American Journal of Epidemiology Vol. 153, No.

4
Copyright © 2001 by The Johns Hopkins University School of Hygiene and Public Health Printed in U.S.A.
All rights reserved

Placental Abruption and Perinatal Mortality Ananth and Wilcox

Placental Abruption and Perinatal Mortality in the United States

Cande V. Ananth1 and Allen J. Wilcox2

Placental abruption is an uncommon obstetric complication associated with high perinatal mortality rates. The
authors explored the associations of abruption with fetal growth restriction, preterm delivery, and perinatal
survival. The study was based on 7,508,655 singleton births delivered in 1995 and 1996 in the United States.
Abruption was recorded in 6.5 per 1,000 births. Perinatal mortality was 119 per 1,000 births with abruption
compared with 8.2 per 1,000 among all other births. The high mortality with abruption was due, in part, to its
strong association with preterm delivery; 55% of the excess perinatal deaths with abruption were due to early
delivery. Furthermore, babies in the lowest centile of weight (<1% adjusted for gestational age) were almost nine
times as likely to be born with abruption than those in the heaviest (≥90%) birth weight centiles. This relative risk

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

progressively declined with higher birth weight centiles. After controlling for fetal growth restriction and early
delivery, the high risk of perinatal death associated with abruption persisted. Even babies born at 40 weeks of
gestation and birth weight of 3,500–3,999 g (where mortality was lowest) had a 25-fold higher mortality with
abruption. The link between fetal growth restriction and abruption suggests that the origins of abruption lie at
least in midpregnancy and perhaps even earlier. Am J Epidemiol 2001;153:332–7.

abruptio placentae; birth weight; fetal growth retardation; infant mortality

In normal pregnancies, placental separation occurs imme- Statistics and derived from the linked birth/infant death data
diately after birth, while in pregnancies complicated by sets. Perinatal mortality was defined to include stillbirths
abruption, the placenta begins to detach before birth (1). (occurring after 20 weeks of gestation) and neonatal deaths
This premature detachment commonly produces pain and (occurring up to 28 days after birth). Abruption was defined
vaginal bleeding, the clinical hallmarks of placental abrup- as the complete or partial separation of the placenta prior to
tion, and occurs in about 0.6–1.0 percent of pregnancies (2). the delivery of the fetus. The diagnosis of abruption is based
Maternal risks associated with abruption include massive on clinical presentation and examination of the delivered
blood loss, disseminated intravascular coagulopathy, renal placenta by the attendant at delivery. It is recorded on cur-
failure, and, less commonly, maternal death (1, 3). rent US birth certificates using a check-box (yes or no) (8),
Abruption is potentially disastrous to the fetus as well, with but grades and severity of abruption are not documented.
perinatal mortality as high as 60 percent (4–7). Surprisingly The diagnosis of abruption on vital statistics is subject to
little is known about the underlying etiology of abruption. potential underascertainment (9), and information on some
We present recent data from the United States to explore the important risk factors and confounders of abruption is
associations of abruption with fetal growth restriction, poorly recorded or unavailable on vital statistics data.
preterm delivery, and perinatal survival. For 95 percent of births, the gestational age assignment
on vital records was based on the date of the last menstrual
period. A clinical estimate was used when the date of the last
MATERIALS AND METHODS
menstrual period was missing or when the reported birth
Analysis was based on 7,508,655 singleton US births for weight was inconsistent with last menstrual period-based
1995 and 1996 assembled by the National Center for Health gestational age (10). Missing data on the month of the last
menstrual period were imputed (11). Growth restriction was
Received for publication January 18, 2000, and accepted for pub- assessed in centiles of weight (<1, 1–2, 3–4, 5–9, and there-
lication June 12, 2000. after in 10-centile increments derived from the complete US
1
Division of Epidemiology and Biostatistics, Department of data set) within strata of gestational age. Babies at or above
Obstetrics, Gynecology, and Reproductive Sciences, University of the 90th centile (who had the lowest risk of abruption)
Medicine and Dentistry of New Jersey, Robert Wood Johnson served as the comparison group.
Medical School, New Brunswick, NJ.
2
Epidemiology Branch, National Institute of Environmental Health One reason that babies born with abruption are at higher
Sciences, Research Triangle Park, NC. risk of mortality is that they are delivered prematurely. In
Correspondence to Dr. Cande V. Ananth, Division of Epidemiology order to estimate the portion of abruption mortality that
and Biostatistics, Department of Obstetrics, Gynecology, and might be explained by early delivery, we performed direct
Reproductive Sciences, University of Medicine and Dentistry of New
Jersey, Robert Wood Johnson Medical School, 125 Paterson Street, standardization by gestational age. Specifically, we applied
New Brunswick, NJ 08901-1977 (e-mail: [email protected]. the gestational age-specific mortality rates of nonabruption
EDU). births to the gestational age distribution of abruption births.

