THE YALE JOURNAL OF BIOLOGY AND MEDICINE 62 (1989), 241-251

Epidemiology of Spontaneous Premature Rupture

of Membranes: Factors in Pre-Term Births
HERBERT C. MILLER, M.D.,a AND JAMES F. JEKEL, M.D., M.P.H.b

aDepartment of Pediatrics, University of Kansas Medical Center, Kansas City,

Kansas; bDepartment of Epidemiology and Public Health, Yale University
School of Medicine, New Haven, Connecticut
Received May 17, 1989
The frequency of spontaneous premature rupture of membranes (PROM) was determined in
the pregnancies of 1,848 white mothers and their singleton infants, born at the University of
Kansas Medical Center between April 1975 and April 1978. The frequency of PROM increased
significantly from a low of 34/707 (4.8 percent) among low-risk mothers, to 40/444 (9.0 percent)
among mothers smoking one to 60 cigarettes a day, to 21/204 (10.3 percent) among mothers with
multiple adverse maternal practices, and to 12/46 (26 percent) among mothers with selected
complications of their pregnancies.
The proportion of low birth weight (LBW) (<2,500 g) pre-term infants born to PROM
mothers increased among the risk factor groups in a similar manner, from a low of 2/34 (6
percent) in low-risk pregnancies to 8/40 (20 percent) among mothers smoking one to 60
cigarettes a day, to 7/21 (33 percent) among mothers with multiple adverse practices, and to 7/12
(58 percent) among mothers with selected complications of pregnancy.
The increased incidence of low birth weight pre-term infants born to mothers with PROM was
associated with evidence of growth retardation among full-term infants in the high-risk groups.
This finding was manifested by reductions in mean birth weights of full-term infants born to
high-risk mothers but not observed in full-term infants born to low-risk mothers. The attained
growth at birth of low birth weight pre-term infants could not be determined, because appropriate
birth weight standards for pre-term infants born to mothers with low-risk pregnancies are not
available.
These results suggest that growth retardation in fetuses increased the probability of the
mothers having PROM prior to the onset of labor, and, if PROM did occur, of having a premature
delivery. We hypothesize that the tensile strength of the amnion and chorion is diminished by the
same conditions that retard fetal growth, and that this reduction in strength of the fetal
membranes contributes to premature rupture of membranes and pre-term delivery.

INTRODUCTION
Modern technology has improved the survival rate of pre-term infants under 1,500 g
at birth at a heavy financial cost and at a considerable risk of long-term deficits among
survivors. Comparatively little progress has been made in preventing pre-term births.
Guidelines for Perinatal Care, second edition, states that the main cause of perinatal
deaths among small pre-term infants is spontaneous premature rupture of membranes
(PROM), and that the cause(s) of PROM remain unknown [1]. A variety of causes
have been suggested: infection, trauma from pelvic examinations or coitus, polyhydramnios, weakened membranes, or incompetent cervix [2]. Changes in barometric
241
Abbreviations: C-H length: crown-heel length LBW: low birth weight PROM: premature rupture of
membranes SES: socioeconomic status
Address reprint requests to: Mr. Jack Jones, Dept. of Pediatrics, U. of Kansas Medical Center, Kansas
City, KS 66103
Copyright e 1989 by The Yale Journal of Biology and Medicine, Inc.
All rights of reproduction in any form reserved.

242

MILLER AND JEKEL

pressure or some "circadian factor" have also been implicated [3]. Relatively little
effort has been made to determine how unfavorable maternal behavior, such as
cigarette smoking or abuse of alcohol and addicting drugs, might be associated with
PROM. In an epidemiologic study published over a decade ago, a threefold increase in
pre-term births under 34 weeks' gestation was observed among cigarette smokers with
PROM as compared to non-smokers with PROM; however, smokers and non-smokers
in the study were not clearly defined [4]. Smokers and non-smokers often adopt other
unfavorable practices, such as gaining too little weight during pregnancy, being too
young, going without any professional prenatal care, or abusing alcohol and other
drugs. The present study was undertaken to determine outcomes in pregnancies that
were complicated by PROM combined with a variety of risk factors. Comparisons
were made between outcomes of pregnancies complicated by PROM combined with
other risk factors and the outcomes of pregnancies complicated by other risk factors
except PROM. Particular emphasis was placed on outcomes of pregnancies complicated by maternal cigarette smoking during pregnancy, with and without PROM.
PROM did occur among mothers with low-risk pregnancies and with medical or
obstetric complications of pregnancy, and the outcomes of these pregnancies were also
investigated.

