the relationship of interpregnancy interval to infant birthweight and length of gestation among...

The relationship of interpregnancy interval toinfant birthweight and length of gestation amonglow-risk women, Georgia

Melissa M. Adams*, Kristin M. Delaney*, Paul W.Stupp*, Brian J. McCarthy* and James S. Rawlings{*WHO Collaborating Center in Perinatal Care and Health ServicesResearch in Maternal Child Health, Division of Reproductive Health,Centers for Disease Control and Prevention, Atlanta, Georgia and{Newborn Medicine, Department of Pediatrics, Madigan ArmyMedical Center, Tacoma, Washington USA

Summary. To examine the association between interpregnancy intervaland low birthweight (5 2500 g), preterm delivery (5 37 weeks' gesta-tion), and inadequate fetal growth, we studied a population-basedsample of 23 388 white and 4885 black women at low risk for adversepregnancy outcomes who delivered their first and second infants inGeorgia from 1980 to 1992. We used fetal death and livebirth certificates.The interpregnancy interval was the time from delivery to the woman'snext conception. For each pregnancy outcome, we stratified by race andused logistic regression to assess the association between interpregnancyinterval and outcome, while controlling for confounders. Intervals 56months were observed for 3.7% of white women and 7.0% of blackwomen and intervals 5 48 months were seen for 16.8% of white womenand 24.8% of black women. Results from logistic regression showed that,for both races, interpregnancy interval was associated with low birth-weight and preterm delivery. Nearly all of the increased risk occurred inintervals 5 6 months or 5 48 months. The magnitude of the increase inrisk associated with these intervals ranged from modest to moderate andwas similar for black and white women. Because short interpregnancyintervals are rare and are weak risk factors among low-risk women,efforts to lengthen interpregnancy intervals are unlikely to reducesubstantially their rates of adverse pregnancy outcomes.

Paediatric and Perinatal Epidemiology 1997, 11, Suppl. 1, 48±62

Address for correspondence: Dr Melissa Adams, MS K-20, Division of Reproductive Health,CDC, 4770 Buford Highway NE, Atlanta, GA 30341-3724, USA.

48# 1997 Blackwell Science Ltd.

Pap7sup DATA

Introduction

Although conventional wisdom suggests that short interpregnancy intervals

increase the risk for a variety of adverse pregnancy outcomes, studies of the

association between interpregnancy interval and pregnancy outcome have yielded

inconsistent results.1±10 The interpregnancy interval is the time from delivery to

the woman's next conception; this interval is sometimes called the `birth-to-

conception' interval.

Interest in the effect of short interpregnancy intervals on delivery of infants

who are preterm (5 37 weeks' gestation) and low birthweight (5 2500 g) was

recently rekindled by a study reporting that short interpregnancy intervals

increase the risk for infants who are simultaneously preterm and low birthweight

more for black women than for white women.10 If confirmed, this finding could

shed light on why black women persistently have a greater risk than white

women of delivering low birthweight infants.11 Because lengthening the

interpregnancy interval is feasible with contraception, confirming this finding

could provide the rationale for public health interventions to lengthen inter-

pregnancy intervals. On the other hand, failing to confirm these findings would

clarify the need to examine other plausible explanations for black women's greater

risk of delivering low birthweight infants. For these reasons, we examined the

association between interpregnancy interval and three pregnancy outcomes ± low

birthweight (5 2500 g), preterm delivery (5 37 weeks' gestation), and inadequate

fetal growth ± among a large population-based sample of low-risk white and black

women who delivered their first and second infants in Georgia from 1980 to 1992.

Two reasons motivated us to study low-risk women. First, previous reports7,12

suggested that women at high risk for adverse pregnancy outcomes were often

overrepresented among women with short interpregnancy intervals; this potential

excess of high-risk women could have biased the analyses. Second, previous

reports5,6 suggested that if an association existed between interval and pregnancy

outcome, the strength of this association was likely to be modest. Modest

associations are most readily observed in low-risk groups, where other factors that

can cause the outcome of interest are less common.

