. In another study, Tulppala et al.5 reported higher
risks of preterm deliveries, intrauterine growth restriction and
impaired glucose tolerance in infants born after previous
miscarriages. In a more recent study, Jivraj et al.6 compared
obstetric and neonatal outcomes of 162 pregnancies in
women with previous recurrent miscarriage with a control
population delivering during the same period. The authors
found increased rates of preterm delivery, growth restriction,
caesarean section and perinatal mortality, although there were

ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1623
 Bhattacharya et al.

no differences in the incidence of hypertensive disorders of ther miscarriage, ectopic pregnancy or termination in their
pregnancy or diabetes. next pregnancy were excluded. Although multiple births were
In this study, we aimed to compare maternal and perinatal excluded from the original cohorts, 17 women, all belonging
outcomes in the next continuing pregnancy in women with to the first control group (group A), had twin deliveries in
and without a spontaneous miscarriage in their first preg- their second pregnancy and were also excluded. The flow
nancy using data from the Aberdeen Maternity and Neonatal chart showing loss to follow up of the exposed cohort and
Databank (AMND). comparison group A is presented in Figure 1.
Sociodemographic characteristics of the women in the
three groups were compared to assess any differences.
Subjects and methods
As regards maternal outcomes, data were obtained on the
Since 1950, the AMND has recorded and stored information incidence of threatened miscarriage (defined as first-trimester
on all pregnancy-related events occurring in a geographically bleeding with ultrasound evidence of fetal cardiac activity),
defined population living in the Grampian region of Scotland. pre-eclampsia, antepartum haemorrhage including placenta
It is possible to link all pregnancy records of individual praevia, abruption and unclassified APH and postpartum
women. It therefore offers the ideal opportunity to study haemorrhage in the exposed and unexposed cohorts. Pre-
the effects of an initial miscarriage on subsequent continuing eclampsia, eclampsia and gestational hypertension are coded
pregnancies in Aberdeen, which has a relatively stable popu- in the AMND according to the International Society for the
lation, minimising loss to follow up. Completeness of the Study of Hypertension in Pregnancy (ISSHP) definition as
database is checked against the number of deliveries recorded follows: Gestational hypertension is defined as a diastolic
in the National Health Service records office at the end of each blood pressure of ‡110 mmHg on any one occasion or dia-
year. There are numerous consistency checks in place to stolic blood pressure of ‡90 mmHg on any two or more
ensure validity of data entry. Although there is no formal occasions at least 4 hours apart; and pre-eclampsia as hyper-
quality assurance mechanism in place for checking the reli- tension as above with proteinuria ‡300 mg/24 hours or pro-
ability of the entire database, several previous surveys and ad teinuria found in two urine specimens collected at least 4
hoc case note reviews have found the data to be more than hours apart. Antepartum haemorrhage is defined as vaginal
90% accurate with regard to the variables studied. The back- bleeding occurring in pregnancy after 28 weeks of gestation
ground and details of the database are described in the website warranting hospital consultation. Further subclassification
www.abdn.ac.uk/amnd.
Variables are coded in the AMND according to the Inter-
national Classification of Diseases (9th Revision). All miscar- First miscarriage (P0+1) First livebirth (P1+0)
riages that warrant a hospital visit are recorded, but those exposed cohort unexposed group A
n =3860 n=20047
occurring at home may be missed.
Formal ethical approval was not considered necessary by
the North of Scotland Research Ethics Committee as only No further pregnancies
No further pregnancies
anonymised routinely collected data were used for the re- recorded
recorded
7742 (39.6%)
search. Permission was obtained from the Caldicott guardians 1161 (30.8%)
of the AMND.
In this retrospective cohort study, data were extracted on Miscarriage in second
Further miscarriage
all women who had a primary pregnancy event between 1986 656 (17.0%) pregnancy
and 2000 from the AMND. Women with multiple pregnan- 728 (3.3%)
cies and major congenital abnormalities diagnosed by ultra-
sound were excluded. Women who had had a spontaneous Ectopic pregnancy
Ectopic pregnancy
204 (1%)
miscarriage resulting in pregnancy loss before 24 completed 262 (6.8%)
weeks of gestation constituted the exposed cohort. There were
two control groups or unexposed cohorts for this study. The
Termination of second Termination of second
first group consisted of women who had had a live birth pregnancy pregnancy
beyond 24 weeks in their first pregnancy (P1+0)—group A. 120 (3.1%) 847 (4.6%)
The second group comprised women in their first continuing
pregnancy (P0+0) beyond 24 weeks during the study period— Second continuing Second continuing
group B. Women in the exposed cohort and group A of the pregnancy pregnancy
1561 (40.4%) 10 549 (52.6%)
unexposed cohort were followed up to 5 years after their
initial pregnancy to identify any subsequent pregnancies. Figure 1. Flow chart showing loss to follow up of exposed and unexposed
Those who did not have any further pregnancies or had a fur- (group A) cohorts.