METHODS
The study group consisted of 1,848 white mothers who delivered singleton infants at
the University of Kansas Medical Center from April 1975 to April 1978. These were
the same mothers and infants who appeared in our previous studies of low birth weight
(LBW) [5-7]. Only white mothers were used because the group of black mothers was
much smaller, and the socioeconomic status (SES) was very different.
The mothers were drawn from a wide range of socioeconomic circumstances,
ranging from professional and executive levels to unemployed and welfare groups,
from those who were college-educated to those with only primary school education, and
from both married and unmarried mothers.
The methods of data collection on mothers and infants also were the same as those
described in our previous studies. Briefly, these methods included the classification of
mothers into low- and high-risk groups, using a special protocol. Four categories of
high risk were established (Table 1): environmental conditions, fetal conditions,
medical and obstetric complications of pregnancy, and adverse maternal practices.
Mothers with high-risk pregnancies were assigned to one of the four categories as
follows: those with environmental conditions were assigned to category 1 regardless of
other risks present (no mothers in the present study were assigned to category 1);
mothers were assigned to category 2 (fetal conditions) even if medical and obstetric
complications and/or adverse practices were present; they were assigned to category 3
(medical and obstetric complications) even if adverse practices were present; and they
were assigned to category 4 (adverse maternal practices) if they had adverse maternal
practices but none of the risks in categories 1, 2, or 3 during their pregnancies. If the
pregnancies had none of the risk factors in Table 1, they were assigned to the
"low-risk" group.
Two subsets of the mothers with medical/obstetrical complications were created:
those with medical/metabolic complications, and those with uterine or placental
abnormalities. Also, for the analysis here, the group of "adverse practices" was divided
into subsets: those mothers who smoked cigarettes throughout pregnancy but had no

PREMATURE RUPTURE OF MEMBRANES

243

TABLE 1
Risk Conditions Associated with Low Birth Weight*
1. Environmental Factors
High altitude
Exposure to specific toxic agents
2. Fetal Factors
Multiple births

Congenital malformations
Fetal infections
Inborn errors of metabolism
Maternal-fetal blood incompatibility, producing disease in the fetus
3. Medical Complications of Pregnency
Toxemia of pregnancy
Chronic hypertension
Severe vaginal bleeding in third trimester
Abnormally high glucose tolerance curves
Malformations of placenta, cord, or uterus
Anemia: hemoglobin level <10 g/dL
Severe chronic maternal disease
Leukemia
Malignant solid tumors
Large ovarian cysts or uterine fibroids
Continuous maternal medication with corticosteroids or immunosuppressive, teratogenic, or fetalgrowth-retarding drugs
Polyhydramnios or oligohydramnios
4. Adverse Maternal Practices
Cigarette smoking during any part of pregnancy
Low weight gain in trimesters 2 and 3a
Low weight for height at conceptionb
Delivery < 17 years of age
Delivery >34 years of age
No professional prenatal care
Use of addicting drugs or consumption of large amounts of alcohol during pregnancy
'Low weight gain, <228 g per week in trimesters 2 and 3
bLow weight, >15 percent below normal Sargent's table for young women (J Nutr 13:318, 1963)
*PROM was considered a special risk that occurred in each of the four high-risk categories and also in
low-risk pregnancies.
Reprinted from Am J Dis Child 137:324, 1983, with permission;
Copyright 1983, American Medical Association.