332
 Placental Abruption and Perinatal Mortality 333

This provided an estimate of total mortality for nonabruption deaths per 1,000 births) (table 1). At the optimal gestational
births had they had the same gestational age distribution of age of 40 weeks (when overall mortality was lowest), peri-
abruption births. The increase in mortality due to the change natal mortality was 19-fold greater for abruption than for
in the gestational age distribution was divided by the total nonabruption births (34.6 vs. 1.8 per 1,000 births, respec-
excess mortality observed among abruption births to esti- tively) (figure 3). Mortality with abruption was more likely
mate the fraction of deaths attributable to the excess of to occur before delivery than after (relative risk for stillbirth
preterm births among women with abruption. In the same was 18 compared with 10 for neonatal death).
fashion, we assessed the contribution of growth restriction to In order to separate fetal growth restriction from early
abruption mortality by applying weight-specific mortality delivery, we examined term babies alone (≥37 weeks). Both
rates to the birth weight distributions of abruption and weight distributions were strikingly Gaussian (figure 4) but
nonabruption births. In the case of birth weight, the analysis with a downward shift of the abruption births by about 250
was further stratified by gestational age to remove the effects g. (This shift was also seen at preterm gestational ages down
of early delivery from the effects of growth restriction. to 28 weeks.) Given this generalized growth restriction with
abruption, there was nonetheless a high risk of mortality
RESULTS among babies of normal fetal size (figure 5). The best sur-
vival (with or without abruption) was among babies weigh-
Placental abruption was recorded in 46,731 pregnancies, ing 3,500–3,999 g. Even among term babies with this opti-
an incidence of 6.2 per 1,000 pregnancies (table 1). mal weight, abruption was associated with a 25-fold
Pregnancies diagnosed with abruption were far more likely mortality risk.

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

to end in preterm delivery; 51 percent ended before 37 com- The strong association of abruption with preterm delivery
pleted weeks compared with 10.1 percent of all other births implies that some of the mortality associated with abruption
(figure 1). This substantial increase in early delivery may simply be due to the delivery of an immature fetus.
resulted in more small babies. Among babies born with Fifty-five percent of the excess perinatal deaths linked with
abruption, 46 percent weighed less than 2,500 g compared an abruption delivery were due to early gestational age
with 6.4 percent among all other pregnancies. alone. In contrast, the small fetal size at birth contributed
The smaller size of babies born with abruption, however, only 9 percent to the perinatal mortality associated with
was not due simply to preterm delivery. Babies born with abruption (despite the strong predictive value of poor fetal
abruption were also smaller at most gestational ages (figure growth as a risk factor for abruption).
2). Since fetal growth restriction precedes delivery, the risk
of abruption at delivery can be expressed as a function of DISCUSSION
fetal size. The relative risk for abruption was highest among
babies in the lowest 1 percent of weight at each gestational Placental abruption is an obstetric complication that poses
age and progressively declined with increasing weight cen- severe hazards to the pregnant woman and her fetus.
tiles (table 2). Although uncommon, abruption accounts for 12 percent of
The perinatal mortality rate associated with abruption was all perinatal deaths. Several studies have reported increased
119.2 per 1,000 births or nearly 12 percent. This rate was frequency of this condition among older women, multi-
nearly 15 times the mortality among other pregnancies (8.2 parous women, smokers, illicit substance users, and those

TABLE 1. Weight-specific perinatal mortality rates in relation to placental abruption, United States,
1995 and 1996

Birth Nonabruption Abruption

Relative
weight Total Total 95% CI†
PMR† PMR risk*
(g) births births

<500 22,817 702.0 1,497 692.1 1.0 0.9, 1.1

500–999 45,299 352.9 4,384 367.7 1.0 1.0, 1.1
1,000–1,499 41,304 122.7 3,795 178.4 1.5 1.4, 1.6
1,500–1,999 80,152 58.6 4,861 128.4 2.2 2.0, 2.4
2,000–2,499 286,461 16.8 6,936 83.6 5.0 4.6, 5.4
2,500–2,999 1,191,673 4.4 9,013 55.3 12.6 11.5, 13.7
3,000–3,499 2,794,679 1.8 9,419 32.4 18.0 16.0, 20.0
3,500–3,999 2,213,070 1.3 5,222 32.6 25.1 21.6, 29.3
4,000–4,499 663,893 1.4 1,382 41.2 28.7 22.1, 37.3
4,500–4,999 109,984 2.7 199 45.2 16.9 8.8, 32.3
≥5,000 12,592 12.2 23 87.0 7.2 1.9, 27.2

Total 7,461,924 8.2 46,731 119.2 14.5 14.2, 14.9

* Relative risk for perinatal mortality.
† CI, confidence interval; PMR, perinatal mortality rate (per 1,000 births).