Methods

Study population

Vital records provided the data for this study. Using fetal death and livebirth

certificates filed in Georgia from 1980 to 1992, we linked pregnancies occurring to

the same woman.13 Women from these linked vital records, whose second

pregnancy occurred during the period between 1989 and 1992, whose reported

gestation for the second pregnancy was 20 weeks or greater, and for whom we

Interpregnancy interval 49

# 1997 Blackwell Science Ltd. Paediatric and Perinatal Epidemiology, 11, Suppl. 1, 48±62

Pap7sup DATA

were able to link the information from their first and second pregnancies, have

been used for this analysis. There were 55 519 white women and 28 233 black

women who satisfied these criteria (Table 1). We excluded women whose second

pregnancy had a gestation of less than 20 weeks, because we felt that they were

likely to be incompletely reported. Too few women of other races were available

for meaningful analysis, so they also were excluded. We restricted the study to

second deliveries occurring from 1989 to 1992 because in only these years did the

certificates include a clinical estimate of the length of gestation and information

about maternal smoking. Because of the associations between birthweight and

plurality and fetal survival, we further restricted the analysis to women whose

first and second pregnancies resulted in singleton livebirths.

We defined low-risk women as those who, at their first pregnancy, were aged

20±34 years, had 12 or more years of education, had started prenatal care in the

first trimester, and listed the father's name on the birth certificate. Other analyses

in Georgia have shown that presence of the father's name on the certificate is

associated with reduced risk for infant mortality and may be a better indicator of

paternal support than is marital status, particularly in groups where marriage is

less common.14 Our definition of `low risk' also required that the woman's first

pregnancy resulted in a live-born infant weighing 5 2500 g who survived the first

year of life and that the woman did not smoke during her second pregnancy. We

50 M. M. Adams et al.


(I) Inclusion criteria for women with first and second pregnanciesMaternal race white or black;Second infant born from 1989 to 92;Length of gestation for second infant 20+ weeks; andFirst and second infants were singleton and live-born.Study subjects: 25 204 white and 5264 black women.

(II) Inclusion criteria for low-risk womenMaternal characteristics at first pregnancy:

Age 20±34 years;Education 12+ years;Started prenatal care in the first trimester; andFather's name listed on birth certificate.

Outcome of first pregnancy:Birthweight 2500+ g; andSurvival to 12 months.

Maternal characteristic at second pregnancy:Nonsmoker.

Low-risk study subjects: 25 204 white and 5264 black women.(III) Exclusion criterion for gestational length of first and second pregnancy:

+2.5 standard deviations of race-, sex-, altitude-, and gestational-age-specificmean birthweight.

Remaining study subjects: 23 388 white and 4885 black women.

Table 1. Selection of study population

Pap7sup DATA

excluded women whose first infant was low birthweight; such women have

increased risk for subsequent delivery of another low birthweight infant.15

Because maternal smoking increases the risks for inadequate fetal growth and

preterm delivery16,17, we excluded women who smoked during their second

pregnancy. Overall, 45.4% (25 204) of white women and 18.5% (5264) of black

women met our definition of low risk.

Definitions

We computed the interpregnancy interval as the date of delivery of the second

pregnancy minus both the clinical estimate of the length of gestation of the second

pregnancy and the date of delivery of the first pregnancy. From other analyses, we

observed that the clinical estimate of length of gestation was less likely to be

missing and more likely to be plausible in relation to birthweight than the LMP-

based estimate. For these reasons we used the clinical estimate of the gestation of

the second pregnancy to determine the approximate date of the conception.

Although the interpregnancy interval initially was computed in days, we

converted it to months. Our initial analyses categorised the interpregnancy

interval into 3-month periods, from 0±2 months to 5 48 months. At interpreg-

nancy intervals of 12 months or more, many of the associations for the 3-month

periods were very similar. Thus, for intervals of 12±23 months, we collapsed the

data into 6-month periods and for those of 24±47 months, we collapsed the data

into 12-month periods.