other risk factors, and those mothers who had multiple adverse practices, usually
smoking plus one other.
The term "adverse practices" was used for risk factors that were under the control of
the mothers, such as the age at which they became pregnant, their weight or weight
gain, and behaviors such as cigarette smoking. The diagnoses of PROM and of medical
and obstetric complications of pregnancy were made by the attending physician in each
case and were obtained from the mothers' obstetrical records. All deliveries were made
by senior staff and resident staff physicians of the University of Kansas Medical
Center. The pediatric investigator (HCM) examined and measured all infants,
reviewed the mothers' medical and obstetric records, and interviewed all mothers
between delivery and discharge from the Medical Center to obtain the history of
cigarette smoking during pregnancy and to clarify information pertaining to any data
included in this study.

244

MILLER AND JEKEL

The diagnosis of pre-term and full-term births was made from a composite of
observations: obstetricians' examinations of the mothers, pediatric examination of the
infants, calculated gestational ages, and Dubowitz scores [9], with the pediatric
investigator (HCM) making the final determination. Whenever possible, gestational
age was calculated from the first day of the mother's last menstrual period. Gestational
ages were accepted only if calculated and estimated gestational ages were the same.
Otherwise, infants were classified only as pre-term or full-term infants. The diagnosis
of prenatal retardation in growth of fetal soft tissues (low ponderal index) and of
growth in crown-heel length was made using the Kansas City fetal growth tables [8].
Mothers in families paying full hospital costs and physicians' fees were classified in
socioeconomic group I if the mother had completed more than 12 years of school and in
group II if she had completed 12 or fewer years of school. Mothers were placed in
socioeconomic group III if their hospital and physician's charges were discounted or
limited to third-party payments; both physicians' charges and hospital charges were
greater than ordinarily would be reimbursed by third-party payors, and their forgiveness of these charges suggested limited financial resources. Mothers were assigned to
socioeconomic group IV if the head of household was on welfare or unemployed.

RESULTS
There were 1,848 white mothers of singleton infants in this study, of whom 1,141
had high-risk factors and 707 were low-risk mothers. The 1,141 high-risk mothers
included 21 mothers with fetal conditions in their pregnancies (category 2, Table 1),
267 mothers with medical or obstetric complications of pregnancy (category 3, Table
1), and 853 mothers with adverse maternal practices (category 4, Table 1). The 707
low-risk mothers had none of the risk factors in Table 1 during their pregnancies.
PROM occurred in 34/707 (4.8 percent) of low-risk mothers; in 3/21 (14 percent) of
mothers with fetal conditions; in 22/267 (8.2 percent) of mothers with medical or
obstetric complications of pregnancy; and in 78/853 (9.1 percent) of mothers with
adverse maternal practices. A total of 137 mothers had PROM.
Data in Table 2 provide information on frequencies of PROM and rates of LBW
pre-term and full-term infants born to mothers with different types of risk factor in
their pregnancies. The data show that PROM occurred earlier in the third trimester,
and with greater frequency, in most categories of high-risk pregnancy, and that
PROM occurred later in the third trimester, and with less frequency, in mothers with
low-risk pregnancies and in pregnancies with only medical/metabolic complications.
With minor exceptions, mothers with the risk factors shown in Table 1 had higher rates
of pre-term infants, low birth weight infants (especially very low birth weight infants),
and PROM.
Data in Table 3 provide information on the mean birth weights of full-term infants
at successive weeks of gestation who were born to mothers in Table 2, grouped
according to the types of risks in their pregnancies. Note that in comparable risk factor
categories, the numbers are smaller in Table 3, because only those deliveries for which
there was agreement between methods of determining gestational ages were used.
These data show that in the various high-risk groups, except for medical/metabolic
problems, the mean birth weights of full-term infants were reduced compared to those
from low-risk pregnancies, whether or not there was PROM, although it was more
marked if there was. This part suggests that there is a reduction in prenatal growth in

245

PREMATURE RUPTURE OF MEMBRANES

TABLE 2
Incidence of Low Birth Weight White Infants, Pre-Term and Full-Term, Born to Mothers of
Different Risk Groups, with and without Spontaneous Premature Rupture of Membranes (PROM)

Pre-Term Infants
Total No.
of Mothers
No.
a. Low-Risk Mothers
No PROM
673
34
PROM

b.

c.

d.

e.

f.