Am J Epidemiol Vol. 153, No. 4, 2001

prgmea.com
 334 Ananth and Wilcox

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

FIGURE 1. Distribution of gestational age at delivery for pregnancies with and without placental abruption, United States, 1995 and 1996.

FIGURE 2. Distributions of mean birth weight for pregnancies with and without placental abruption, United States, 1995 and 1996.

Am J Epidemiol Vol. 153, No. 4, 2001

prgmea.com
 Placental Abruption and Perinatal Mortality 335

TABLE 2. Relative risk for placental abruption in relation to The incidence of abruption may be increasing with time.
centiles* of birth weight, United States, 1995 and 1996 Recorded abruptions in the United States increased 29 per-
cent between 1979 and 1987 (8.2–11.5 per 1,000 births)
Birth weight All births Term births (≥37 weeks) (12). Similar data from Norway (13) indicated a 31 percent
centile Relative Relative
(%) 95% CI† 95% CI increase in the incidence between 1967–1971 and
risk risk
1987–1991 (5.7–8.3 per 1,000 births). At least some of this
<1 8.8 8.2, 9.5 9.8 9.0, 10.5 increase may reflect improvements in ultrasound tech-
1–2 5.1 4.8, 5.4 4.6 4.3, 5.0 niques to detect mild or partial abruption. The prevalence of
3–4 4.2 4.0, 4.5 3.5 3.2, 3.8 some risk factors for abruption including cocaine and drug
5–9 3.6 3.4, 3.8 2.7 2.5, 2.9
use (7, 14) increased between 1980 and 1987 (15), which
10–19 3.1 3.0, 3.3 2.0 1.9, 2.1
20–29 2.8 2.7, 2.9 1.7 1.6, 1.8
may also have contributed to an increased incidence.
30–39 2.5 2.4, 2.7 1.5 1.4, 1.6 Multiple births due to assisted reproductive methods have
40–49 2.3 2.2, 2.5 1.4 1.3, 1.5 been increasing over time, with multiple fetuses raising the
50–59 2.1 2.0, 2.3 1.4 1.3, 1.5 risk for abruption (1, 5).
60–69 1.8 1.7, 1.9 1.2 1.1, 1.3 In searching for the causes of abruption, clinical studies
70–79 1.5 1.4, 1.6 1.2 1.2, 1.3 have sometimes focused on events near the time of delivery,
80–89 1.3 1.2, 1.3 1.1 1.0, 1.2 such as trauma (1). While abruption often appears to be an
≥90 1.0 Referent 1.0 Referent acute event, its association with poor fetal growth suggests
* Birth weight centiles were adjusted for gestational age. that the origins of abruption may lie at least in midpreg-

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

† CI, confidence interval. nancy, perhaps extending back to the earliest stages of preg-
nancy. The association with fetal growth restriction is so
strong that growth restriction in itself can serve as a marker
with hypertensive disorders. However, most cases of abrup- for abruption risk (table 2). The chronic processes underly-
tion occur with no known cause. ing abruption may also contribute to the risk of preterm

FIGURE 3. Gestational age-specific perinatal mortality rates (plotted on a logarithmic scale) for pregnancies with and without placental abrup-
tion, United States, 1995 and 1996.

Am J Epidemiol Vol. 153, No. 4, 2001

prgmea.com
 336 Ananth and Wilcox

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

FIGURE 4. Distribution of birth weight for term pregnancies (≥37 weeks) with and without placental abruption, United States, 1995 and 1996.

FIGURE 5. Birth weight-specific perinatal mortality rates (plotted on a logarithmic scale) in term pregnancies (≥37 weeks) with and without
placental abruption, United States, 1995 and 1996.

Am J Epidemiol Vol. 153, No. 4, 2001

prgmea.com
 Placental Abruption and Perinatal Mortality 337

delivery, which accounts for the majority of excess mortal- Anthony Vintzileos, and Jun Zhang for their comments, crit-
ity that accompanies abruption. icisms, and insightful discussions. They are also grateful to
The diagnosis of abruption on vital statistics data is sub- Susan Fosbre for secretarial assistance.
ject to misclassification (9) and is recorded with knowledge This paper was presented at the Society for Pediatric and
of the outcome of pregnancy. A pregnancy with severe hem- Perinatal Epidemiologic Research meeting in Baltimore,
orrhage may be more likely to be recorded as an abruption Maryland, June 1999.
if the baby dies. The present analysis also does not take into
account some known or suspected risk factors for abruption
including smoking, crack use, chronic hypertension,
preeclampsia, chorioamnionitis, and prolonged rupture of REFERENCES
membranes. These factors are reported poorly or not at all
on vital statistics. Some of these factors might contribute to 1. Cunningham GF, MacDonald PC, Gant NF, et al, eds. Chap 37.
the results noted here, but it is unlikely that these factors Obstetrical hemorrhage. In: Williams obstetrics. 19th ed.
could explain more than a small portion of the powerful Norwalk, CT: Appleton & Lange, 1993:819–51.
2. Ananth CV, Savitz DA, Williams MA. Placental abruption and
associations with fetal growth and preterm delivery. its association with hypertension and prolonged rupture of
Pathologic studies suggest that abruption is associated membranes: a methodologic review and meta-analysis. Obstet
with abnormal placental vasculature, thrombosis, and Gynecol 1996;88:309–18.
reduced placental perfusion. Genetic variations may predis- 3. Brame RG, Harbert GM Jr, McGaughey HS, et al. Maternal
risk in abruption. Obstet Gynecol 1968;31:224–7.
pose to these problems. A high prevalence of mutations in