Our three outcomes of interest were birthweight, length of gestation, and

adequacy of fetal growth. Birthweight was categorised as 5 2500 g or 5 2500 g.

Gestation length was categorised as 5 37 weeks or 5 37 weeks. Despite using the

clinical estimate of the length of gestation, we observed some infants with

biologically implausible combinations of birthweight and gestation. Using birth-

weight distributions developed at CDC (Yip R, McLaren N. Optimal birthweight

distribution at specific gestational age of the United States infants: The CDC

intrauterine growth standard [in preparation]), we excluded records in which the

infant's birthweight fell outside of a range defined by the race-, sex-, altitude- and

gestational-age-specific mean birthweight plus or minus 2.5 times its standard

deviation. Acceptable race-, sex-, and gestational-age-specific birthweight ranges

for infants born at sea level are shown in the Appendix.

The observation that mothers tend to repeat birthweight across successive

pregnancies17 prompted us to consider fetal growth (as measured by birthweight)

in relative rather than absolute terms. To assess the adequacy of fetal growth in the

second pregnancy, we used an approach that adjusted for the birthweight of the

mother's first infant, relative to all other firstborn infants. Because of the increase

in absolute birthweight from the first to the second pregnancy,18 we devel-

oped parity-specific birthweight distributions. Thus, for first and second births



Pap7sup DATA

separately, we computed gestation-, race-, and sex-specific birthweight distribu-

tions. Then we assigned each infant to a birthweight quartile, based on the

mother's parity, length of gestation, race and the infant's sex and birthweight.

Because the analysis of fetal growth required an accurate estimate of the length

of gestation of the first birth, we edited the gestations of the first births, using the

birthweight criteria described above for second births. Because clinical estimates

of the length of gestation were not available for all first births, gestations for first

births were computed as the difference between the date of delivery and the date

of the LMP. For consistency, all analyses presented here exclude women with

missing or implausible gestations for their first or second pregnancies. These edits

excluded 7.2% of white women and 7.2% of black women, leaving 23 288 white

and 4885 black women available for analysis. The distribution of interpregnancy

intervals did not differ between women who were excluded or retained on the

basis of the edits for gestational age.

Because there were too few births to compute stable birthweight distributions

for pre- and post-term gestations, we restricted the analysis of fetal growth to

women whose first and second pregnancies resulted in term deliveries (37±42

weeks). Thus, 20 419 white women (87.3% of 23 388) and 4062 black women

(83.2% of 4885) were included in the analyses of fetal growth.

Analysis

Except for the analyses of interactions between race and interpregnancy interval,

all analyses were stratified by race. We began by computing the distribution of

interpregnancy intervals. Next, for each interpregnancy interval, we examined the

distribution of potentially confounding factors. Then, for infants born after each

interval, we computed the mean and median birthweights; the percentages of

infants who were low birthweight or preterm; and the percentage of infants with

birthweights in the first quartile of their gestation-, race-, and sex-specific birth-

weight distribution.

We used logistic regression to assess the association between interpregnancy

interval and pregnancy outcome. Separate models were analysed for each

outcome. All variables were entered into the models as categorical variables. The

first step of the modelling strategy involved estimating the crude odds ratios

(ORs) and 95% confidence intervals (CIs) for the association with interpregnancy

interval. Next we entered potential confounders into the model. We considered

confounding to be present if the adjusted OR differed from the crude OR by more

than 10%. We assessed the adequacy of each model by reviewing the Hosmer and

Lemeshow goodness-of-fit statistic.

For the analyses of low birthweight and preterm delivery, we considered the

following variables from the second pregnancy as potential confounders in the

logistic model: maternal age and education, prenatal care initiation, and presence



Pap7sup DATA

of father's name on the birth certificate. For the analysis of quartile of birthweight

in the second pregnancy, we also considered quartile of birthweight for the

mother's first pregnancy as a potential confounder in the model.