<2,000 g
No.

1
(95.2)
(4.8)
0
Mothers with Uterine/Placental
Complications of Pregnancya
5
34
NoPROM
(73.9)
12
PROM
(26.1)
3
Mothers with Medical/Metabolic
Complications of Pregnancyb
No PROM
1
55
(98.2)
1
PROM
(1.8)
0
Mothers with Medical/Metabolic
Complications and Adverse Practices'
No PROM
66
3
(90.4)
7
PROM
0
(9.6)
Mothers Who Smoked
Throughout Pregnency
4
No PROM
404 (91.0)
6
PROM
40
(9.0)
Mothers with Multiple
Adverse Practices
1
No PROM
183
(89.7)
PROM
21
5
(10.3)

Full-Term Infants

2,000-2,499 g
No.

2,000-2,499 g
No.

2,500+ g

No.

(0.1)

8
2

(1.2)
(5.9)

4
0

(0.5)

660
32

(98.1)
(94.1)

(14.7)
(25.0)

1
4

(2.9)
(33.3)

4
2

(11.8)
(16.7)

24
3

(70.6)
(25.0)

(1.8)

1
0

(1.8)

0
0

53
1

(96.4)
(100.0)

(4.6)

2
0

(3.0)

5
0

(7.6)

56
7

(84.8)
(100.0)

(1.0)
(15.0)

11
2

(2.7)
(5.0)

5
2

(1.2)
(5.0)

384
30

(95.0)
(75.0)

(0.5)
(23.8)

5
2

(2.7)
(9.5)

10
2

(5.5)
(9.5)

167
12

(91.3)
(57.1)

aUterine/placental complications of pregnancy in part b: infections (n = 14), placenta abruptio (n = 11),

placenta accreta (n = 5), battledore placenta (n = 4), polyhydramnios (n = 4), bicornuate uterus (n = 3),
multiple placental infarcts (n = 4), uterine fibroids (n = 1)
bMedical/metabolic complications of pregnancy in part c: pre-eclampsia (n = 35), hypertension (n = 7),
diabetes mellitus (n = 8), gestational diabetes (n = 6)
cMedical/metabolic complications of pregnancy in part d: pre-eclampsia (n = 37), hypertension (n = 25),
diabetes mellitus (n = 2), gestational diabetes (n = 9)

these groups, and we hypothesize that this growth reduction may play a role in the
etiology of PROM.
Low-Risk Pregnancies
The frequency of PROM in low-risk pregnancies in Table 2 was low, 34/707 (4.8
percent). Second, the incidence of pre-term infants born to low-risk mothers with
PROM was 2/34 (5.9 percent), which was low compared to the incidence of 22
pre-term infants among 81 high-risk mothers with PROM (27 percent), and the birth
weights of these two were over 2,000 g; however, the incidence of pre-term infants born
to low-risk mothers without PROM was even lower, 9/673 (1.3 percent). Note in
Table 3 that the mean birth weights of full-term infants born to low-risk mothers with
and without PROM were comparable and were among the highest in this study.

246

MILLER AND JEKEL

TABLE 3
Mean Weights (g) of White, Full-Term Infants Born to Mothers of Different Risk Groups, with and
without PROM, by Gestational Age
Gestational Age (Weeks)a
37

38

39

40

41

42+

Average
(No.)

Average

Average
(No.)

Average
(No.)

Average
(No.)

Average
(No.)

Total

PROM
Status

Infants
No.

