Annals of Epidemiology xxx (2018) 1e7

Contents lists available at ScienceDirect

Annals of Epidemiology
journal homepage: www.annalsofepidemiology.org

Original article

Short interpregnancy interval and adverse birth outcomes in women

of advanced age: a population-based study
Timothy O. Ihongbe, MBBS, MPH a, *, Jordyn T. Wallenborn, MPH, PhD a,
Sylvia Rozario, MBBS, MPH a, Saba W. Masho, MD, MPH, DrPH a, b, c
a
Division of Epidemiology, Department of Family Medicine and Population Health, School of Medicine, Virginia Commonwealth University, Richmond, VA
b
Department of Obstetrics and Gynecology, School of Medicine, Virginia Commonwealth University, Richmond, VA
c
Institute for Women's Health, Virginia Commonwealth University, Richmond, VA

a r t i c l e i n f o a b s t r a c t

Article history: Purpose: Short interpregnancy interval (IPI) has been linked with adverse birth outcomes. However, the
Received 7 January 2018 association in advanced age women needs further investigation. This study aims to examine the asso-
Accepted 15 June 2018 ciation between short IPI and adverse birth outcomes including preterm birth, post-term birth, low birth
Available online xxx
weight, and macrosomia, in a population of advanced age U.S. women.
Methods: The 2016 U.S. public-use natality data was analyzed. Analysis was restricted to women with
Keywords:
second-order singleton live births who were
35 years at first live birth (n ¼ 46,684). Multinomial logistic
Interpregnancy interval
regression analysis was used to examine the association between short IPI and adverse birth outcomes.
Preterm birth
Post-term birth
Results: Short IPI in advanced age women was significantly associated with higher odds of extremely
Low birth weight preterm birth (0e5 months IPI: adjusted odds ratio [AOR] ¼ 2.43, 95% confidence interval [CI] ¼ 1.07
Macrosomia adverse birth outcomes e5.52; 6e11 months IPI: AOR ¼ 2.17, 95% CI ¼ 1.09e4.31), very preterm birth (0e5 months IPI: AOR ¼
Advanced maternal age 1.63, 95% CI ¼ 1.04e2.56), and extremely low birth weight (0e5 months IPI: AOR ¼ 2.43, 95% CI ¼ 1.28
e4.60) in the second delivery. An inverse relationship between short IPI and post-term birth was
observed and no significant association between short IPI and macrosomia was found.
Conclusions: Short IPI in advanced age women increases the odds of adverse birth outcomes in the
second delivery.
© 2018 Elsevier Inc. All rights reserved.

Introduction
Over the past 2 decades, women have increasingly initiated
childbirth later in life due to career and educational goals [1]. In the
United States, from 2007 to 2016, birth rates for women in their late
30s and early 40s rose by 11% and 19%, respectively [2], while birth
rates declined to record lows for women in all age groups under
30 years [3]. Consequently, women of advanced age (35 years and
older at first live birth) often have short pregnancy intervals in an
attempt to have their desired number of children and avoid preg-
nancy complications that may arise due to increasing age. Results
from a study that examined the 2006e2010 National Survey of
* Corresponding author. Division of Epidemiology, Department of Family Medi-
cine and Population Health, School of Medicine, Virginia Commonwealth Univer-
sity, P.O. Box 980212, Richmond, VA 23298. Tel.: þ1-803-476-6318; fax: þ1-804-
828-9773.
E-mail address: ihongbeto@vcu.edu (T.O. Ihongbe).
Family Growth (NSFG) reported that women who initiated child-
bearing at 30 years and older were significantly more likely to have
short interpregnancy intervals (IPIs) than women who initiated
childbearing before the age of 30 years [4].
IPI is the time interval between a previous live birth and
conception of the subsequent pregnancy [5]. While a strict cutoff
for optimum IPI is yet to be established, research has shown that
short IPI (less than 18 months) is associated with poor birth out-
comes when compared to an IPI of 18e23 months [6e10]. A
retrospective cohort study among Dutch women of reproductive
ages demonstrated that short IPI was associated with preterm birth
and low birth weight [9]. Similarly, a population-based cohort
study of reproductive age U.S. women reported that short IPI
increased the risk of preterm birth and its recurrence [11]. In
addition, a meta-analysis by Kozuki et al. demonstrated that short
IPI increased the risk of small for gestational age, infant mortality,
and preterm birth in women of reproductive ages [12].
A few studies have reported positive associations for the rela-
tionship between short IPI and adverse perinatal outcomes in

https://doi.org/10.1016/j.annepidem.2018.06.007
1047-2797/© 2018 Elsevier Inc. All rights reserved.

Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
2 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7

women of advanced age [7,9]. However, these studies used a
methodological approach that compared the effect of short IPI on
perinatal outcomes in advanced age women relative to women of
nonadvanced age (e.g., 25e29 years). Although these studies
attempted to adjust for confounding effects that may arise from
reproductive and fertility differences between women of advanced
age and younger age women [13], the choice of younger age women
as comparison group may result in residual and/or unmeasured
confounding that may bias the study findings [14]. As a result, effect
sizes may be underestimated or overestimated. A simulation study
that examined the impact of residual and unmeasured confounding
in epidemiologic studies demonstrated that effect sizes of the
magnitude frequently reported in observational epidemiologic
studies may be generated by residual and/or unmeasured con-
founding alone [14].
To address this challenge, the present study uses women of
advanced age as the comparison group, that is, advanced age
women who had short IPI were compared with other advanced
aged women who had “optimal IPI,” to determine the odds of
adverse birth outcomes when compared to normal birth outcomes.
This methodological approach will effectively control for repro-
ductive and fertility differences that may otherwise act as residual
or unmeasured confounding and will produce a more accurate
estimate of the true effect of short IPI on adverse birth outcomes in
women of advanced age. Therefore, this study, using a nationally
representative sample of advanced age women, aims to (1)
examine the association between short IPI and adverse birth out-
comes among women of advanced age in the United States and (2)
address the challenge of residual or unmeasured confounding by
comparing women of advanced age who have short IPI with other
advanced aged women who have “optimal IPI.”
Material and methods
Data source and study population
This cross-sectional study used the 2016 public-use natality data
(National Vital Statistics System, National Center for Health
Statistics) for all births registered in the 50 states of the United
States, the District of Columbia, and New York City [15]. The public-
use natality data set is a collaborative effort between the states,
National Center for Health Statistics, and the National Vital Statis-
tics System that collates birth records such as demographic char-
acteristics, medical and obstetric history, and perinatal outcomes
from birth certificates. Necessary information required to measure
IPI was available in the 2003 revised birth certificate, which was
used by all states in the United States and the District of Columbia
as of January 1, 2016.
The study population was restricted to women who were aged
35 years or older at the time of their first live birth. In addition, only
women with a second-order singleton live birth of at least
20 weeks' gestation were included in the analysis. Maternal age at
first live birth was computed by subtracting the time interval be-
tween the first and second live births from the maternal age at the
second live birth. Younger age women (less than 35 years at first
live birth), women with first-order singleton births (i.e., the index
delivery is the first live birth) or third or higher order singleton
births, multifetal births (e.g., twins, triplets), or live births with
congenital birth defects were excluded from the analysis (Fig. 1).
Third-order singleton births were excluded to avoid the effect of
clustering, and fourth or higher order singleton births were
excluded because such births could not be linked to the first birth.
Also, women with multifetal births and congenital birth defects
were excluded because of their higher risk for adverse birth out-
comes [16,17]. This yielded a total of 46,684 women. Because the
data set is publicly available and deidentified with no protected
health information, ethical review by the Virginia Commonwealth
University institutional review board was not required.
Measures
Outcomes
Study outcomes include preterm birth, post-term birth, low
birth weight, and macrosomia. Preterm birth and post-term birth
were measured using the gestational age at time of delivery. The
gestational age at time of delivery was measured using the obstetric

All births registered in the 50

states of the US, Washington
D.C., and New York City in
2016
N = 3,956,112

Excluded (N = 3,909,428)
• Women <35 years at first live birth
N=3,821,554
• Women with first, third or higher order
births N=81,083
• Women with multifetal births (e.g. twins,
triplets) N=6,472
• Women with births with congenital defects
N=194
• Women with births <20 weeks gestation
N=25
• Women with implausible interpregnancy
interval N=100

Participants in study population

N = 46,684

Fig. 1. Selection of study population.

Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7
estimate of gestation at delivery, which has been reported to have
greater validity over the last menstrual periodebased measure of
gestational age [18]. Based on the World Health Organization
classification [19], preterm birth was categorized into extremely
preterm (<28 weeks), very preterm (28 to <32 weeks), and
moderate-to-late preterm (32 to <37 weeks) births. Post-term birth
was measured as the birth of a baby on or after 42 weeks of
gestation (
42 weeks). Term delivery (37 to <42 weeks) was used
as the reference.
Low birth weight and macrosomia were measured using the
weight of the baby at birth. Low birth weight was categorized based
on the World Health Organization classification [20] into extremely
low birth weight (<1000 g), very low birth weight (1000 ge1499 g),
and moderately low birth weight (1500 ge2499 g). Macrosomia
was measured as the birth of a baby having a birth weight of at least
4000 g (
4000 g). Normal birth weight (2500 ge3999 g) was used
as the reference.
Exposure
The exposure, IPI, was defined as the time interval between a
live birth and conception of a subsequent live birth. It was
calculated as the time interval (months) between the first and
second deliveries minus the obstetric estimate of the gestational
age of the second infant [21]. IPI was classified into five mutually
exclusive categories: 0e5, 6e11, 12e17, 18e23, and greater than or
equal to 24 months, consistent with classifications used in previ-
ous studies [7,9,22]. Short IPI consisted of IPI categories less than
18 months (0e5, 6e11, and 12e17 months). As only the month and
year of the births were available in the natality data set, the day
was assumed to be the 15th day of the month in both cases for all
records. IPI obtained from birth certificate data has been validated
against that obtained from the NSFG, and the distribution of IPI
from birth certificate data was reported to be similar to that of the
NSFG [21]. Women who had implausible IPI (i.e., IPIs that were
negative or shorter than 30 days; <1%) were excluded from the
analysis.
Potential confounders
Sociodemographic factors such as maternal age at first and
second births (continuous), race/ethnicity (non-Hispanic [NH]
white, NH black, Hispanic, NH other, and NH multiple race), ed-
ucation (no high school diploma or General Education Develop-
ment, high school diploma or General Education Development, or
some college, or bachelor's degree or higher), marital status
(married or unmarried), and insurance status (private, Medicaid,
self-pay, and other) were examined. “Other” category included
Indian Health Service, CHAMPUS/TRICARE, other government
(federal, state, local), and charity. In addition, medical and ob-
stetric history such as prepregnancy body mass index (under-
weight [<18.5], normal [18.5e24.9], overweight [25.0e29.9], or
obese [
30]), weight gain during the second pregnancy (contin-
uous), medical morbidity including prepregnancy diabetes,
gestational diabetes, prepregnancy hypertension, and gestational
hypertension (yes or no), previous preterm birth (yes or no),
previous termination of pregnancy (yes or no), use of infertility
treatment (yes or no), previous cesarean section (yes or no), and
adequacy of prenatal care utilization were examined. Adequacy of
prenatal care utilization was measured using the Revised-
Graduated Prenatal Care Utilization Index (R-GINDEX). The R-
GINDEX measures adequacy of prenatal care using the gestational
age of the infant, the trimester during which prenatal care began,
and the total number of prenatal care visits during pregnancy [23].
Risky behavior such as smoking in the third trimester of preg-
nancy (yes or no) was also examined.
Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
3
Statistical analysis
Descriptive analyses highlighting the characteristics of the study
population by IPI were computed, and differences between IPI
categories were tested using c2 test and t test as appropriate.
Multinomial logistic regression models were created to examine
the association between short IPI and adverse birth outcomes using
maximum likelihood estimation. Advanced age women who had
short IPI were compared with other advanced aged women who
had “optimal IPI” to determine the odds of adverse birth outcomes
when compared to normal birth outcomes (i.e., term birth and
normal birth weight). Potential confounders whose inclusion in the
regression model resulted in a change of 10% or more in the un-
adjusted estimate were retained in the adjusted model [24]. In
addition, known confounders of the association between short IPI
and adverse birth outcomes, which have been established in the
literature [8,9,11,25,26], were included in the regression models.
For the model containing categories of preterm birth and post-term
birth, confounders controlled for include maternal age at second
live birth, race/ethnicity, marital status, insurance, weight gain
during second pregnancy, prenatal care utilization (R-GINDEX),
smoking, previous termination of pregnancy, previous preterm
birth, previous cesarean section, use of infertility treatment, and
medical morbidity. For the model containing categories of low birth
weight and macrosomia, confounders controlled for include
maternal age at second live birth, race/ethnicity, education, marital
status, insurance, weight gain during second pregnancy, prenatal
care utilization (R-GINDEX), smoking, previous cesarean section,
use of infertility treatment, and medical morbidity. Odds ratios and
95% confidence intervals (CIs) were computed. Analyses were
conducted using SAS 9.4 (SAS Institute, Cary, NC) using a signifi-
cance level of 5% a priori.
Results
Characteristics of the study population by IPI are described in
Table 1. Almost half (46.7%) of the study population had short IPI
(6.0%, 0e5 months; 16.9%, 6e11 months; and 23.8%,
12e17 months). About 17% of women had IPIs between 18 and
23 months and over a third (36.2%) had long IPIs (
24 months).
Over half of the women in the study population were NH white,
married, had a bachelors' degree or higher, and used private in-
surance, respectively. The mean age at the first and second live
births were approximately 37 (SD, 1.8) and 39 (SD, 2.2) years,
respectively.
Figures 2 and 3 show the prevalence rate of the various cat-
egories of preterm birth and low birth weight as well as post-
term birth and macrosomia across the different categories of IPI
for the second birth. The prevalence rate of extremely preterm,
very preterm, and moderate-to-late preterm births was highest
among women with IPI of 0e5 months and 24 months and above.
Similarly, the prevalence of extremely low birth weight, very low
birth weight, and moderately low birth weight were highest
among women with IPI of 0e5 months and 24 months and above.
The distribution followed a Ueshaped curve, with lowest preva-
lence rates at 6e11, 12e17, and 18e23 months IPIs. However, the
prevalence rate of macrosomia was lowest at IPI of 0e5 months
and highest at the 6e11, 12e17, and 18e23 months IPIs. The
prevalence rate of post-term birth was highest at the 0e5 month
IPI and steadily decreased as the IPI increased to 18e23 months. A
slight increase in the prevalence rate of post-term birth was
observed, as the IPI increased from 18e23 months to 24 months
and above.
Table 2 shows the odds ratios of the multinomial logistic
regression analyses of adverse birth outcomes for each IPI category.
A U-shaped relationship between IPI and adverse birth outcomes
4 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7

Table 1
Characteristics of study population by interpregnancy interval, U.S. natality data, 2016

Characteristics Interpregnancy interval (months)

Total (n ¼ 46,684) 0e5 (n ¼ 2800) 6e11 (n ¼ 7894) 12e17 (n ¼ 11,119) 18e23 (n ¼ 7960)
24 (n ¼ 16,911) P value*
(n [%]) (n [%]) (n [%]) (n [%]) (n [%]) (n [%]) (c2 test)