As a final, formal test for differences between blacks and whites in the asso-

ciation between pregnancy outcome and interpregnancy interval, we computed

models that included data for white and black women and terms for the

interaction between race and interpregnancy interval. We judged interactions to

be present if the maximum likelihood ratio test for the interaction terms was

statistically significant.

Results

Distribution of interpregnancy interval

In this sample of low-risk women, short interpregnancy intervals were not

common (Table 2). The distribution of intervals differed by race (P5 0.001): black

women were more likely than white women to have either short or long intervals.

For example, 0.6% of white women compared with 1.7% of black women had

intervals of 5 3 months. Similarly, 16.8% of white women compared with 24.8%

of black women had intervals 5 48 months.

Association between maternal attributes and interval

For both races, the distribution of maternal attributes varied by interpregnancy

interval. Women with shorter intervals tended to have younger ages, fewer years



Table 2. Distribution of second births by interpregnancy interval and race in low-riskwomen, Georgia

Distribution of second births (%)ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

Interpregnancy interval White women Black women(months) (n = 23 388) (n = 4885)

0±2 0.6 1.73±5 3.1 5.36±8 5.3 6.19±11 7.1 7.0

12±17 15.9 13.618±23 15.3 10.724±35 23.4 17.336±47 12.6 13.5548 16.8 24.8

Total 100.0 100.0

Pap7sup DATA

of education, and less first-trimester prenatal care in their second pregnancy

(Table 3). White women with short intervals were slightly less likely to have had

firstborn infants with birthweights in the first quartile (lowest birthweight) of the

distribution for their race, length of gestation, and sex. In contrast, black women

with short intervals were slightly more likely to have had firstborn infants with

birthweights in the first quartile. Independent of interval, white women had a

more favourable profile for positive pregnancy outcome than did black women.

For example, 37% of white women and 23% of black women had 16 or more years

of education.

Pregnancy outcomes

As expected in these low-risk women, the crude risks of adverse pregnancy

outcomes were very low (Table 4). The median interval-specific birthweight

ranged from 3515 to 3572 g for white infants and from 3232 to 3374 g for black

infants. The crude risk for low birthweight was 1.7% for white women and 4.7%

for black women. Overall, 3.5% of white women and 6.1% of black women

delivered preterm infants. As expected by definition, approximately 25% of

infants had birthweights in the first quartile of their race-, gestation-, and sex-

specific distribution, with small variation caused by rounding. For whites, the

greatest risks for low birthweight occurred during interpregnancy intervals of 0±2

months (5.8%), 6±8 months (1.9%), and 5 48 months (2.5%); the greatest risks for

preterm delivery occurred during intervals of 5 9 months (range 3.8±5.8%) and

5 48 months (4.5%); and the greatest risk for birthweight in the first quartile

occurred during the intervals of 3±5 (31.4%) and 6±8 months (28.6%). For all three

outcomes together, the risks were lowest at intervals from 12 to 35 months.

Similar to white infants, black infants had the highest risk for adverse

outcomes at the shortest and longest intervals. The risks for low birthweight were

highest at interpregnancy intervals of 0±2 months (7.1%), 3±5 months (9.7%) and

5 48 months (6.1%). The risks for preterm delivery were greatest at intervals of 3±

11 months (range 7.3±10.9%) and 5 48 months (7.0%). The risks for birthweight in

the first quartile were highest at intervals less than 9 months (range 28.2±30.5%).

Logistic regression

Results from logistic regression showed that, for both races, short and long

interpregnancy intervals were associated with an increased risk for low

birthweight (Table 5). Compared with women with intervals of 24±35 months,

the adjusted odds for low birthweight among white women were appreciably

increased only for the very shortest interval (0±2 months) and the longest interval

(5 48 months). For black women, the comparable analysis showed significantly

elevated adjusted odds for intervals of 3±5 months and 5 48 months. For both



Pap7sup DATA

races, the ORs were adjusted for maternal age and trimester of initiation of

prenatal care at the second pregnancy. The test for an interaction between interval

and race was not significant, suggesting that race did not modify the association

between interval and risk of low birthweight.