(No.)

a. Low-Risk Mothers
No PROM
490a
2,945 (27) 3,286 (32) 3,424 (99) 3,545 (168)
29
PROM
2,961 (4) 3,214 (4) 3,458 (9) 3,781 (8)
b. Mothers with Uterine/Placental Complications of Pregnancy
No PROM
26
3,105 (1) 2,913 (3) 2,849 (5) 2,905 (6)
Mean of all ages = 2,700
5
PROM
c. Mothers with Medical/Metabolic Complications of Pregnancy
No PROM
38
3,120 (7) 3,316 (6) 3,544 (11)
PROM
0
d. Mothers with Medical/Metabolic Complications and Adverse Practices
51
2,736 (2) 3,043 (8) 3,250 (17) 3,082 (14)
No PROM
e. Mothers Who Smoked Throughout Pregnancy
314
No PROM
3,035 (13) 3,073 (34) 3,224 (64) 3,352 (101)
Average no.

cigarettes
smoked
PROM

24

14
2,920 (2)

16
3,179 (3)

3,695 (111) 3,652 (53)

3,010 (1)
3,607 (3)
3,501 (3)

2,913 (8)b

3,748 (10)

4,100 (4)

3,847 (5)

3,139 (5)

3,386 (70)

3,596 (32)

18

17

17

15

3,070 (6)

3,188 (5)

3,134 (5)

3,348 (3)

17

23

Average no.

cigarettes
12
27
15
15
smoked
f. Mothers with Multiple Adverse Practices
140
No PROM
2,856 (15) 2,838 (20) 3,156 (30) 3,242 (33)
10
PROM
2,745 (3) 2,593 (1) 3,038 (3) 3,659 (3)

3,409 (30)

3,414 (12)

aThese numbers are smaller than those in Table 2 because the dates of gestation could not be reconciled
between methods; there was complete agreement on dates for those pregnancies shown in Table 3.
bThese infants were full-term but exact dates were uncertain.

In Tables 2 and 3, comparison of rates and birth weights for low- and high-risk
mothers illustrates that in high-risk mothers: (1) the frequency of PROM was
increased compared to that for low-risk mothers; (2) the occurrence of PROM before
37 weeks of pregnancy was increased in mothers with high-risk factors compared to
low-risk mothers; and (3) the mean weights of full-term infants of mothers who smoked
or had multiple adverse practices were lower than those from low-risk pregnancies,
suggesting a degree of prenatal growth retardation. Also, the average birth weights by
week of gestation were generally lower within a risk group for the infants of those
mothers who had PROM compared to those who did not.
Mothers with Uteroplacental Complications
The highest frequency of PROM, the highest incidence of pre-term infants, and the
lowest mean birth weights of full-term infants occurred in a group of mothers with
uteroplacental complications of pregnancy (part b of Tables 2 and 3). Among the 46
mothers with uteroplacental complications, 12 (26 percent) had PROM, and seven (58

PREMATURE RUPTURE OF MEMBRANES

247

percent) of these 12 mothers with PROM had pre-term infants under 2,500 g, the
highest incidence in this study.
The incidence of pre-term infants born to 34 mothers with uterine/placental
complications but no PROM was six (18 percent) compared to 7/12 (58 percent)
among similar mothers with PROM (Fisher exact test, two-tailed: p = 0.021). The
mean birth weight of the five full-term infants born to mothers with uterine/placental
complications and PROM was 2,700 g, which was below the mean birth weight of all
the 26 full-term infants born to similar mothers with no PROM.
Medical or Obstetrical Complications
Mothers with medical or obstetric complications of pregnancy had the widest
variations in the incidence of PROM, the time of pregnancy in which PROM occurred,
incidences of pre-term births, and mean birth weights of full-term infants. Data in part
c of Tables 2 and 3 show that mothers with medical/metabolic complications of
pregnancy (pre-eclampsia, hypertension of more than a month's duration preceding
delivery, diabetes mellitus, or gestational diabetes) had the lowest frequencies of
PROM. There were 56 mothers who had a single medical/metabolic complication and
no other known risk factors of the type in Table 1, of whom only one (2 percent) had
PROM, and her infant was full-term. In the group of 55 mothers without PROM,
there were two mothers (3.6 percent) who had pre-eclampsia and pre-term infants
after induction of labor or Cesarean section without labor.
There were 73 mothers with medical-metabolic complications who also had other
risk factors, usually adverse practices-mainly cigarette smoking; seven (9.6 percent)
of these had PROM, but all seven of their infants were full-term (Table 2, part d). Five
of the 66 mothers with metabolic conditions without PROM had pre-eclampsia and
pre-term births (7.6 percent); only two of the five mothers of pre-term infants had
spontaneous onsets of labor.
In Table 3, the mean birth weights of infants born to mothers with complications of
pregnancy combined with other risks (part d) were usually below the mean birth
weights of infants born to mothers with a single complication of pregnancy and no
other risks. The mean birth weights of infants born to mothers with medical/metabolic
complications (part c) were similar to or greater than the mean birth weights of infants
born to low-risk mothers (part a).