Downloaded from http://aje.oxfordjournals.org/ by guest on January 5, 2012

4. Hibbard BM, Jeffcoate TNA. Abruptio placentae. Obstet
the genes that encode or regulate factor V Leiden, pro- Gynecol 1966;27:155–67.
thrombin, methylenetetrahydrofolate reductase, and homo- 5. Williams MA, Lieberman E, Mittendorf R, et al. Risk factors
cysteine has been seen among women with abruption com- for abruptio placentae. Am J Epidemiol 1991;134:965–72.
pared with other uncomplicated pregnancies (16, 17). These 6. Karegard M, Gennser G. Incidence and recurrence rate of
abruptio placentae in Sweden. Obstet Gynecol 1986;67:523–8.
genetic factors may also be contributing to the fetal growth 7. Ananth CV, Berkowitz TS, Savitz DA, et al. Placental abruption
restriction and preterm delivery associated with abruption, and adverse perinatal outcomes. JAMA 1999;282:1646–51.
perhaps through compromise of placental function. 8. Taffel SM, Ventura SJ, Gay GA. Revised US certificate of
Since abruption is apparently a disease of the placenta, birth—new opportunities for research on birth outcome. Birth
1989;16:188–93.
aspects of placental development (even preconception fac- 9. Buescher PA, Taylor KP, Davis MH, et al. The quality of the
tors) merit attention. Placental growth is primarily under the new birth certificate data: a validation study in North Carolina.
control of fetal genes inherited from the father (18). Am J Public Health 1993;83:1163–5.
Imprinted paternal alleles regulate the formation of the pla- 10. MacDorman MF, Atkinson JO. Infant mortality statistics from
centa and membranes surrounding the embryo, whereas the the linked birth/infant death data set—1995 period data. Mon
Vital Stat 1998;46(suppl 2):1–22.
development of the embryo itself requires contribution from 11. Taffel S, Johnson D, Heuser R. A method of imputing length of
the maternally derived alleles. A search for genetic suscepti- gestation on birth certificates. Vital Health Stat 2 1982;93:1–11.
bility to abruption should include a consideration for pater- 12. Saftlas AF, Olson DR, Atrash HK, et al. National trends in the
nally imprinted genes. incidence of abruptio placentae, 1979–1987. Obstet Gynecol
1991;78:1081–6.
13. Rasmussen S, Irgens LM, Bergsjo P, et al. The occurrence of
placental abruption in Norway 1967–1991. Acta Obstet Gynecol
Scand 1996;75:222–8.
14. Shiono PH, Klebanoff MA, Nugent RP, et al. The impact of
cocaine and marijuana use on low birth weight and preterm birth:
ACKNOWLEDGMENTS a multicenter study. Am J Obstet Gynecol 1995;172:19–27.
15. Abelson HI, Miller JD. A decade of trends in cocaine use in the
household population. NIDA Res Monogr 1985;61:35–49.
At the time of this study, Dr. Ananth was supported, in 16. Kupferminc MJ, Eldor A, Steinman N, et al. Increased fre-
part, through grant 029553 from the Robert Wood Johnson quency of genetic thrombophilia in women with complications
Foundation, New Jersey, awarded to the Center for Perinatal of pregnancy. N Engl J Med 1999;340:9–13.
Health Initiatives. 17. de Vries JIP, Dekker GA, Huijgens PC, et al. Hyperhomo-
cysteinaemia and protein S deficiency in complicated pregnan-
The authors are indebted to Drs. Kitaw Demissie, Lorentz cies. Br J Obstet Gynaecol 1997;104:1248–54.
Irgens, Robert Knuppel, Dawn Misra, George Rhoads, 18. Bartolomei MS, Tilghman SM. Genomic imprinting in mam-
Andrew Rowland, Amy Sayle, John Smulian, John Thorp, mals. Annu Rev Genet 1997;31:493–525.

Am J Epidemiol Vol. 153, No. 4, 2001