Obstetric Risk Factors Associated with Placenta Previa Development: Case-Control Study

44(6):728-733,2003
CLINICAL SCIENCES
Obstetric Risk Factors Associated with Placenta Previa Development: Case-Control
Study
Lea Tuzoviæ, Josip Djelmiš, Marcela Ilijiæ
Department of Obstetrics and Gynecology, Zagreb University Hospital Center and School of Medicine, Zagreb,
Croatia
Aim. To evaluate potential risk factors and perinatal outcome of pregnancies complicated with placenta previa in Cro-
atian population of pregnant women recruited from the largest tertiary care perinatal center in Croatia.
Methods. This retrospective case-control study included a total of 202 singleton pregnancies with placenta previa dur-
ing a 10-year study period and 1,004 randomly selected simple singleton controls. Data on potential risk factors for pla-
centa previa development were carefully extracted from medical records, reviewed, and compared with a control
group of women. Data were statistically analyzed with chi-square test and Mann-Whitney U test, and crude odds ratio
(OR) with 95% confidence interval (95% CI) were provided.
Results. The incidence of placenta previa was 0.4%. Factors significantly associated with a placenta previa develop-
ment were advanced maternal age (especially >34 years, even after adjustment for high parity), gravidity of 3 and
more (OR, 4; 95% CI, 2.5-6.6), more than one previous delivery (OR, 2.76; 95% CI, 1.7-4.3), history of previous cesar-
ean sections (OR, 2.0; 95% CI, 1.17-3.44), abortions (OR, 2.8; 95% CI, 2.04-3.83), and presence of various uterine ab-
normalities (OR, 8.5; 95% CI, 1.75-44.5). The risk was significantly increased after two previous cesarean sections
(OR, 7.32; 95% CI, 2.1-25) and after one previous abortion (OR, 4.8; 95% CI, 2.7-8.3). No difference between the
groups was found regarding the history of previous placenta previa, drug abuse, and male sex at birth. Smoking history
was significantly less frequent in women with placenta previa than controls (16.3% vs 25.6%, chi-square=7.9,
p=0.007). The main perinatal complication was preterm birth, with 14-fold higher risk in women with placenta previa.
Preterm infants of mothers with placenta previa were more likely to have lower first- (6 vs 10, p<0.001) and fifth-min-
ute median Apgar scores (8 vs 10, p<0.045). Term infants of mothers with placenta previa had significantly lower birth
weight then their controls (3,300 vs 3,500 g, p<0.001).
Conclusion. The most important obstetric factors for placenta previa development were advanced maternal age espe-
cially >34 years, 3 or more previous pregnancies, parity of 2 and more, rising number of previous abortions, and his-
tory of previous cesarean section, but not child sex at birth, history of drug abuse and previous placenta previa. Smok-
ing cigarettes was significantly less frequent in women with placenta previa. Preterm delivery still remains the greatest
problem in pregnancies complicated with placenta previa.
Key words: cesarean section; Croatia; placenta previa; risk factors; smoking
Placenta previa is a rare form of impaired placen-
placenta previa is highly suggestive, the etiology of
tation where placenta lies low in the uterine cavity,
this condition still remains obscure. The strongest
covering completely or partially the internal cervical
connection was found between previous history of
ostium and thereby preventing normal vaginal deliv-
cesarean section (5,10-14), high parity (10,11,14),
ery. It is one of the main causes of vaginal bleeding in
and advanced maternal age (15), but the strength of
the third trimester (1), and a significant cause of ma-
the connection varies from study to study. Moreover,
ternal (2,3) and perinatal morbidity and mortality (4).