1624 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
 Pregnancy outcomes subsequent to initial miscarriage

into placenta praevia, abruption or nonspecific APH is made

retrospectively on ultrasound diagnosis. Postdelivery blood
loss as assessed by the attending midwife or doctor is recorded
in the AMND as a continuous variable in millilitres. This was
later recoded for analysis purposes into a binary variable—
postpartum haemorrhage if it was more than 500 ml.7 Labour
and delivery characteristics like induction of labour and
instrumental or caesarean delivery, which are available in
the AMND, were also compared. Perinatal outcomes included
preterm delivery before 37 and 34 completed weeks, mal-
presentation or position at term, birthweight <2500 g, still-
birth and early neonatal death. Gestational age at delivery is
recorded in the AMND according to ultrasound assessment
from 1986 onwards when it became universally available. For
ease of analysis, this variable was categorised into preterm
delivery (before 37 completed weeks of gestation) and very
preterm delivery (before 34 completed weeks).
Data on smoking status were missing in 24.8% of the
exposed cohort and 22.1 and 20.3% of comparison groups
A and B, respectively. Maternal weight at the first antenatal
visit was not recorded in 7.5% women with prior miscarriage
and 8.0 and 8.2% women, respectively, in groups A and B.
Missing values were entered as categories into the final logistic
regression model to maximise the use of available data.
Statistical analysis was carried out using Statistical Pack-
age for Social Scientists (SPSS) version 13 (SPSS Inc.,
Chicago, IL, USA). Continuous variables were compared
using Student’s t test or analysis of variance, while categor-
ical variables by chi-square test and crude odds ratios cal-
culated. Possible confounders identified by univariate
analysis were adjusted for by means of binary logistic
regression, and results were presented as adjusted odds
ratios with 99% CI. Statistical significance was set at 1%
significance level.
Results
A total of 1561 women with an initial miscarriage were
identified (exposed cohort). Of these, 1404 women miscar-
ried in the first trimester and 157 women in the second
trimester.
The first unexposed cohort (group A) comprised 10 549
women who had previously delivered after 24 completed
weeks in their first pregnancy and were pregnant for the sec-
ond time. The second unexposed cohort (group B) included
21 118 women who delivered after 24 weeks in their first
pregnancy during the study period.
The median (interquantile range, IQR) time between the
first and the second pregnancies was 15 (4.8) weeks in the
group with miscarriage, while in the women with a delivery in
the first pregnancy, it was 2.9 (0.2) years.
Table 1 shows the comparison of sociodemographic
characteristics of the women in the three different groups.
Compared with women with a previous delivery (group A),
women with an initial miscarriage were younger (mean [SD]
age 27.4 [5.3] versus 28.7 [4.9] years), more likely to be mar-
ried or cohabiting (1197 [85.1%] versus 8513 [80.7%]) and
smokers (702 [50%] vs. 3597 [34.1%]). They were less likely
to belong to social class I/II as defined by the Registrar
General’s classification (316 [22.5%] versus 4039 [38.3%]).
The women in the miscarriage group were also more likely
to have had infertility treatment in comparison with women
in group A, although this variable failed to reach statistical
significance. In comparison with primigravidae (group B),
women with a previous miscarriage were older (mean age
27.4 [5.3] versus 26 [5.2] years), more likely to be smokers
(50 versus 33.6%), married or cohabiting (85.1 versus 80.4%)
and less likely to belong to social class I/II (22.5 versus 25.4%).
Type 1 diabetes was more common in the exposed cohort