All participants 100 6.0 16.9 23.8 17.1 36.2 <.0001

Age at first live birth (years) 36.6 (1.8) 36.8 (2.0) 36.7 (2.0) 36.6 (1.8) 36.6 (1.9) 36.5 (1.8) 0.8679y
(mean [SD])
Age at second live birth (years) 39.3 (2.2) 37.8 (2.0) 38.4 (2.0) 38.6 (1.8) 39.1 (1.9) 40.5 (2.1) <.0001y
(mean [SD])
Race/ethnicity <.0001
NH white 28,888 (62.9) 1601 (57.9) 4975 (64.0) 7346 (67.2) 5140 (65.7) 9826 (59.1)
NH black 3446 (7.5) 343 (12.4) 637 (8.2) 684 (6.3) 493 (6.3) 1289 (7.8)
Hispanic 6469 (14.1) 463 (16.8) 1115 (14.3) 1378 (12.6) 986 (12.6) 2527 (15.2)
NH otherz 6321 (13.8) 319 (11.5) 919 (11.8) 1348 (12.3) 1058 (13.5) 2677 (16.1)
NH multiple race 800 (1.7) 38 (1.4) 132 (1.7) 182 (1.7) 145 (1.9) 303 (1.8)
Insurance <.0001
Private insurance 38,358 (82.4) 2023 (72.6) 6384 (81.2) 9468 (85.4) 6752 (85.1) 13,731 (81.5)
Medicaid 5703 (12.3) 590 (21.2) 1068 (13.6) 1105 (10.0) 790 (10.0) 2150 (12.8)
Self-pay 1131 (2.8) 85 (3.1) 213 (2.7) 268 (2.4) 200 (2.5) 545 (3.2)
Otherx 1161 (2.5) 90 (3.2) 201 (2.6) 249 (2.3) 189 (2.4) 432 (2.6)
Education <.0001
No high school diploma or GED 1381 (3.0) 142 (5.1) 233 (3.0) 253 (2.3) 179 (2.3) 574 (3.5)
High school diploma or GED 3287 (7.2) 296 (10.7) 600 (7.8) 646 (5.9) 469 (6.0) 1276 (7.7)
Some college 7539 (16.5) 642 (23.2) 1296 (16.7) 1613 (14.8) 1237 (15.8) 2751 (16.6)
Bachelor's degree or higher 33,629 (73.4) 1683 (60.9) 5616 (72.5) 8417 (77.0) 5951 (75.9) 11,962 (72.2)
Marital status <.0001
Married 41,554 (89.0) 2263 (80.8) 6961 (88.2) 10,078 (90.6) 7198 (90.4) 15,054 (89.0)
Not married 5130 (11.0) 537 (19.2) 933 (11.8) 1041 (9.4) 762 (9.6) 1857 (11.0)
Prepregnancy BMI <.0001
Underweight 1038 (2.3) 41 (1.5) 160 (2.1) 260 (2.4) 191 (2.5) 386 (2.4)
Normal weight 22,938 (50.5) 1100 (40.5) 3776 (49.2) 5719 (52.9) 4110 (53.0) 8233 (50.2)
Overweight 11,725 (25.8) 758 (27.9) 2022 (26.3) 2687 (24.9) 1952 (25.2) 4306 (26.2)
Obese 9687 (21.3) 819 (30.1) 1724 (22.4) 2146 (19.9) 1509 (19.4) 3489 (21.3)
Weight gain during second 29.0 (12.8) 27.6 (13.6) 28.9 (13.0) 29.5 (12.5) 29.1 (12.5) 28.8 (13.0) 0.0463
pregnancy (lbs) (mean [SD])
Prenatal care utilization (R-GINDEX) <.0001
No care 198 (0.4) 24 (0.9) 37 (0.5) 38 (0.4) 29 (0.4) 70 (0.4)
Inadequate 2088 (4.6) 206 (7.6) 388 (5.1) 442 (4.1) 323 (4.2) 729 (4.4)
Intermediate 17,282 (38.1) 1119 (41.2) 3092 (40.3) 4206 (38.9) 2954 (38.2) 5911 (36.0)
Adequate 21,668 (47.8) 1134 (41.8) 3092 (46.4) 4206 (48.3) 2954 (48.3) 5911 (48.8)
Intensive 4126 (9.1) 233 (8.6) 595 (7.8) 903 (8.4) 697 (9.0) 1698 (10.3)
Previous cesarean section 17,370 (37.2) 986 (35.2) 2912 (36.9) 3984 (35.8) 2890 (36.3) 6598 (39.0) <.0001
Previous termination of pregnancy 17,950 (38.5) 944 (33.8) 2648 (33.6) 3900 (35.1) 3001 (37.7) 7457 (44.2) <.0001
Previous preterm birth 1714 (3.7) 137 (4.9) 281 (3.6) 397 (3.6) 293 (3.7) 606 (3.6) 0.0123
Infertility treatment used 2855 (6.1) 62 (2.2) 340 (4.3) 634 (5.7) 512 (6.4) 1307 (7.7) <.0001
Medical morbidityk 7510 (16.1) 496 (17.7) 1264 (16.0) 1644 (14.3) 1139 (14.3) 2967 (17.6) <.0001
Smoking, third trimester 464 (1.0) 71 (2.6) 105 (1.3) 77 (0.7) 52 (0.7) (0.9) <.0001