in some cases the results of the studies are contradic-
The incidence of placenta previa in pregnant women
tory and deserve further evaluation. Other potential
is approximately 0.3-0.8%, depending upon the pop-
risk factors with more confounding effect on the de-
ulation investigated (3,5-7). A trend of increasing pla-
velopment of placenta previa include history of previ-
centa previa incidence was observed in the past de-
ous spontaneous or induced abortions (8,11), increas-
cade mainly because of an increasing cesarean sec-
ing number of previous cesarean sections (12,13),
tion rate (8) and advancing maternal age at the time of
previous uterine operations, previous placenta previa
first pregnancy (6,9). Although the clinical course of
(16), smoking (17,18) or substance abuse during preg-
728 www.cmj.hr

---

Tuzoviæ et al: Obstetric Risk Factors and Placenta Previa
Croat Med J 2003;44:728-733
nancy (19), multiple gestation (20), and child sex at
abortions, history of previous placenta previa or any other uterine
birth (21,22). As results of the studies in risk factors and
operation or anomaly, history of substance abuse during preg-
nancy (smoking and illicit drugs), child sex, pathological presen-
outcome of placenta previa pregnancies vary around
tations (breech, transverse, or oblique lie), delivery data, and neo-
the world (10-14,16,28), we decided to evaluate po-
natal outcome data (gestational age at delivery, birth weight and
tential risk factors and perinatal outcome of pregnan-
height, Apgar scores, and cord blood pH value).
cies complicated with placenta previa on a large popu-
Statistical Analysis
lation of pregnant women recruited from the largest
All data were analyzed with statistical package program
tertiary care perinatal center in Zagreb, Croatia.
STATISTICA, version 6.0 (StatSoft, Inc., Tulsa, OK, USA). Patients
with placenta previa were compared with those without placenta
Patients and Methods
previa, ie, controls. For the study purpose, detailed multiple vari-
able database was formed. All the data were collected either as
A retrospective case-control study encompassed the 10-year
dichotomous categorical variables (e.g., “yes” or “no” for history
period between January 1992 and December 2001 and was con-
of previous cesarean section), variables with set of multiple differ-
ducted in Women’s Hospital, Zagreb University School of Medi-
ent categories (e.g., different age groups), or as continuous nu-
cine. This is the largest tertiary care center in Croatia, with ap-
meric variables.
proximately 5,000-6,000 deliveries annually. In the same period,
After testing for normality of distribution, continuous vari-
204 cases of placenta previa were identified, with 202 of them
ables were expressed as median because the distribution was not
being singleton pregnancies and 2 multiple twin gestations.
normal. For statistical comparison, non-parametric Mann-Whit-
Study Sample
ney U-test was used. Dichotomous categorical variables were
Study group included 202 singleton pregnancies with pla-
given as percentages. To test independence between two dichot-
centa previa. Placenta previa was defined as a placenta that com-
omous variables, Pearson’s chi-square test was used. Fischer’s ex-
pletely or partially covered the internal cervical ostium, or as the
act test was performed when a single cell in a 2x2 contingency ta-
placenta whose margin reached the edge of internal cervical
ble had an expected frequency less than 5. Crude odds ratio, with
ostium at the time of delivery. For diagnosis of marginal placenta
95 % confidence interval, was also calculated to test a connec-
previa, a cut-off ultrasonographic margin within 1 cm from inter-
tion between an independent and factor variable. Calculated
nal cervical ostium was used. Cases with low-lying placentas
odds ratio served as an approximation of relative risk. For vari-
(n=3) and incomplete data (n=1) were excluded from further
ables with a set of different categories, Mantel-Hanszel chi-square
analysis and were not considered to have placenta previa (Fig. 1).
test for linear trend was used. P-value of less than 0.05 was con-
The diagnosis of placenta previa was established by transabdo-
sidered significant.
minal ultrasonographic imaging performed by trained attending
physicians, and the last ultrasonographic examination before de-
Results
livery was used to establish correct diagnosis. This was particu-
larly important in order to exclude the cases of placenta previa that
Out of a total of 53,042 deliveries at our Hospital
resolved spontaneously during the course of pregnancy. Further-
during the study period, 204 were cases of placenta
more, the diagnosis was confirmed by direct inspection of the pla-
cental location at the time of cesarean section or in the rare cases
previa. From those, 202 were singleton deliveries and
of vaginal delivery by palpating the edge of placenta adjacent to
2 were multiple twin gestations. The calculated inci-
the internal cervical orifice in the presence of complete cervical di-
dence of placenta previa was 0.4 % in our population
latation. These data were derived from operation protocol descrip-
of pregnant women. The incidence was stable be-
tions. Control group consisted of 1,004 simple randomly selected
tween 1992 and 1998 (0.31%-0.4%), but showed a
singleton pregnancies of women either delivered vaginally or by
cesarean section in the 10-year study period recruited from peri-
slight but insignificant increase in 1999 and 2000
natal birth registry from a total number of 53,042 deliveries. For
(0.6%). We analyzed potential risk factors for placen-
each case, 5 randomly selected unmatched controls were chosen.