Table 1. Comparison of characteristics between women with a first-trimester miscarriage in their first pregnancy and groups A and B

Characteristics* Group A (P110) P value Miscarriage (P011) Group B (P010) P value**

(n 5 10 549) (n 5 1404) (n 5 21 118)

Age at second pregnancy event (years) 28.7 (4.9) <0.001 27.4 (5.3) 26.0 (5.2) <0.001
Interpregnancy interval (years) 2.9 (0.2) <0.001 0.27 (0.04) N/A
Height (cm) 162.8 (6.4) 0.790 162.9 (6.3) 162.6 (6.3) 0.38
Body mass index (Kg/m2) 24.20 (2.61) 0.013 25.53 (1.92) 24.11 (2.1) 0.011
Married or cohabiting 8513 (80.7%) <0.001 1197 (85.1%) 16976 (80.4%) <0.001
Husband/partner’s social class (I/II) 4039 (38.3%) <0.001 316 (22.5%) 5357 (25.4%) 0.02
Smokers 3597 (34.1%) <0.001 702 (50.0%) 7802 (33.6%) <0.001
Diabetes (type 1) 32 (0.3%) 0.502 7 (0.3%) 56 (0.2%) <0.001
Chronic hypertension 157 (1.0%) 0.600 16 (1.1%) 245 (1.2%) 0.528
Infertility treatment 225 (1.4%) 0.053 28 (2.0%) 336 (1.6%) 0.081

*Expressed as mean (SD), median (IQR) or n (%).

**Statistically significant p-values are presented in bold.

ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1625
 Bhattacharya et al.

than in group B. The women in the miscarriage group had
a higher body mass index than the women in both compar-
ison groups. The average heights of women in the three
groups were comparable.
The risks of obstetric complications were different in the
two comparison groups.
Exposed cohort versus group A
In comparison with women who had had a previous delivery,
women with an initial miscarriage were 3.3 times (99% CI
2.6–4.6) more likely to suffer from pre-eclampsia, 1.7 times
(99% CI 1.5–2.0) more likely to have threatened miscarriage
and 1.3 times (99% CI 1.1–1.6) more prone to have nonspe-
cific antepartum haemorrhage. Women with an initial mis-
carriage were also more likely to have induction of labour (adj
OR 2.2 [99% CI 1.7–2.6]), instrumental vaginal delivery (adj
OR 5.9 [99% CI 5.0–6.9]) and manual removal of placenta
(adjusted OR 1.5 [99% CI 1.1–2.1]). Elective caesarean sec-
tion was, however, more common in women who had had
a previous delivery. These findings are shown in Table 2.
Table 3 shows the risks of perinatal complications in the
exposed cohort (P0+1) in comparison with group A (P1+0).
Preterm delivery, after 34 weeks of gestation (adjusted OR 1.6
[99% CI 1.2–2.0]), was more common in the miscarriage
group, as was malpresentation (adjusted OR 1.7 [99% CI
1.5–1.8]). After adjusting for gestational age and sex of the
baby, babies with birthweight <2500 g were 1.6 times more
common (99% CI 1.3–2.1) in mothers who had had a pre-
vious miscarriage.
Exposed cohort versus group B
In comparison with primigravidae (group B), threatened mis-
carriage (adjusted OR 1.5 [99% CI 1.4–1.7]), induction of
labour (adjusted OR 1.3 [99% CI 1.2–1.5]) and postpartum
haemorrhage (adjusted OR 1.4 [99% CI 1.2–1.6]) were more
common in the miscarriage group. These findings are pre-
sented in Table 4.
In terms of perinatal complication (Table 5), preterm
delivery after 34 weeks of gestation (adjusted OR 1.5 [99%
CI 1.2–1.8]) was higher in the exposed cohort after adjusting
for confounders.
The median gestational period of miscarriage in the
exposed cohort was 9 weeks. We conducted a subgroup anal-
ysis of all outcomes (data shown in additional files) comparing
them between women who had had a first-trimester miscar-
riage with those who had a miscarriage in their second trimes-
ter. There were no statistically significant differences in the
outcomes in the two subgroups studied except that the risk
of preterm delivery was higher in the women with second-
trimester miscarriage (adjusted OR 2.8 [95% CI 1.2–3.6]).
Discussion
In comparison with women with a previous successful preg-
nancy, our data suggest that women with an initial miscar-
riage have an increased risk of some obstetric complications.
These include pre-eclampsia, threatened miscarriage, nonspe-
cific antepartum haemorrhage, induced labour, instrumental
delivery and manual removal of placenta. They are also more