Frequencies may not add up to total due to missing values, and percentages may not add up to 100% due to rounding.
GED ¼ general educational development; BMI ¼ body mass index; R-GINDEX ¼ revised-graduated prenatal care utilization index.
*
Test for differences between IPI categories.
y
t test for difference in means.
z
American Indian or Alaska Native, Native Hawaiian, or other Pacific Islander.
x
Indian Health Service, CHAMPUS/TRICARE, other government (federal, state, local), and charity.
k
Prepregnancy diabetes, gestational diabetes, prepregnancy hypertension, and gestational hypertension.

was observed. Results from the adjusted regression analyses
showed that women with IPI between 0 and 5 months, compared to
women with IPI of 18e23 months, had significantly greater odds of
having babies who were extremely preterm (adjusted odds ratio
[AOR] ¼ 2.43, 95% CI ¼ 1.07e5.52) and very preterm (AOR ¼ 1.63,
95% CI ¼ 1.04e2.56), relative to women with term births. Similarly,
women with IPI between 6 and 11 months, compared to women
with IPI of 18e23 months, had significantly greater odds of having
extremely preterm babies (AOR ¼ 2.17, 95% CI ¼ 1.09e4.31), relative
to women with term births. The odds of extremely preterm, very
preterm, or moderate-to-late preterm birth in advanced age
women with IPI between 12 and 17 months were not statistically
significant. An inverse relationship between short IPI and the odds
of post-term birth was observed. As the IPI increased from
0e5 months to 12e17 months, the odds of post-term birth signif-
icantly decreased.
Advanced age women with IPI between 0 and 5 months,
compared to women with IPI of 18e23 months, had significantly
greater odds of having extremely low birth weight babies (AOR ¼
2.43, 95% CI ¼ 1.28e4.60), relative to women with normal birth
weight babies (Table 2). The odds of extremely low birth weight,
very low birth weight, and moderately low birth weight babies in
women with IPI of 6e11 and 12e17 months were not statistically
significant. No significant association was observed between short
IPI and macrosomia in women of advanced age.
Discussion
Findings from this study show that almost half of U.S. women of
advanced age with second born, singleton live births had short IPI
(<18 months). This may be due to the desire to have more than one
child and the realization that conception chances decline with

Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7 5

Extremely preterm birth Very preterm birth

Moderate-to-late preterm birth Post-term birth
10

6
(%)

0
0-5 6-11 12-17 18-23 ≥24
Interpregnancy interval (months)

Fig. 2. Percent distribution of extremely preterm birth, very preterm birth, moderate-to-late preterm birth, and post-term birth by interpregnancy interval, U.S. natality data, 2016.
Extremely preterm birth (<28 weeks), very preterm birth (28 to <32 weeks), moderate-to-late preterm birth (32 to <37 weeks), post-term birth (
42 weeks).

increasing age [9,27]. It is also important to note that over a third of
U.S. women of advanced age had long IPI. The long IPI observed
may, however, be attributable to a decrease in fecundity rather than
a conscious desire to wait for longer periods after the birth of the
first child. This is demonstrated by findings from the present study,
which showed that the rate of use of infertility treatment increased
with increasing IPI.
Furthermore, study findings demonstrate that short IPI
(particularly 0e5 and 6e11 months) in women of advanced age is
independently associated with increased odds of extremely pre-
term birth, very preterm birth, and extremely low birth weight
babies. These findings are consistent with a retrospective cohort
study among Dutch women that examined short IPI and perinatal
outcomes in advanced age women compared to younger women

Extremely low birth weight Very low birth weight

Moderately low birth weight Macrosomia
12
11
10
9
8
7
(%)

6
5
4
3
2
1
0
0-5 6-11 12-17 18-23 ≥24
Interpregnancy interval (months)

Fig. 3. Percent distribution of extremely low birth weight, very low birth weight, moderately low birth weight, and macrosomia by interpregnancy interval, U.S. natality data, 2016.
Extremely low birth weight (<1000 g), very low birth weight (1000 ge1499 g), moderately low birth weight (1500 ge2499 g), macrosomia (
4000 g).

Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
6 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7

Table 2
Association between interpregnancy interval and adverse birth outcomes in women of advanced age, U.S. natality data, 2016

Interpregnancy Adverse birth outcome

interval
Unadjusted odds ratio (95% CI)
(months)
Extremely preterm birth Very preterm birth Moderate-to-late preterm birth Post-term birth

0e5 2.58 (1.23e5.43)* 1.77 (1.16e2.69)* 1.02 (0.87e1.21) 2.56 (2.07e3.17)*

6e11 1.97 (1.06e3.68)* 0.64 (0.42e0.97) 0.96 (0.85e1.08) 1.69 (1.41e2.03)*
12e17 1.28 (0.68e2.41) 0.93 (0.66e1.31) 0.83 (0.74e0.93) 1.27 (1.06e1.52)*
18e23 1.00 1.00 1.00 1.00