Smoking During Pregnancies

Data in Table 2, part e, show that in the 444 mothers who reported smoking one to 60
cigarettes a day throughout pregnancy, but who had none of the other risks in Table 1,
the frequency of PROM was 40/444 (9 percent), which was significantly higher than
the frequency of 34/707 (4.8 percent) in low-risk pregnancies (X2 = 8.0, p = 0.0047).
The incidence rate of pre-term infants born to smoking mothers who had PROM was
8/40 (20 percent), more than three times as high as the incidence rate of pre-term
births in low-risk mothers with PROM, and five times as high as the incidence rate of
pre-term infants born to smoking mothers who had no PROM, 15/404 (3.7 percent). It
is important to note that in Table 3 the average number of cigarettes smoked per day
was about the same among mothers with and without PROM and that the mean birth
weights of full-term infants born to cigarette smokers was lower among mothers with
PROM than among mothers who had no PROM (except at 38 weeks).

248

MILLER AND JEKEL

Multiple Adverse Practices

Part f of Tables 2 and 3 show data on mothers with multiple adverse maternal
practices in their pregnancies and none of the other risks in Table 1. These data
illustrate a greater increase in frequency of PROM, a higher incidence of pre-term
infants, and a greater decrease in mean birth weight of full-term infants among
mothers with PROM than were observed among mothers who had a single adverse
practice: smoking cigarettes. It is interesting that among the 204 mothers with
multiple adverse practices, 196 (96 percent) smoked cigarettes as one of their multiple
adverse practices. The main additional practices were, in descending order, low weight
gains, being underweight for height at conception (for definitions, refer to Table 1),
and being under 17 or over 34 years of age at time of delivery. Abuse of alcohol or
addicting drugs was limited to one mother who was on heroin and methadone.
The frequency of PROM among the 204 mothers with multiple adverse practices, as
shown in part f of Table 2, was 21 (10.3 percent), twice the frequency of 34/707 (4.8
percent) in low-risk mothers. The incidence of pre-term births among the 21 mothers
with PROM was seven (33 percent), compared to an incidence of 6/183 (3.3 percent)
among mothers with multiple adverse practices but no PROM.
The mean birth weights of the 140 full-term infants born to mothers with multiple
adverse practices but no PROM were substantially lower than the mean birth weights
of full-term infants in Table 3 born to mothers who had a single adverse maternal
practice, except at 41 weeks, at which time they were approximately the same.
Prenatal Growth Retardation
The types of prenatal growth retardation among full-term infants born to mothers
with PROM are shown in Table 4. The two main types of prenatal growth retardation
were fetal skeletal growth retardation (short crown-heel length for gestational age, sex,
and race of infant) and fetal malnutrition (low ponderal index with a satisfactory
crown-heel length for gestational age).
Neither of the two main types of growth retardation appeared among the 32
full-term infants born to low-risk mothers with PROM. As shown in Table 4, in the
group of 78 full-term infants born to high-risk mothers with PROM there were 15
infants who at birth had fetal skeletal growth retardation, and there were five others
who had low ponderal indices with birth weights below the fifth percentiles for their
sex, gestational ages, and race. Two other infants had birth weights below the fifth
percentiles, and their crown-heel lengths were below the tenth percentiles. Two infants
had a small head circumference only. Altogether, in the group of 78 high-risk mothers
with PROM there were 15 full-term infants with short crown-heel lengths, five others
with low ponderal indices, and five others still with other evidence of prenatal growth
retardation. The first group strongly suggests prenatal growth retardation beginning
fairly early in pregnancy. The five infants with low ponderal indices and satisfactory
crown-heel lengths for their gestational ages had growth retardation of soft tissues,
probably only in late pregnancy. The overall difference in the incidence of prenatal
growth retardation in full-term infants born to high- and low-risk mothers with PROM
was great: 32 percent in the high-risk group vs. 6 percent in the low-risk group
(p = .004).
The important question of whether or not skeletal and/or soft tissue growth
retardation occurred prenatally in pre-term infants born to 673 low-risk mothers
(n = 9) is more difficult to answer, because so few pre-term infants under 2,500 g were