ta previa development in the study population and
Exclusion criteria were multiple gestations, placenta previa or any
controls (Table 1). The median age of pregnant wo-
other placental abnormality (adherent placenta, placenta accreta,
men with placenta previa was significantly higher in
placenta succenturiata, or placental abruption), and incomplete
data (Fig. 1). Correct gestational age was derived from the first day
women with placenta previa than in controls (31 vs
of the last menstrual period and was checked with ultrasonograp-
28, p<0.001). The distribution according to age
hic evaluation of gestational age.
groups revealed a significantly higher frequency of
women older than 34 years in the placenta previa
Selected for the study group
Randomly selected for the control group
group than in the control group (25.7% vs 13.6%, re-
(n=207)
(n=1,035)
spectively), and at the same time significantly lower
frequency of women younger than 25 (7.9% vs
Excluded from the study:
Excluded from the study:
26.3%, respectively; Table 2). Women with placenta
- multiple gestation (n=2)
- multiple gestation (n=8)
previa were also more likely to be of higher gravidity
- low-lying placenta (n=2)
- placental abruption (n=7)
and parity (Table 2). The risk for placenta previa de-
- incomplete data (n=1)
- adherent placenta (n=6)
- succenturiate placenta (n=1)
velopment increased with increasing number of pre-
- placenta accreta (n=1)
vious pregnancies. Whereas a stable trend of decreas-
- incomplete data (n=8)
ing gravidity toward higher gravidity groups (4+) was
observed in the control group of women, there was an
increasing percentage of women with 3 or more pre-
Placenta previa group (n=202)
Control group (n=1,004)
vious pregnancies among the women with placenta
Figure 1. Selection of pregnant women for the study.
previa. Women with 5+ previous pregnancies had
more than 7-fold higher risk for placenta previa devel-
Outcome Measures
opment. A trend of increasing parity was also ob-
For all women with placenta previa and their controls,
served. The frequency of multiparous women was sig-
medical records were carefully reviewed and multiple parame-
nificantly higher in the group of women with placenta
ters regarding potential obstetric risk factors were extracted and
compared. The following data were obtained: age of pregnant
previa than in the control group (Table 1). The distri-
women, gravidity, previous parity separately with total number of
bution according to different parity groups showed
previous cesarean sections, history of spontaneous or induced
that this was the consequence of significantly higher
729

---

Tuzoviæ et al: Obstetric Risk Factors and Placenta Previa
Croat Med J 2003;44:728-733
Table 1. Multiple potential risk factors for placenta previa (PP) development in placenta previa and control pregnancies*
No. (%) of women
Parameter
with PP (n=202)
without PP (n=1,004)
Chi-square
p
Crude odds ratio (CI)
Age (years):
<30
75 (37.1)
613 (61.1)
39.3
<0.001
2.63 (1.92-3.7)
>30
127 (62.9)
391 (38.9)
Multiparity:
yes
145 (71.6)
536 (53.4)
23.2
<0.01
2.2 (1.59-3.09)
no
57 (28.4)
468 (46.6)
Previous cesarean section:
yes
20 (9.8)
52 (5.2)
6.7
0.01
2.0 (1.17-3.44)
no
182 (90.2)
952 (94.8)
Previous abortions:
yes
92 (45.5)
231 (23.0)
43.6
<0.001
2.8 (2.04-3.83)
no
110 (54.5)
773 (77.0)
Previous placenta previa:
yes
0 (0)
1 (0.096)
0.2
0.832
–
no
202 (100)
1003 (99.9)
Smoking:
yes
33 (16.3)
257 (25.6)
7.9
0.007
0.58 (0.38-0.88)
no
169 (83.7)
747 (74.4)
Drug abuse:
yes
0 (0)
2 (0.2)
0.4
0.69
–
no
202 (100)
1002 (99.8)
Patologic presentation:
yes
40 (20.5)
39 (3.9)
73.4
<0.001
6.26 (3.84-10.0)
no
155 (79.5)
965 (96.1)
Uterine abnormalities:
yes
5 (2.5)
3 (0.3)
12.1
0.005
8.47 (1.75-44.5)
no
197 (97.5)
1001(99.7)
Newborn's sex:
male
114 (57.6)
520 (51.8)
2.2
0.136
1.26 (0.9-1.74)
female
84 (42.4)
484 (48.2)
Premature labor <37 weeks:
yes
83 (41.1)
48 (4.8)
161.4
<0.001
13.9 (9.1-21.2)
no
119 (58.8)
955 (95.2)
*Abbreviations: PP – placenta previa; 95% CI – 95% confidence interval.