24 2.34 (1.34e4.09)* 1.22 (0.90e1.65) 1.20 (1.09e1.34)* 1.15 (0.97e1.37)
Extremely low birth weight Very low birth weight Moderately low birth weight Macrosomia
0e5 2.85 (1.58e5.16)* 2.14 (1.01e4.52)* 1.15 (0.93e1.42) 0.89 (0.77e1.02)
6e11 1.44 (0.76e2.72) 1.14 (0.64e2.02) 0.87 (0.74e1.03) 0.96 (0.87e1.06)
12e17 0.97 (0.51e1.86) 1.10 (0.64e1.87) 0.82 (0.70e0.96)* 0.96 (0.88e1.06)
18e23 1.00 1.00 1.00 1.00

24 2.10 (1.22e3.61)* 1.76 (1.10e2.83)* 1.18 (1.03e1.35)* 0.95 (0.87e1.03)

Adjusted odds ratio (95% CI)y,z

Extremely preterm birth Very preterm birth Moderate-to-late preterm birth Post-term birth

0e5 2.43 (1.07e5.52)* 1.63 (1.04e2.56)* 0.98 (0.82e1.17) 2.41 (1.90e3.06)*

6e11 2.17 (1.09e4.31)* 0.69 (0.45e1.06) 0.97 (0.86e1.10) 1.71 (1.39e2.09)*
12e17 1.58 (0.79e3.17) 0.95 (0.66e1.37) 0.86 (0.70e1.02) 1.30 (1.07e1.59)*
18e23 1.00 1.00 1.00 1.00

24 2.02 (1.07e3.81)* 1.08 (0.78e1.51) 1.06 (0.95e1.18) 1.26 (1.04e1.53)*
Extremely low birth weight Very low birth weight Moderately low birth weight Macrosomia
0e5 2.43 (1.28e4.60)* 1.99 (0.88e4.53) 1.07 (0.85e1.35) 0.97 (0.83e1.13)
6e11 1.57 (0.78e3.16) 1.06 (0.57e1.98) 0.86 (0.72e1.04) 0.98 (0.89e1.09)
12e17 1.15 (0.65e2.06) 1.09 (0.53e2.23) 0.86 (0.72e1.01) 0.96 (0.87e1.06)
18e23 1.00 1.00 1.00 1.00

24 1.78 (1.06e2.98)* 1.56 (0.84e2.89) 1.04 (0.89e1.20) 0.98 (0.89e1.08)

Extremely preterm birth (<28 weeks), very preterm birth (28 to <32 weeks), moderate-to-late preterm birth (32 to <37 weeks), term birth (37 to <42 weeks), post-term birth
(
42 weeks), extremely low birth weight (<1000 g), very low birth weight (1000 ge1499 g), moderately low birth weight (1500 ge2499 g), normal birth weight (2500 ge3999
g), macrosomia (
4000 g).
Term birth and normal birth weight served as referent categories, respectively.
*
p < 0.05.
y
Model containing categories of preterm birth and post-term birth: adjusted for maternal age at second live birth, race/ethnicity, marital status, insurance, weight gain
during second pregnancy, prenatal care utilization (R-GINDEX), smoking, previous termination of pregnancy, previous preterm birth, previous cesarean section, use of
infertility treatment, and medical morbidity.
z
Model containing categories of low birth weight and macrosomia: adjusted for maternal age at second live birth, race/ethnicity, education, marital status, insurance,
weight gain during second pregnancy, prenatal care utilization (R-GINDEX), smoking, previous cesarean section, use of infertility treatment, and medical morbidity.

(25e29 years) [9]. De Weger et al. reported adverse effect of short
IPI on preterm delivery in both younger and advanced age women;
however, the effect was weaker among advanced age women. They
suggested that this may be because of better ovarian reserve in
advanced age women due to a self-selection process that results in
reproductively healthier women in the short IPI group. However, de
Weger et al. did not control for important reproductive factors such
as previous poor birth outcomes, previous termination of preg-
nancy, previous cesarean section, and presence of medical mor-
bidities, which may confound the true estimate of the association.
Using advanced age women as comparison group in the present
study and controlling for important reproductive factors, we were
able to provide estimates of the association that addressed the
problem of reproductive and fertility differences.
The present study found an inverse relationship between short
IPI and the odds of post-term birth among advanced age women. As
the IPI increased from 0e5 months to 12e17 months, the odds of
post-term birth significantly decreased. This finding is in contrast
with those from two previous studies conducted among women of
reproductive ages in the general population [28,29]. DeFranco et al.,
in a study to assess the influence of IPI on birth timing, reported
that short IPI had a reduced risk of post-term birth compared to
long IPI [28]. However, DeFranco et al. did not stratify the associ-
ation by maternal age to assess whether a differential association
exists across maternal age. Increasing maternal age (particularly
advanced maternal age) has been reported to be one of the stron-
gest risk factors for post-term birth [30]. Furthermore, higher
prevalence of risk factors for post-term birth such as overweight/
obesity [30,31] was observed in women with short IPI
(0e5 months) relative to women with IPIs of 18e23 and greater
than or equal to 24 months. Overweight/obesity has been shown to
account for almost 20% of all post-term pregnancies in women [30].
In this study population, 58.0% of women with IPIs of 0e5 months
were overweight or obese compared to 44.6% and 47.5% of women
with IPIs of 18e23 and greater than or equal to 24 months,
respectively. Although we controlled for gestational weight gain,
which is an important risk factor for post-term birth (we did not
control for prepregnancy body mass index in the second pregnancy
in the analysis because it was an intermediate of the association
between IPI and adverse birth outcome in the second birth), factors
underlying the inverse relationship are unclear and future studies
are needed to further examine the association between IPI and
post-term birth in advanced age women.
No significant association was observed between short IPI and
macrosomia in advanced age women. Although, no study to the
authors' knowledge has previously examined the effect of short IPI
on macrosomia in advanced age women, the finding is not sur-
prising. The null finding may be partly explained by the maternal
depletion hypothesis [32]. In advanced age women with short IPI,
insufficient recovery from the physiological stresses of a previous
pregnancy and/or subsequent lactation may compromise the
woman's ability to support fetal growth optimally [33], and this
may preclude the development of a macrosomic baby.
This study has several strengths. First, the use of advanced age
women as comparison group helped to eliminate unmeasured and/
or residual confounding that may arise from the use of younger age

Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
 T.O. Ihongbe et al. / Annals of Epidemiology xxx (2018) 1e7
women as comparison group. Second, the use of a recent national
cohort of advanced age women with singleton live births in the
United States allows for findings to be generalized to similar
advanced age women in the United States. Third, this study
examined the effect of short IPI on two adverse birth outcomes
(post-term birth and macrosomia), which previously have not been
examined in women of advanced age. Fourth, we measured
gestational age using the obstetric estimate of gestation at delivery,
which has been reported to have greater validity over the last
menstrual periodebased measure of gestational age. Finally,
despite the cross-sectional design of the study, we were able to
establish temporal sequence between IPI and adverse birth out-
comes because of the order of events. However, the study is not
without limitations. First, we were unable to adjust for some con-
founders such as pregnancy intention, previous low birth weight,
previous macrosomia, and previous post-term birth due to their
unavailability in the public-use natality data. Second, computation
of IPI and age at last live birth resulted in a small percentage (<1%)
of negative or implausible values; however, the numbers were
small and did not significantly impact the estimates of the associ-
ation. Finally, maternal age in the public-use natality data was
available as age in completed years and did not account for the
number of months before the next completed year. So while
computing age at first live birth as described in the Material and
methods section, a proportion of women may have been mis-
classified as not being of advanced age (<35 years at first live birth).
However, this misclassification is nondifferential and may not
significantly impact the study estimates.
Conclusion
Findings from this study demonstrated that short IPI in women
of advanced age is associated with greater odds of extremely pre-
term birth, very preterm birth, and extremely low birth weight
babies. The present study also showed an inverse relationship be-
tween short IPI and post-term birth. Furthermore, short IPI showed
no association with macrosomia in women of advanced age. This
study highlights the negative effect of short IPI on birth outcomes
in women who initiate childbirth at an advanced age.
Acknowledgments
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
References
[1] Goldin C. The quiet revolution that transformed women's employment, edu-
cation, and family. Am Econ Rev 2006;96(2):1e21.
[2] Martin JA, Hamilton BE, Osterman MJK. Births in the United States, 2016.
NCHS Data Brief 2017;(287):1e8.
[3] Hamilton BE, Martin JA, Osterman MJK, Driscoll AK, Rossen LM. Births: Pro-
visional data for 2016. Vital statistics rapid release; no 2. Hyattsville, MD:
National Center for Health Statistics; 2017. https://www.cdc.gov/nchs/data/
vsrr/ report002.pdf. [Accessed 26 October 2017].
[4] Gemmill A, Lindberg LD. Short interpregnancy intervals in the United States.
Obstet Gynecol 2013;122(1):64e71.
[5] Fuentes-Afflick E, Hessol NA. Interpregnancy interval and the risk of prema-
ture infants. Obstet Gynecol 2000;95(3):383e90.
[6] Conde-Agudelo A, Belizan JM, Norton MH, Rosas-Bermudez A. Effect of the
interpregnancy interval on perinatal outcomes in Latin America. Obstet
Gynecol 2005;106(2):359e66.
[7] Nabukera SK, Wingate MS, Kirby RS, Owen J, Swaminathan S, Alexander GR,
et al. Interpregnancy interval and subsequent perinatal outcomes among
Please cite this article in press as: Ihongbe TO, et al., Short interpregnancy interval and adverse birth outcomes in women of advanced age: a
population-based study, Annals of Epidemiology (2018), https://doi.org/10.1016/j.annepidem.2018.06.007
7
women delaying initiation of childbearing. J Obstet Gynaecol Res 2008;34(6):
941e7.
[8] van Eijsden M, Smits LJ, van der Wal MF, Bonsel GJ. Association between short
interpregnancy intervals and term birth weight: the role of folate depletion.
Am J Clin Nutr 2008;88(1):147e53.
[9] de Weger FJ, Hukkelhoven CW, Serroyen J, te Velde ER, Smits LJ. Advanced
maternal age, short interpregnancy interval, and perinatal outcome. Obstet
Gynecol 2011;204(5):421.e1e9.