249

PREMATURE RUPTURE OF MEMBRANES

TABLE 4
Types of Prenatal Growth Retardation in Full-Term Infants Born to High- and Low-Risk Mothers
with Spontaneous Premature Rupture of Membranes (PROM)
n = 78

Low-Risk
n = 32

8
3
5
2
1
2
25

0
0
0
0
0
2
0
2

High-Risk
Measurements Below Fifth Percentilesa

Crown-heel lengths
Crown-heel lengths, low ponderal indices and birth weight
Crown-heel lengths and birth weight
Crown-heel lengths only
Low ponderal indices and birth weight
Birth weight onlyb
Birth weight and head circumference
Head circumference only
Totalc

aBased on Kansas City fetal growth tables [8]

bThese two infants had crown-heel lengths and ponderal indices below the tenth percentiles.

cChi-square overall proportion of growth retardation among high- vs. low-risk group mothers = 8.16 on
one degree of freedom; p = 0.004.

born to them. Figure 1 shows the crown-heel lengths by week of gestation for pre-term
births to low-risk mothers with no PROM and to two groups of high-risk mothers with
PROM; this graph was done only for infants whose calculated and gestational ages
were the same. In Fig. 2, a similar plot was done for the same three risk groups, except
that none of the infants in Fig. 2 were from pregnancies with PROM. In Fig. 1, the
crown-heel lengths for low-risk pre-term births appear to be higher than the average of
the two high-risk groups. The same general pattern appears to hold in Fig. 2, where
none of the infants were from pregnancies with PROM, but the differences between
the groups appear less striking.

48

Crrown Heel Length (Cm.)

47
46
46
44
43
42
41
40

39
38

+
+

+
*

27

28

29

30

36

37

31

32

33

34

36

Week of Gestation
*
*

Adv. Pract, PROM

ComplIcatlons, PROM

Risk Group
+ Low Risk, No PROM

FIG. 1. Pre-term deliveries with PROM: Crown-heel length by week of gestation.

250

49
48

MILLER AND JEKEL

Crown-Heel Length (Cm.)

-+*

47

46 -+
46
44
43
42
41
40
39
38
27

*
*

l
34

36

l
28

*
l
29

l
30

l
31

l
32

l
33

l
36

37

Week of Gestation
* Adv. Pract, No PROM
* Comps., No PROM

Low Risk, No PROM

FIG. 2. Pre-term deliveries without PROM: Crown-heel length by week of gestation.

Table 5 shows that, within comparable risk groups, the frequency of PROM did not
vary greatly between different socioeconomic (SES) levels of heads of households. It is
interesting, however, that in both risk factor categories, PROM occurred more
frequently in the upper two SES groups than in the two lower SES groups.
DISCUSSION
The analysis in this and one other paper [10] suggest that the causal pathway is
twofold: (1) risk factors directly increase the incidence of low birth weight, and (2) risk
factors increase the incidence of PROM, which, in turn, further increases the incidence
of low birth weight. In the presence of other risk factors, PROM appeared to have
serious consequences, possibly because it produced a premature delivery imposed on
growth retardation. Data from Table 2 for low-risk pregnancies suggests, however, an
absence of prenatal growth retardation in full-term infants born to' low-risk mothers,
whether or not PROM occurred, and the PROM did not appear to have serious
consequences.
The present results do not determine the causes of PROM, but they do suggest an
hypothesis: the conditions causing fetal growth retardation also impair the development of normal tensile strength of the amnion and chorion, which results in more
premature rupture of these membranes. The data in this study suggest that an analysis
of the strength and histochemical composition of fetal membranes of pregnancies with
PROM and a control group without PROM might reveal important differences. The
hypothesis is supported by the observation that the frequency of PROM was
significantly increased among mothers of undergrown infants at birth compared to the
frequency of PROM among mothers of newborn infants who attained adequate levels
of growth. The frequency of PROM was not only increased significantly among
mothers of undergrown newborn infants, but PROM also occurred much earlier in
pregnancy in undergrown infants. The incidence of pre-term infants in the birth range