percentage of women who delivered 2 or 3+ times in
Not a single case of illicit drug abuse was found in
the placenta previa group, whereas the frequency of
placenta previa group, whereas in the control group
women with 1 previous delivery was the same in both
0.2% of women had a history of illicit drug abuse
groups (Table 2). To control for the most important
(heroine and methadone). No women in placenta
confounding effect for women’s age to parity, we per-
previa group had the evidence of previous placenta
formed stratified analysis of women’s age according
previa. Slight, but statistically non-significant predom-
to different parity groups (multiparous vs primipa-
inance of male newborns was noticed in the placenta
rous). Adjusted and crude odds ratios proved that
previa group in comparison with control group
women’s age was a significant risk factor even after
(57.6% vs 51.8%, respectively). We found signifi-
controlling for high parity (Table 3). The effect of par-
cantly lower frequency of smokers among women
ity was further studied according to the mode of deliv-
with placenta previa than among controls (Table 1).
ery (vaginal or cesarean section) and history of previ-
The risk of having preterm delivery was almost
ous spontaneous or induced abortions. Women with
14-fold higher in the placenta previa group (41.1% vs
previous cesarean section had a 2-fold higher risk for
4.8%, p<0.001). Stratified analysis of neonatal out-
placenta previa development (Table 1). Among wo-
come data according to time of delivery (preterm vs
men with placenta previa, there was a significantly
term delivery) showed no significant difference in me-
higher frequency of those with 2 or more previous ce-
dian birth weight and height of preterm infants be-
sarean sections, whereas at the level of one previous
tween the two groups (Table 4). However, infants of
cesarean section no significant difference was found
mothers with placenta previa had significantly lower
(Table 2). The number of previous spontaneous/in-
first- (6 vs 10) and fifth-minute (8 vs 10) median Apgar
duced abortions was also significantly higher in the
scores than their controls. The difference between the
group of women with placenta previa (45.5% vs
groups in Apgar scores and cord blood pH value was
23.0%, p<0.001). Furthermore, the risk significantly
insignificant. Term infants of mothers with placenta
increased with increasing number of previous abor-
previa had significantly lower birth weight than
tions (chi-square for linear trend=73.23, p<0.001;
infants of the mothers in control group (3,300 g vs
Table 2). The rate of pathological fetal presentations
3,500 g, p<0.001).
was significantly higher in women with placenta
previa than in the control group (20.5 % vs 3.9%,
Discussion
p<0.001). Women with placenta previa had also
higher rate of different uterine abnormalities, such as
Placenta previa complicated 0.4% of all deliver-
uterine septum or myomatous uterus (Table 1). No as-
ies, which was within the range of 0.3-0.8% observed
sociation was found between placenta previa devel-
in other studies (3,5-7). In the past two decades, a sig-
opment and drug abuse during pregnancy (Table 1).