In logistic regression models with preterm delivery in the second pregnancy as

the outcome of interest, we adjusted for maternal age, education, and trimester

when prenatal care was initiated in the second pregnancy, and whether the same

man had fathered the first child. Results from logistic regression showed that

interpregnancy interval was associated with preterm delivery for both races.

Compared with women with intervals of 24±35 months, the adjusted odds for

preterm delivery among white women were significantly increased only for the

very longest interval (5 48 months). For black women, the comparable analysis

showed significantly elevated odds for the interval of 3±5 months and marginally

elevated odds for intervals of 9±11 and 5 48 months. Again, the test for an

interaction between interval and race was not significant.

For white, but not black, women, interpregnancy interval was associated with

the odds of having a baby whose birthweight was in the first quartile. Among

white women, the adjusted odds were significantly increased for intervals of 3±5,

6±8, and 36±47 months, compared with intervals of 24±35 months. All of these

increased ORs were of modest magnitude. The logistic regression model was

adjusted for the quartile of birthweight in the first pregnancy. The test for

modification of the association between interval and quartile of birthweight by

race was not significant.

Length of gestation is notoriously difficult to measure accurately. To explore

the possible impact of inaccuracies, we collapsed the intervals of 0±2 months and

3±5 months and repeated the analyses. The results were essentially the same as

described above (Table 5). For both races, interpregnancy interval was associated

with each outcome, and none of the interactions between interval and race were

significant. The only notable finding was that the OR for low birthweight for an

interval of 0±5 months differed by race. The OR was 1.2 [95% Cl: 0.7±1.8] for white

women and 2.3 [1.3±3.9] for black women.

Discussion

Perhaps the most notable feature of these analyses is the nearly uniform absence of

a strong trend between interpregnancy interval and pregnancy outcome. Nearly

all of the increased risk for adverse outcomes occurs in intervals of 5 6 months or

5 48 months. The magnitude of the increase in risk associated with these inter-

vals ranges from modest to moderate and is roughly comparable for black and

white women.

The strengths of this study include the large number of subjects, the use of pop-

ulation-based data for subject selection, the availability of birth records for first and



Pap7sup DATA

second pregnancies, and the selection of low-risk women. Although our focus on

low-risk women may have limited the generalisability of the findings, we judge that

this approach avoided potential confounding effects that could have biased analysis

of the association between interpregnancy interval and pregnancy outcome.

The limitations of this study include the inaccuracies in vital records data, the

small number of women with very short interpregnancy intervals, and the

omission of women whose first and second pregnancies were not linked.

Inaccuracies in the length of gestation recorded on vital records have been a

longstanding problem for investigators.19 To minimise inaccuracies in gestational

age, we excluded records with biologically implausible combinations of gestation

and birthweight. As apparent from some of the wide confidence intervals shown

in Table 5, statistical power for some of the analyses was limited. From a previous

evaluation of our sibship linkage,20 we know that we were unable to link the

pregnancies of some women, such as those whose first delivery occurred outside

of Georgia. Women with second livebirths during 1989±92 for whom we did not

link their previous livebirth comprised approximately 22% of all women with

second births during these years. Quartile of birthweight in the second pregnancy

may have been too crude an outcome to permit detection of an association

between fetal growth and interpregnancy interval.

To the best of our knowledge, the investigation by Rawlings et al.10 is the only

other study to explicitly examine the risk of adverse pregnancy outcome among

low-risk women. Rawlings et al. studied women who delivered at an Army

Medical Center from 1983 to 1993 and defined low-risk women as those whose

previous pregnancy had not resulted in preterm delivery or a low birthweight

infant. Among white low-risk women, interpregnancy interval was not associated

with delivery of a preterm, low birthweight infant. However, among black low-

risk women intervals of 5 9 months significantly increased the risk for this

outcome (relative risk=4.0 [95% CI 1.1±14.8]).