251

PREMATURE RUPTURE OF MEMBRANES

TABLE 5
Effect of Socioeconomic Level of Heads of Household among Mothers with Low and High
Frequencies of Spontaneous Premature Rupture of Membranes (PROM)

Frequency of PROM
High-Risk Mothersb

Low-Risk Mothers'

Socioeconomic
Group
Group 1
Group II
Group III
Group IV
Total

With PROM
Total
No.

No.

317
259
108
145
829

17
12
6
5
40

With PROM

Total
No.

No.

5.4
4.6
5.6
3.4
4.8

82
236
137
232
687

13
27
9
24
73

15.9
11.4
6.6
10.3
10.6

Tests of independence for frequency of PROM between different socioeconomic status (SES) groups:
Low-Risk: Chi-square = 0.094; d.f. = 3; p = 0.81; Not significant
High-Risk: Chi-square = 4.92; d.f. = 3; p = 0.178; Not significant
aLow-risk: Includes low-risk mothers and mothers with pre-eclampsia, hypertension, diabetes mellitus,
and gestational diabetes
bHigh-risk: Includes high-risk groups shown in Table 2, except for parts a and f.

of under 2,500 g born to mothers with PROM was several times greater than the
incidence of pre-term infants born to mothers with no PROM.
The tendency of pre-term infants born to mothers with PROM to have very low birth
weights (under 2,000 g) was associated with the adverse practices adopted by mothers
during the pregnancy, especially cigarette smoking. Thus, this study provides suggestive evidence that reducing adverse practices, especially smoking, may reduce the
incidence of PROM, and, if PROM still occurs, the incidence of small pre-term infants
would probably be lower.
ACKNOWLEDGEMENT
The authors wish to thank Donna Calandro for assistance with the word processing for this paper.

REFERENCES
1. Guidelines for Perinatal Care, second edition. American Academy of Pediatrics and American College
of Obstetrics and Gynecologists, 1988
2. Oxorn H: Human Labor and Birth. Norwalk, CT, Appleton-Century-Crofts, 1988
3. Cooperstock M, England JE, Wolfe RA: Circadian incidence of premature rupture of the membranes in
term and preterm births. Am J Obstet Gynecol 69:936-941, 1980
4. Meyer MD, Tonascia JA: Maternal smoking, pregnancy complications and perinatal mortality. Am J
Obstet Gynecol 128:494-502, 1977
5. Miller HC, Jekel JF: The effect of race on the incidence of low birth weight: Persistence of effect after
controlling for socioeconomic, educational, marital, and risk status. Yale J Biol Med 60:221-232, 1987
6. Miller HC, Jekel JF: Incidence of low birth weight infants born to mothers with multiple risk factors.
Yale J Biol Med 60:397-404, 1987
7. Miller HC, Jekel JF: The epidemiology of white full-term infants with short crown-heel lengths for
gestational ages at birth. Yale J Biol Med 62:1-12, 1989
8. Miller HC: Intrauterine growth retardation: An unmet challenge. Am J Dis Child 135:944-948, 1981
9. Dubowitz LMS, Dubowitz V, Goldberg C: Clinical assessment of gestational age in newborn infants. J
Pediat 77:1-10, 1970
10. Jekel JF, Miller HC: Epidemiology of spontaneous premature rupture of membranes. Submitted for
publication