nificantly increasing trend in incidence of placenta
730

---

Tuzoviæ et al: Obstetric Risk Factors and Placenta Previa
Croat Med J 2003;44:728-733
Table 2. Age, gravidity, parity, previous cesarean section, and abortion distribution in women with placenta previa and control
pregnancies*
No. (%) of women
Parameter
with PP (n=202)
without PP (n=1,004)
p†
Crude odds ratio (95% CI)
Age‡:
<25
16 (7.9)
264 (26.3)
<0.001
1
25-29
59 (29.2)
349 (34.8)
0.128
2.79 (1.5-5.3)
30-34
75 (37.1)
254 (25.3)
<0.001
4.87 (2.7-9.1)
>34
52 (25.7)
137 (13.6)
<0.001
6.26 (3.3-12.5)
Gravidity§:
1
32 (15.8)
379 (37.7)
<0.001
1
2
41 (20.3)
322 (32.1)
<0.001
1.5 (0.9-2.5)
3
58 (28.7)
170 (16.9)
<0.001
4 (2.5-6.6)
4
31 (15.3)
68 (6.8)
<0.001
5.4 (2.9-10)
5+
40 (19.8)
65 (6.5)
<0.001
7.29 (4.2-12.5)
ParityII:
0
57 (28.2)
468 (46.6)
<0.001
1
1
70 (34.7)
344 (34.3)
0.915
1.67 (1.1-2.5)
2
42 (20.8)
125 (12.5)
0.008
2.76 (1.7-4.3)
3+
33 (16.3)
67 (6.7)
<0.001
4.0 (2.38-4)
Previous cesarean section:¶
0
182 (90.1)
952 (94.8)
<0.001
1
1
13 (6.4)
47 (4.7)
0.295
1.45 (0.73-2.9)
2+
7 (3.5)
5 (0.1)
<0.001
7.32 (2.1-25)
Previous abortions:**
0
99 (49.0)
763 (76)
<0.001
1
1
53 (26.2)
184 (18.3)
0.009
2.22 (1.5-3.2)
2
26 (12.9)
42 (4.2)
<0.001
4.77 (2.7-8.3)
3
15 (7.4)
13 (1.3)
<0.001
8.89 (3.9-20)
4+
9 (4.5)
2 (0.2)
<0.001
34.7(-)
*Abbreviations: PP – placenta previa; 95% CI – 95% confidence interval.
†Chi-square for the differences between women with or without placenta previa.
‡Chi-square=4.47, p=0.034.
§Chi-square=90.60, p<0.001.
||Chi-square=39.20, p<0.001.
¶Chi-square=6.95, p=0.009.
**Chi-square=73.23, p<0.001.
previa was reported in some studies. One of the larg-
Table 3. Stratified analysis of women’s age according to differ-
est meta-analysis (8), which compared incidences of
ent parity groups
placenta previa in different studies around the world,
No. (%) of women aged (years)
showed that in studies conducted between 1975 and
Parity
<30
>30
OR (95% CI)*
1984 the overall incidence was 0.36%. However, the
Primiparous:
studies conducted between 1985 and 1995 showed
placenta previa
32 (56.1)
25 (43.9)
2.513 (1.42-4.42)†
control
357 (76.3)
111 (23.7)
that the incidence increased to 0.48%. In our study, a
Multiparous:
slight increase in incidence was observed in years
placenta previa
43 (29.7)
102 (70.3)
2.169 (1.46-3.21)‡
1999 and 2000, although it was not significant. This
control
256 (47.8)
280 (52.2)
placenta previa
76 (37.1)
127 (62.9)
2.63 (1.92-3.7)‡
could be partly explained with an increasing rate of
control
613 (61.1)
391(38.9)
cesarean sections observed in our population during
*Abbreviations: OR – odds ratio; 95% CI – confidence interval. Adjusted OR for
the last decade. According to the recent reports, the
primiparity and multiparity, and crude OR for total sample.
†p=0.001.
incidence of cesarean section is in constant increase,
‡p<0.001.
reaching the incidence of more than 15% in tertiary
care centers (23).
Table 4. Neonatal outcome data in placenta previa and con-
Our study clearly demonstrated that women
trol pregnancies
older than 30 years had more than 2.5-fold higher risk
Pregancies (median, range)
for placenta previa development. The distribution ac-
Variable
placenta previa
controls
p*
cording to different age groups proved that this is the
Term delivery
consequence of significantly higher frequency of
(>37 weeks):
women older than 35 years in the study group and at
birth weight (g)
3,300 (1,450-4,750) 3,500 (2,000-5,040) 0.001
birth height (cm)
50 (42-57)
51(44-58)
0.067
the same time, significantly lower frequency of wo-
cord blood pH
7.23 (6.68-7.41)
7.27 (6.83-7.36)
0.054
men younger than 25. Because the group with pla-
Apgar score:
centa previa had significantly higher percentage of
first minute
10 (1-10)
10 (2-10)
0.102
multiparous women, and parity could have a con-
fifth minute
10 (3-10)
10 (4-10)
0.240
founding effect on risk associated with age, we ad-
Preterm delivery
(<37 weeks)
justed maternal age for different parity groups. How-
birth weight (g)
2,140 (940-3,750)
2,200 (870-3,100)
0.167
ever, this had no effect on adjustment. Other authors
birth height (cm)
45 (34-53)
45 (32-52)
0.085
reported a similar observation (6,15), although there
cord blood pH
7.22 (7.05-7.43)
7.3 (6.67-7.41)
0.241
were some who could not prove this association (11).