Table 2. Risk of subsequent obstetric complications and interventions in women with miscarriage compared with women with a pregnancy
beyond 24 weeks in their first pregnancy (group A)

Complications Miscarriage* Group A (P110)* Crude OR (99% CI) Adjusted OR** (99% CI)
(n 5 1404) (n 5 10 549)

Pre-eclampsia 62 (4.4) 147 (1.4) 3.2 (2.4–4.3) 3.3 (2.6–4.6)

Threatened miscarriage 381 (27.1) 1835 (17.8) 2.7 (2.4–3.0) 1.7 (1.5–2.0)
Placental abruption 11 (0.8) 50 (0.5) 1.5 (0.9–2.1)
Placenta praevia 5 (0.4) 38 (0.4) 1.0 (0.7–2.8)
Other APH*** 151 (10.7) 829 (8.0) 1.3 (1.1–1.6) 1.3 (1.1–1.6)
Malpresentation**** 253 (18.0) 532 (5.1) 2.7 (2.2–3.2) 1.7 (1.5–1.8)
Induced labour 464 (33.0) 1925 (18.6) 2.2 (1.9–2.5) 2.2 (1.7–2.6)
Instrumental delivery 375 (26.7) 597 (5.8) 5.9 (5.1– 6.7) 5.9 (5.0–6.9)
Elective caesarean section 59 (4.2) 828 (8.0) 0.5 (0.4–0.6) 0.5 (0.3–0.6)
Postpartum haemorrhage (PPH)***** 169 (12.0) 898 (8.7) 1.5 (1.2–1.7) 1.1 (0.7–1.3)
Manual removal 55 (3.9) 244 (2.4) 1.7 (1.3–2.3) 1.5 (1.1–2.1)
of placenta

*Values expressed as n (%).

**All odds ratios adjusted for age, year of delivery, interpregnancy interval, marital status, body mass index, husband/partner’s social class
and smoking.
***Odds ratio for other APH also adjusted for threatened miscarriage.
****Malpresentation also adjusted for preterm delivery.
*****Odds ratio for PPH also adjusted for induced labour, instrumental delivery, caesarean section and manual removal of placenta.

1626 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
 Pregnancy outcomes subsequent to initial miscarriage

Table 3. Risk of subsequent perinatal complications in women with a miscarriage compared with women with a pregnancy beyond 24 weeks in
their first pregnancy (group A)

Complications Miscarriage* Group A (P110)* Crude OR (99% CI) Adjusted OR** (99% CI)
(n 5 1404) (n 5 10 549)

Preterm delivery*** 128 (9.1) 540 (3.4%) 2.8 (2.3–3.5) 2.1 (1.6–2.8)
,34 weeks 39 (2.8) 162 (1.0%) 2.8 (2.0–3.5) 1.5 (1.1–2.2)
.34 weeks 89 (6.3) 378 (2.4%) 2.8 (2.2–3.5) 1.6 (1.2–2.0)
Spontaneous preterm delivery 19 (1.4) 234 (1.1%) 1.1 (0.9–2.1)
<34 weeks
Spontaneous preterm delivery 53 (3.8) 591 (2.8%) 1.4 (1.1–1.8)
>34 weeks
Stillbirth 14 (1.0) 42 (0.3%) 3.6 (2.0–6.5) 1.9 (1.1–3.6)
Neonatal death**** 11 (0.8) 24 (0.2%) 5.1 (2.5–10.1) 2.3 (1.1–4.8)
Birthweight <2500 g***** 120 (8.5) 472 (4.6%) 1.9 (1.6–2.4) 1.6 (1.3–2.1)
Apgar at 5 minutes ,7 48 (3.3) 597 (3.8) 1.1 (0.8–1.7)

*Values expressed as n (%).