Our findings suggest that, at least for low-risk women, only the very shortest

or longest interpregnancy intervals are likely to increase the risk for adverse

pregnancy outcomes and very short intervals increase the risk only modestly.

Because so few low-risk women have very short intervals, we judge that efforts

directed at low-risk women to lengthen interpregnancy intervals are unlikely to

substantially reduce their rates of adverse pregnancy outcomes. Further work is

needed to determine if these findings also hold for women with obstetric or

demographic risks for adverse pregnancy outcomes.

Acknowledgements

Michael Lavoie, Director, Vital Statistics Office, Center for Health Information

Branch, Division of Public Health, State of Georgia provided vital records data

and consulted on their interpretation.



Pap7sup DATA

References

1 Brody DJ, Bracken MB. Short interpregnancy interval: a risk factor for low birthweight.American Journal of Perinatology 1987; 4:50±54.2 Eisner V, Brazie JV, Pratt MW, Hexter AC. The risk of low birthweight. AmericanJournal of Public Health 1987; 69:887±893.3 Erickson JD, Bjerkedal T. Interpregnancy interval: association with birth weight,stillbirth, and neonatal death. Journal of Epidemiology and Community Medicine 1978; 32:124±130.4 Ferraz EM, Gray RH, Fleming PL, Maia TM. Interpregnancy interval and low birthweight: findings from a case-control study. American Journal of Epidemiology 1988; 128:1111±1116.5 Greenwood R, Samms-Vaughan M, Golding J, Ashley D. Past obstetric history and riskof perinatal death in Jamaica. Paediatric and Perinatal Epidemiology 1994; 8(Suppl 1):40±53.6 Kallan JE. Effects of interpregnancy intervals on preterm birth, intrauterine growthretardation, and fetal loss. Social Biology 1992; 39:231±245.7 Klebanoff MA. Short interpregnancy interval and the risk of low birthweight. AmericanJournal of Public Health 1988; 78:667±670.8 Lieberman, E, Lang JM, Ryan KJ, Monson RR, Schoenbaum SC et al. The association ofinterpregnancy interval with small for gestational age births. Obstetrics and Gynecology 1989;74:1±5.9 Mavalankar DV, Gray RH, Trivedi CR. Risk factors for preterm and term lowbirthweight in Ahmedabad, India. International Journal of Epidemiology 1992; 21:263±272.10 Rawlings JS, Rawlings VB, Read JA. Prevalence of low birth weight and pretermdelivery in relation to the interval between pregnancies among white and black women.New England Journal of Medicine 1995; 332:69±74.11 NCHS. Advance report of final natality statistics, 1992. (Monthly vital statistics report;vol 43, no. 5, suppl). Hyattsville, Maryland: US Department of Health and Human Services,Public Health Service, CDC, 1994.12 Fedrick J, Adelstein P. Influence of pregnancy spacing on outcome of pregnancy. BritishMedical Journal 1973; 4:753±756.13 Adams MM, Wilson HG, Casto DL, Berg CJ, McDermott JC, Gaudino JA et al.Construction of reproductive histories by linking vital records. American Journal ofEpidemiology 1997; (in press).14 Gaudino JA, Jenkins W, Rochat, R, Griffin G. No father's name: a risk factor for infantmortality in Georgia. American Journal of Epidemiology 1993; 138:655.15 Bakketeig LS, Hoffman HJ, Harley EE. The tendency to repeat gestational age and birthweight in successive births. American Journal of Obstetrics and Gynecology 1979; 135:1086±1103.16 Meis PG, Micielutte R, Peters TJ, Wells HB, Sands RE, Coles EC et al. Factors associatedwith preterm birth in Cardiff, Wales. II. Indicated and spontaneous preterm birth. AmericanJournal of Obstetrics and Gynecology 1995; 173:597±602.17 Shu XO, Hatch MC, Mills J, Clemens J, Susser M. Maternal smoking, alcohol drinking,caffeine consumption, and fetal growth: results from a prospective study. Epidemiology 1995;6:115±120.18 Kline J, Stein Z, Susser M. Conception to Birth. New York: Oxford University Press, 1989;p. 222.19 Alexander GR, Tompkins ME, Petersen DH, Hulsey TC, Mor J. Discordance betweenLMP-based and clinically estimated gestational age: implications for research, programs,and policy. Public Health Reports 1995; 110:396±402.