Apgar score:
fist minute
6 (1-10)
10 (2-10)
<0.001
The mechanism by which advanced maternal age im-
fifth minute
8 (3-10)
10 (5-10)
0.045
pairs normal placental development is not well un-
*Mann-Whitney U Test.
derstood. One of the possible explanations could be
731

---

Tuzoviæ et al: Obstetric Risk Factors and Placenta Previa
Croat Med J 2003;44:728-733
that the percentage of sclerotic changes on intramyo-
of male newborns. Some previous studies managed to
metrial arteries increases with increasing age, thereby
prove 2-4 fold higher risk for placenta previa in smok-
reducing blood supply to placenta (15). We further
ers (17,18), opposite to our findings, which further
studied the gravidity and parity distribution in both
supports the fact that different factors seem to be im-
groups of women in our study. We observed that
portant in different population. Drug abuse and smok-
women with placenta previa had significantly higher
ing during pregnancy was also insignificant in our
frequency of women with 3 or more previous preg-
population of pregnant women, which has not been
nancies. In the study of Abu Heija et al (11), the
reported by other authors. On of the possible explana-
gravidity became important after 5 or more previous
tion could be the fact that cocaine abuse, which
pregnancies. Parity distribution showed that differ-
shows the strongest association with placenta previa
ence between the groups was not significant for one
(19), was not observed in our population of pregnant
previous delivery, but it was significant for women
women. Contrary to some well-developed countries
having 2 or more previous deliveries. Some earlier
like USA where cocaine abuse is widely dispersed,
studies showed that parity became significant after 4
the most frequent illicit drug used among Croatian
or more previous pregnancies (11,24). Effect of parity
pregnant women are opiates. The role of previous pla-
was further studied separately for the effect of previ-
centa previa, which implies genetic base for placenta
ous cesarean sections. Our study confirmed that the
previa development, was not of importance in our
frequency of previous cesarean sections was signifi-
study. None of the women with placenta previa had a
cantly higher in placenta previa group than in the con-
history of placenta previa. However, there are some
trol group, which corresponded to 2-fold higher risk
indications from other studies that previous placenta
for placenta previa development. Several studies con-
previa could be a risk factor for its development in
ducted around the world confirmed a 2-5 fold in-
current pregnancy. Gorodeski et al (16) found recur-
creased risk for placenta previa development in wo-
rence risk for placenta previa to be 6 times higher
men with history of previous cesarean section (5,10-
than in general population of pregnant women, but
13,25). The risk determined in our study was at lower
they did not control for potential confounding factors.
border of significance. Whereas most of these studies
This important topic is yet to be clarified on a large
agree that one previous cesarean section significantly
population of pregnant women. Among other risk fac-
increases the risk of placenta previa development, the
tors, we found significant connection between pla-
impact of multiple repeated cesarean sections is more
centa previa and various uterine abnormalities, such
confusing. Some studies managed to prove that the
as uterine septum or myomas, which could act as me-
risk increased with increasing number of previous ce-
chanical barrier for normal placental implantation.
sarean sections (13,25), but others did not (11,12). In
Women with known uterine abnormalities had al-
our study, the effect of multiple repeated cesarean
most 8.5-fold higher risk, which was within the range
sections revealed that the frequency of placenta
observed in other studies (28). We further studied
previa increased more than 7-fold in women with 2
perinatal risk factors associated with pregnancies
previous cesarean sections. The exact mechanism of
complicated with placenta previa. In the last 10 years,
previous uterine scar predisposing to low implanta-
the advances in obstetric and neonatal care signifi-
tion of placenta is not well understood. It has been re-
cantly reduced perinatal mortality associated with
cently shown that uterine scar prevented migration of
placenta previa. However, preterm delivery still re-
placentas during the course of pregnancy toward the
mains one of the main problems (4,29). In our study,
more vascularized uterine fundus (26). This is sup-
41% of women with placenta previa delivered prema-
ported by the fact that the incidence of placenta
turely. Stratification according to different gestational
previa is significantly higher early in gestation than at
age groups showed that premature babies from moth-
term (26,27), and that its persistence mostly depends
ers with placenta previa had significantly lower first-
on type of placenta previa in the third trimester and
and fifth-minute Apgar scores. In term infants the only
on history of previous cesarean section (26). The role
significance was observed regarding birth height,
of previous abortions, either spontaneous or induced,
which was significantly lower in placenta previa
was proved to be important for placenta previa devel-
group. This could reflect significantly higher frequency
opment in our population of pregnant women. The
of intrauterine growth restriction among women with
percentage of previous abortions was significantly
placenta previa, although some authors were not able
higher among women with placenta previa, which
to prove this association (29,30).