**All odds ratios adjusted for age, year of delivery, interpregnancy interval, marital status, body mass index, husband/partner’s social class
and smoking.
***Preterm delivery also adjusted for threatened miscarriage, pre-eclampsia, antepartum haemorrhage and induction of labour.
****Neonatal death also adjusted for preterm delivery.
*****Low birthweight adjusted for preterm delivery and sex of baby.

prone to preterm delivery, malpresentation and low birth-
weight. Many of these risks, however, are no higher than those
in primigravidae. Thus, women who have an initial early
pregnancy loss behave like ‘virtual primigravidae’ in their
next pregnancy not only in terms of their labour and delivery
characteristics but also with regard to pregnancy complica-
tions and neonatal outcomes.
In comparison with primigravid women in a continuing
pregnancy, the risk of threatened miscarriage, induced
labour, postpartum haemorrhage and preterm delivery
were higher in women with an initial miscarriage. The
results of our study leave no room for doubt that women
with a first miscarriage have a higher risk of adverse out-
comes in the subsequent pregnancy compared with
women who have a successful first pregnancy. When com-
pared with primigravidae, however, interpretation of our
findings is not so straightforward. Although there is some
evidence of incremental risk of selected complications,

Table 4. Risk of obstetric complications and interventions in women with miscarriage in their first pregnancy compared with primigravidae
(group B)

Complications Miscarriage* (n 5 1404) Group B* (n 5 21 118) Unadjusted OR (99% CI) Adjusted OR** (99% CI)

Pre-eclampsia 62 (4.4%) 811 (3.8%) 1.1 (0.9–1.3)

Threatened miscarriage 381 (27.1%) 3938 (18.6%) 1.6 (1.4–1.8) 1.5 (1.4–1.7)
Placental abruption 11 (0.8%) 150 (0.7%) 1.1 (0.6–2.1)
Placenta praevia 5 (0.4%) 44 (0.2%) 1.8 (0.7–4.6)
Other APH 151 (10.7%) 2024 (9.6%) 1.1(1.0–1.4)
Malpresentation*** 253 (18.0%) 3268 (15.5%) 1.2 (1.1–1.3) 1.2 (1.0–1.4)
Induced labour 464 (33.0%) 5681 (26.9%) 1.4 (1.2–1.5) 1.3 (1.2–1.5)
Instrumental delivery 375 (26.7%) 5358 (25.4%) 1.1 (0.9–1.2)
Elective caesarean section 59 (4.2%) 749 (3.5%) 1.2 (0.9–1.6)
PPH**** 169 (12.0%) 1792 (8.5%) 1.5 (1.3–1.8) 1.4 (1.2–1.6)
Manual removal 55 (3.9%) 698 (3.3%) 1.2 (0.9–1.6)
of placenta

*Values expressed as n (%).

**All odds ratios adjusted for age, year of delivery, marital status and smoking.
***Malpresentation also adjusted for preterm delivery.
****Odds ratio for PPH also adjusted for induced labour.

ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology 1627
 Bhattacharya et al.
Table 5. Risk of perinatal complications in women with miscarriage in their first pregnancy compared with primigravidae (group B)
Complications Miscarriage* Group B*
(n 5 1404) (n 5 21 118)
Preterm delivery*** 144 (9.2%) 1384 (6.8%)
,34 weeks 39 (2.8%) 456 (2.2%)
.34 weeks 89 (6.3%) 928 (4.4%)
Stillbirth 14 (1.0%) 133 (0.6%)
Neonatal death**** 11 (0.8%) 77 (0.4%)
Birthweight <2500 g***** 120 (8.5%) 1509 (7.1%)
Apgar at 5 minutes <7 48 (3.3%) 646 (3.1%)
*Values expressed as n (%).
**All odds ratios adjusted for age, year of delivery, marital status and smoking.
***Preterm delivery also adjusted for induction of labour.
****Neonatal death also adjusted for preterm delivery.
*****Low birthweight adjusted for preterm delivery and sex of baby.
the absolute risk is small and may not be clinically
important.
Despite an extensive literature search, we were unable to
find any papers that had specifically looked at pregnancy out-
comes following a single miscarriage. Thus, to our knowledge,
this is the only population-based study of its kind. The AMND
records information contemporaneously as the obstetric events
occur, thereby minimising recall bias. Stringent coding criteria
and consistency checks are in place adding to the reliability of
the data. Most previous studies in this area have used an
unselected population attending a particular hospital during
a specified time period as their control group. We compared
our exposed cohort with two control groups—primigravidae
and second gravidae. The influence of parity when looking at
certain obstetric complications like pre-eclampsia is obvious.8
Jivraj et al.6 divided the women with history of recurrent
miscarriage into two groups: primary, that is nulliparous,
and secondary, that is parous, and compared the outcomes.
They suggest that the control population should ideally be
matched on age and parity with the exposed cohort. The
parity of women who miscarry in their first pregnancy is
debatable. In epidemiological terms, these women are multi-
gravidae and should therefore be compared with multigravid
controls. In clinical terms, they could be expected to behave
like primiparae and should therefore be matched with nullip-
arous women.9 We have performed both, adding strength and
validity to our conclusions.
Like other observational studies of its kind, this study has
several limitations. Its retrospective nature could result in
selection bias, and the tertiary referral centre setting could
have influenced the prevalence of some of the outcomes. This
is, however, likely to be minimal as only women residing in
Aberdeen city and district were selected as the control pop-
ulation. As Aberdeen Maternity Hospital is the only hospital
in the area, it provides secondary as well as tertiary level care.
1628 ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
Unadjusted OR (99% CI) Adjusted OR** (99% CI)
1.4 (1.3–1.6)
1.3 (0.9–1.8)
1.5 (1.2–1.8) 1.5 (1.2–1.8)
1.6 (0.9–2.8)
2.1 (1.2–3.9) 2.1 (1.0–3.9)
1.2 (1.0–1.5) 1.2 (0.9–1.4)
1.1 (0.9–1.2)
In addition, we were only able to capture miscarriages that led
to hospital contact and as such cannot generalise our findings
to all women with a first miscarriage. As we have studied
women over a long time period (15 years), changes in obstet-
ric practice are bound to have influenced the outcomes. How-
ever, as the unexposed cohort were identified from the same
setting and similar time period, this is unlikely to have a sig-
nificant effect. Moreover, we entered the year of delivery as an
independent variable in the logistic regression model to adjust
for this. It was not possible to do any subgroup analyses on
the basis of aetiology of miscarriage because of unavailability
of data. Furthermore, miscarriages occurring in the second
trimester are usually of a different aetiology, and any subse-
quent pregnancies are likely to have different outcomes com-
pared with those occurring in the first trimester. As our
miscarriage group consisted mainly of first-trimester loss,
we did a subgroup analysis with second-trimester miscar-
riages. This group showed a higher risk of preterm delivery.
Lastly, multiple comparisons using a large data set invariably
show some associations that may or may not be clinically
significant. We set alpha at 0.01 level to take this into account.
Our results show that any pregnancy following an initial
miscarriage is associated with a degree of increased obstetric
and perinatal risk. This study questions the evidence behind the
attribution of risk only to women who have had three or more
consequent miscarriages (recurrent miscarriage). At the same
time, the magnitude of this risk is likely to be small as women
with a history of recurrent miscarriage achieve good preg-
nancy outcomes with support from an early pregnancy unit.10
Unfortunately, we could not find any other study in the
literature that had attempted to quantify the obstetric risks
following a single miscarriage that could corroborate or refute
our findings. Our findings agree with those presented in some
studies that examined these risks in women with recurrent
miscarriage. Perinatal adverse outcomes such as preterm