[10] Nerlander LM, Callaghan WM, Smith RA, Barfield WD. Short interpregnancy
interval associated with preterm birth in US adolescents. Matern Child Health
J 2015;19(4):850e8.
[11] DeFranco EA, Stamilio DM, Boslaugh SE, Gross GA, Muglia LJ. A short inter-
pregnancy interval is a risk factor for preterm birth and its recurrence. Obstet
Gynecol 2007;197(3):264.e1e6.
[12] Kozuki N, Lee AC, Silveira MF, Victora CG, Adair L, Humphrey J, et al. The
associations of birth intervals with small-for-gestational-age, preterm, and
neonatal and infant mortality: a meta-analysis. BMC Public Health
2013;13(3):S3.
[13] Cleary-Goldman J, Malone FD, Vidaver J, Ball RH, Nyberg DA, Comstock CH,
et al. Impact of maternal age on obstetric outcome. Obstet Gynecol
2005;105(5 Pt 1):983e90.
[14] Fewell Z, Davey Smith G, Sterne JA. The impact of residual and unmeasured
confounding in epidemiologic studies: a simulation study. Am J Epidemiol
2007;166(6):646e55.
[15] National Center for Health Statistics. Data file documentations, Public-use
Natality Data, 2016. Hyattsville, Maryland: National Center for Health Statis-
tics; 2017. https://www.cdc.gov/nchs/data_access/vitalstatsonline.htm.
[Accessed 26 October 2017].
[16] Su R, Zhu W, Wei Y, Wang C, Feng H, Lin L, et al. Maternal and neonatal
outcomes in multiple pregnancy: a multicentre study in the Beijing popula-
tion. Chronic Dis Transl Med 2015;1(4):197e202.
[17] Dolan SM, Gross SJ, Merkatz IR, Faber V, Sullivan LM, Malone FD, et al. The
contribution of birth defects to preterm birth and low birth weight. Obstet
Gynecol 2007;110(2 Pt 1):318e24.
[18] Martin JA, Osterman MJK, Kirmeyer SE, Gregory ECW. Measuring gestational
age in vital statistics data: transitioning to the obstetric estimate. Natl Vital
Stat Rep 2015;64(5):1e20.
[19] World Health Organization. Preterm birth. http://www.who.int/mediacentre/
factsheets/fs363/en/. [Accessed 26 October 2017].
[20] World Health Organization. World Health Organization, Geneva. Low birth
weight: a tabulation of available information, WHO/MCH/92.2, 2. New York:
UNICEF; 1992.
[21] Copen CE, Thoma ME, Kirmeyer S. Interpregnancy intervals in the United
States: data from the birth certificate and the national survey of family
growth. Natl Vital Stat Rep 2015;64(4):1e11.
[22] Ball SJ, Pereira G, Jacoby P, de Klerk N, Stanley FJ. Re-evaluation of
link between interpregnancy interval and adverse birth outcomes: retro-
spective cohort study matching two intervals per mother. BMJ 2014;349:
g4333.
[23] Kogan MD, Martin JA, Alexander GR, Kotelchuck M, Ventura SJ,
Frigoletto FD. The changing pattern of prenatal care utilization in the
United States, 1981-1995, using different prenatal care indices. JAMA
1998;279(20):1623e8.
[24] Maldonado G, Greenland S. Simulation study of confounder-selection strate-
gies. Am J Epidemiol 1993;138(11):923e36.
[25] Klerman LV, Cliver SP, Goldenberg RL. The impact of short interpregnancy
intervals on pregnancy outcomes in a low-income population. Am J Public
Health 1998;88(8):1182e5.
[26] Rodrigues T, Barros H. Short interpregnancy interval and risk of sponta-
neous preterm delivery. Eur J Obstet Gynecol Reprod Biol 2008;136(2):
184e8.
[27] Delpisheh A, Brabin L, Attia E, Brabin BJ. Pregnancy late in life: a hospital-
based study of birth outcomes. J Womens Health (Larchmt) 2008;17(6):
965e70.
[28] DeFranco E, Ehrlich S, Muglia L. Influence of interpregnancy interval on birth
timing. BJOG 2014;121(13):1633e40.
[29] Olesen AW, Basso O, Olsen J. An estimate of the tendency to repeat postterm
delivery. Epidemiology 1999;10(4):468.
[30] Roos N, Sahlin L, Ekman-Ordeberg G, Kieler H, Stephansson O. Maternal risk
factors for postterm pregnancy and cesarean delivery following labor induc-
tion. Acta Obstet Gynecol Scand 2010;89(8):1003e10.
[31] Usha Kiran T, Hemmadi S, Bethel J, Evans J. Outcome of pregnancy in a woman
with an increased body mass index. BJOG 2005;112(6):768e72.
[32] Miller JE. Birth intervals and perinatal health: an investigation of three hy-
potheses. Fam Plann Perspect 1991;23:62e70.
[33] Conde-Agudelo A, Rosas-Bermudez A, Castan ~ o F, Norton MH. Effects of birth
spacing on maternal, perinatal, infant, and child health: a systematic review of
causal mechanisms. Stud Fam Plann 2012;43(2):93e114.