yielded a risk of 2.75. The risk increased with increas-
Our retrospective study has some limitations.
ing number of previous abortions (1 or more). Our
Since it was a hospital-based study, its results are not
findings are in accordance with most studies dealing
applicable on the whole population of Croatian preg-
with this topic, although there are some studies that
nant women. Furthermore, although a large number
could not confirm this association (10,11). The mech-
of different parameters was tested on a large popula-
anism how previous abortions predispose to placenta
tion of pregnant women with placenta previa, the
previa development could be explained with possible
univariate model used in this study could not entirely
endometrial damage during repeated abortions, which
prevent the possible confounding influence of differ-
impedes successful fundal implantation of placenta.
ent variables on the amount of the risk associated with
Contrary to some previous studies where an asso-
each single variable. However we tried to attenuate
ciation between male sex of the newborn and pla-
this effect by using stratification method of data in
centa previa was observed (21,22), our study showed
variables of special interest (age, parity, and neonatal
only a slight, statistically insignificant predominance
outcome data).
732

---

Tuzoviæ et al: Obstetric Risk Factors and Placenta Previa
Croat Med J 2003;44:728-733
The results of our study indicate that knowing ob-
14 Gilliam M, Rosenberg D, Davis F. The likelihood of pla-
stetric factors predisposing women for placenta pre-
centa previa with greater number of cesarean deliveries
via development in our population is important for
and higher parity. Obstet Gynecol 2002;99:976-80.
choosing adequate preventive measures for these
15 Zhang J, Savitz DA. Maternal age and placenta previa: a
women. Physician should suspect placenta previa es-
population-based, case-control study. Am J Obstet
pecially if woman is over 34 years of age, has had 3 or
Gynecol 1993;168:641-5.
more previous pregnancies, parity of 2 and more, and
16 Gorodeski IG, Bahari CM. The effect of placenta previa
raising number of previous abortions and cesarean
localization upon maternal and fetal-neonatal outcome.
sections. These women should receive counseling as
J Perinat Med 1987;15:169-77.
soon as pregnancy is confirmed. This is especially im-
17 Chelmow D, Andrew DE, Baker ER. Maternal cigarette
portant in non-compliant women with possible poor
smoking
and
placenta
previa.
Obstet
Gynecol
1996;87(5 Pt 1):703-6.
antenatal care. Careful monitoring of these high-risk
pregnancies is of utmost importance, especially re-
18 Handler AS, Mason ED, Rosenberg DL, Davis FG. The
relationship between exposure during pregnancy to
garding careful ultrasonographic examination with
cigarette smoking and cocaine use and placenta previa.
exact placental location during the second trimester
Am J Obstet Gynecol 1994;170:884-9.
of pregnancy. Early recognition and proper monitor-
19 Macones GA, Sehdev HM, Parry S, Morgan MA, Berlin
ing of placenta previa could minimize the possibility
JA. The assotiation between maternal cocaine use and pla-
of poor outcome in sudden massive vaginal bleeding.
centa previa. Am J Obstet Gynecol 1997;177: 1097-100.
20 Francois K, Johnson JM, Harris C. Is placenta previa
more common in multiple gestations? Am J Obstet
References
Gynecol 2003;188:1226-7.
21 Demissie K, Breckenridge MB, Joseph L, Rhoads GG.
1 Cunningham FG, Gant NF, Leveno KL, Gilstrap III LC,
Placenta previa: preponderance of male sex at birth.
Hauth JC, Wenstrom KD. Williams obstetrics. 21st ed.
Am J Epidemiol 1999;149:824-30.
New York (NY): McGraw Hill; 2001.
22 Wen SW, Demissie K, Liu S, Marcoux S, Kramer MS.
2 Crane JM, Van den Hof MC, Dodds L, Armson BA,
Placenta previa and male sex at birth: results from a
Liston R. Maternal complications with placenta previa.
population-based study. Paediatr Perinat Epidemiol
Am J Perinatol 2000;17:101-5.