Pap7sup DATA

20 Adams MM, Berg CJ, McDermott M, Gaudino JA, Casto DL, Wilson HG et al.Evaluation of reproductive histories constructed by linking vital records. Paediatric andPerinatal Epidemiology 1997; 11:78±92.



Appendix Acceptable combinations of birthweight (g) and gestational age for infantsborn at sea level, by maternal race and infant sex

Infants of white women Infants of black womenWeeks of ÐÐÐÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐÐÐÐgestation Males Females Males Females

526 41468 41323 41405 4132726 468±1468 483±1323 450±1405 427±132727 467±1732 467±1597 437±1677 403±157328 479±2004 456±1886 431±1961 405±185529 522±2277 475±2175 455±2245 427±212730 611±2541 540±2450 523±2523 492±239231 742±2729 655±2710 640±2785 600±265032 918±3043 817±2952 801±3036 757±289233 1125±3280 1018±3178 1003±3273 948±312334 1356±3506 1246±3386 1224±3489 1163±333835 1593±3728 1486±3581 1459±3694 1391±354136 1828±3943 1723±3768 1692±3887 1818±392837 2043±4158 1945±3950 1910±4070 1827±390738 2235±4365 2140±4130 2100±4245 2011±407639 2392±4567 2300±4305 2258±4413 2162±423740 2519±4759 2432±4472 2386±4571 2284±438941 2625±4930 2538±4628 2489±4719 2489±471942 2727±5072 2641±4756 2589±4849 2469±466943±44 52727 52641 52589 52469

Pap7sup DATA

#1997

Black

well

Scien

ceL

td.

Paediatric

and

Perin

atalE

pidemiology,

11,S

up

pl.

1,48±62

Interpregn

ancy

interval

55

Table 3. Percentage distribution of maternal characteristics, by interpregnancy interval and race in low-risk women, GeorgiaWhite women (n = 23 388)

Interpregnancy interval (months)0±2 3±5 6±8 9±11 12±17 18±23 24±35 36±47 5 48

ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ(a) White women (n = 23 388) (n = 137) (n = 727) (n = 1227) (n = 1661) (n = 3711) (n = 3566) (n = 5475) (n = 2950) (n = 3934)

Agea (years)20±24 54.7 35.4 26.2 23.7 18.2 14.7 11.8 6.4 0.125±29 32.1 41.7 44.5 44.6 47.2 46.9 45.9 48.5 30.030±34 12.4 21.1 26.9 27.9 30.4 32.9 36.1 37.2 44.935+ 0.7 1.9 2.4 3.7 4.3 5.5 6.2 8.0 16.0

Educationa(years)12 51.9 40.6 33.0 29.9 29.0 31.3 34.0 41.6 50.013±15 24.4 27.7 25.3 27.1 26.7 26.3 26.2 25.9 26.416+ 23.7 31.7 41.8 43.0 44.4 42.4 39.8 32.6 23.5

Prenatal care initiation (trimester)a

1st 81.8 84.9 89.1 90.9 94.1 94.7 95.0 95.5 94.4Other/unknown 18.2 15.1 10.9 9.1 5.9 5.3 5.0 4.5 5.6