delivery, intrauterine growth restriction and perinatal death
generally appear to be higher3,5,6 in women with recurrent
miscarriage in comparison with the general population.
Hughes et al.,4 however, found these risks to be no higher
than those in the local obstetric population. In another pub-
lished report, Hammoud et al.11 found that in nulliparous
women, the risk of preterm delivery increased exponentially
with increasing number of previous miscarriages, although
the risk associated with one previous miscarriage was small
(adj OR 1.13, 95% CI 1.01–1.26). We found a similar risk of
preterm delivery (adj OR 1.5, 95% CI 1.2–1.8), but only of
deliveries after 34 completed weeks of gestation. It is difficult
to explain the increased risk of preterm delivery in women
with a history of miscarriage given that most early pregnancy
losses occur as a result of implantation defects.12 It is possible
that this may be related to surgical management of the pre-
vious incomplete miscarriage. The growing trend in expectant
or medical management of incomplete miscarriages will
enable us in a few years’ time to accumulate sufficient data
to compare outcomes based on different management re-
gimes of previous miscarriages.
Conclusions
Women with an initial miscarriage are at increased risk of
some obstetric and perinatal complications in comparison
with women who have a successful initial pregnancy. While
overall risks are less pronounced in comparison with primi-
gravid women, the risks of threatened miscarriage, induced
labour and postpartum haemorrhage are higher in women
with an initial miscarriage. These incremental risks are small
and may not be clinically significant.
Disclosure of interest
The authors declare that they have no conflict of interest.
Contribution to authorship
Sohinee Bhattacharya was responsible for designing the study,
analysing the data and writing the first draft of the report. JT
provided statistical support and commented on the final ver-
sion of the paper. AS was involved in designing the study and
providing comments on the final version of the paper. DC
helped to design the study and facilitated data extraction.
Siladitya Bhattacharya conceived the research question and
provided expert knowledge during the design, analysis and
writing of the paper. All the authors contributed to the final
draft of the manuscript.
ª 2008 The Authors Journal compilation ª RCOG 2008 BJOG An International Journal of Obstetrics and Gynaecology
Pregnancy outcomes subsequent to initial miscarriage
Details of ethical approval
The protocol for this project was submitted to the North of
Scotland Research Ethics Committee, who declared in a letter
that no formal ethical approval was necessary as researchers
would be using anonymised routinely collected data.
Funding
Endowment grant from National Health Services Research
and Development, Grampian region, Scotland.
Acknowledgements
The authors would like to thank Mrs Linda Murdoch for
extracting the data from the AMND. j
References
1 Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer JP,
Canfield RE, et al. Incidence of early loss of pregnancy. N Engl J Med
1988;319:189–94.
2 Regan L, Braude PR, Trembath PL. Influence of past reproductive
performance on risk of spontaneous abortion. BMJ 1989;299:541–5.
Erratum appears in BMJ 1989;299:1082.
3 Reginald PW, Beard RW, Chapple J, Forbes PB, Liddell HS, Mowbray JF,
et al. Outcome of pregnancies progressing beyond 28 weeks gestation
in women with a history of recurrent miscarriage. Br J Obstet Gynaecol
1987;94:643–8.
4 Hughes N, Hamilton EF, Tulandi T. Obstetric outcome in women after
multiple spontaneous abortions. J Reprod Med 1991;36:165–6.
5 Tulppala M, Palosuo T, Ramsay T, Miettinen A, Salonen R, Ylikorkala O.
A prospective study of 63 couples with a history of recurrent sponta-
neous abortion: contributing factors and outcome of subsequent preg-
nancies. Hum Reprod 1993;8:764–70.
6 Jivraj S, Anstie B, Cheong YC, Fairlie FM, Laird SM, Li TC. Obstetric and
neonatal outcome in women with a history of recurrent miscarriage:
a cohort study. Hum Reprod 2001;16:102–6.
7 World Health Organization. International Statistical Classification of
Diseases and Related Health Problems, Tenth revision. Geneva: WHO,
1992.
8 Seidman DS, Ever-Hadani P, Stevenson DK, Gale R. The effect of abor-
tion on the incidence of pre-eclampsia. Eur J Obstet Gynecol Reprod
Biol 1989;33:109–14.
9 Nurminen T. On adjusting for the outcome of previous pregnancies in
epidemiologic reproductive studies. See comment. Epidemiology
1995;6:84–6.
10 Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained
recurrent first trimester miscarriage. Hum Reprod 1997;12:387–9.
11 Hammoud AO, Merhi ZO, Diamond M, Baumann P. Recurrent preg-
nancy loss and obstetric outcome. Int J Gynecol Obstet 2007;96:28–9.
12 Webb PD, Glasser SR. Implantation. In: Wolf DP, Quigley MA, editors.
Human in Vitro Fertilization and Embryo Transfer. New York: Plenum
Press; 1984. pp. 341–66.
1629