2000;14:300-4.
3 Love CD, Wallace EM. Pregnancies complicated by
23 Dra anèiæ A. Perinatal mortality in Republic of Croatia
placenta previa: what is appropriate management? Br J
in the year 2001. Gynecologia et Perinatologia 2002;
Obstet Gynaecol 1996;103:864-7.
11:1-13.
4 Crane JM, Van den Hof MC, Dodds L, Armson BA,
24 Babinszki A, Kerenyi T, Torok O, Grazi V, Lapinski RH,
Liston R. Neonatal outcomes with placenta previa.
Berkowitz RL. Perinatal outcome in grand and
Obstet Gynecol 1999;93:541-4.
great-grand multiparity: effects of parity on obstetric risk
5 Sheiner E, Shoham-Vardi I, Hallak M, Hershkowitz R,
factors. Am J Obstet Gynecol 1999;181:669-74.
Katz M, Mazor M. Placenta previa: obstetric risk factors
25 To WW, Leung WC. Placenta previa and previous ce-
and pregnancy outcome. J Matern Fetal Med 2001;
sarean section. Int J Gynaecol Obstet 1995;51:25-31.
10:414-9.
26 Dashe JS, McIntire DD, Ramus RM, Santos-Ramos R,
6 Frederiksen M, Glassenberg R, Stika C. Placenta previa:
Twickler DM. Persistence of placenta previa according
a 22-year analysis. Am J Obstet Gynecol 1999;180(6 Pt
to gestational age at ultrasound detection. Obstet
1):1432-7.
Gynecol 2002;99(5 Pt 1):692-7.
7 Iyasu S, Saftlas AK, Rowley DL, Koonin LM, Lawson
27 Taipale P, Hiilesmaa V, Ylostalo P. Transvaginal ultra-
HW, Atrash HK. The epidemiology of placenta previa
sonography at 18-23 weeks in predicting placenta previa
in the United States, 1979 through 1998. Am J Obstet
at delivery. Ultrasound Obstet Gynecol 1998; 12:422-5.
Gynecol 1993;168:1424-9.
28 Rasmussen S, Albrechtsen S, Dalaker K. Obstetric history
8 Ananth CV, Smulian JC, Vintzielos A. The association of
and the risk of placenta previa. Acta Obstet Gynecol
placenta previa with history of cesarean delivery and
Scand 2000;79:502-7.
abortion: a metaanalysis. Am J Obstet Gynecol 1997;
29 Ananth CV, Demissie K, Smulian JC, Vintzileos AM. Re-
177:1071-8.
lationship among placenta previa, fetal growth restric-
tion, and preterm delivery: a population-based study.
9 Ziadeh S, Yahaya A. Pregnancy outcome at age 40 and
Obstet Gynecol 2001;98:299-306.
older. Arch Gynecol Obstet 2001;265:30-3.
30 Wolf EJ, Mallozzi A, Rodis JF, Egan JF, Vintzileos AM,
10 Parazzini F, Dindelli M, Luchini L, La Rosa M, Potenza
Campbell WA. Placenta previa is not an independent
MT, Frigerio L, et al. Risk factors for placenta previa.
risk factor for a small for gestational age infant. Obstet
Placenta 1994;15:321-6.
Gynecol 1991;77:707-9.
11 Abu-Heija A, El-Jallad F, Ziadeh S. Placenta previa: ef-
fect of age, gravidity, parity and previous cesarean sec-
Received: August 8, 2003
tion. Gynecol Obstet Invest 1999;47:6-8.
Accepted: October 29, 2003
12 Hershkowitz R, Fraser D, Mazor M, Leiberman JR. One
or multiple previous cesarean sections are associated
Correspondence to:
with similar increased frequency of placenta previa. Eur
Lea Tuzoviæ
J Obstet Gynecol Reprod Biol 1995;62:185-8.
Department of Obstetrics and Gynecology
13 Hendricks MS, Chow YH, Bhagavath B, Singh K. Previ-
Zagreb University School of Medicine
ous cesarean section and abortion as risk factors for de-
Petrova 13
veloping placenta previa. J Obstet Gynaecol Res
10000 Zagreb, Croatia
1999;25:137-42.
ltuzovic@vip.hr
733

---