Paternitya

Same as 1st pregnancy 100.0 98.6 98.5 98.4 98.4 97.6 96.3 93.0 77.6Otherb 0.0 1.4 1.5 1.6 1.6 2.4 3.7 7.0 22.4

Birthweight quartilec

1 (lightest) 22.2 23.5 24.7 26.2 25.3 25.9 25.5 26.5 26.52 35.0 25.6 23.2 25.9 25.5 25.3 26.1 25.0 24.93 23.9 27.8 26.1 25.1 24.5 24.3 24.1 24.2 23.64 (heaviest) 18.8 23.2 26.0 22.7 24.8 24.5 22.2 24.3 25.1

Pap7sup DATA

#1997

Black

well

Scien

ceL

td.

Paediatric

and

Perin

atalE

pidemiology,

11,S

up

pl.

1,48±62

56M

.M

.A

dams

etal.

Table 3 contd.

Interpregnancy interval (months)0±2 3±5 6±8 9±11 12±17 18±23 24±35 36±47 5 48

ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ(b) Black women (n = 4885) (n = 85) (n = 257) (n = 300) (n = 341) (n = 666) (n = 521) (n = 845) (n = 659) (n = 1211)

Age a(years)20±24 55.3 53.3 47.7 45.8 35.9 32.4 20.0 10.3 0.225±29 34.1 33.5 32.7 38.4 41.9 44.5 50.2 54.0 42.430±34 10.6 12.8 17.7 13.5 18.6 19.8 25.2 29.9 43.435+ 0.0 0.4 2.0 2.4 3.6 3.3 4.6 5.8 14.1

Education a(years)12 48.8 54.0 52.2 42.7 45.1 45.2 43.5 46.7 48.013±15 36.9 31.8 26.4 36.5 31.0 28.1 31.7 28.2 31.116+ 14.3 14.3 21.4 20.8 23.9 26.7 24.8 25.1 20.9

Prenatal care initiation (trimester)a

1st 56.5 62.3 67.7 76.0 78.4 82.7 85.2 85.3 83.7Other/unknown 43.5 37.7 32.3 24.0 21.6 17.3 14.8 14.7 16.3

Paternitya

Same as 1st pregnancy 89.4 85.2 88.7 86.2 87.2 84.3 81.8 79.7 64.5Otherb 10.6 14.8 11.3 13.8 12.8 15.7 18.2 20.3 35.5

Birthweight quartilec

1 (lightest) 27.5 27.0 29.4 23.6 26.2 25.7 24.2 24.4 26.32 31.9 20.0 22.5 19.9 26.0 25.1 26.3 26.3 23.43 27.5 30.0 24.1 30.4 24.5 25.1 23.5 23.7 25.14 (heaviest) 13.0 23.0 24.1 26.1 23.3 24.2 26.0 25.6 25.3

aMaternal attribute at second pregnancy. Education missing for 218 white and 57 black women.bOther=Different fathers in first and second pregnancies or father's name not stated for first or second pregnancy. Because of rounding,percentages may not sum to 100%.cRelative birthweight of infant delivered in first pregnancy compared with other infants of the same gestation, sex, and race (see text forcomputation of quartiles). Computed only for the subset of women whose first and second pregnancies were term deliveries.

Pap7sup DATA

#1997

Black

well

Scien

ceL

td.

Paediatric

and

Perin

atalE

pidemiology,

11,S

up

pl.

1,48±62

Interpregn

ancy

interval

57

Table 4. Outcomes of second births, by interpregnancy interval and race, in low-risk women, Georgia

Outcome of second birthÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

White women (n = 23 388) Black women (n = 4885)ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

Median Low First quartile Median Low First quartileInterpregnancy birthweight birthweight of birthweight birthweight birthweight of birthweightinterval (months) (g) (%) Preterm (%) (%) (g) (%) Preterm (%) (%)

0±2 3544 5.8 5.8 24.8 3232 7.1 5.9 29.03±5 3515 1.2 4.5 31.4 3232 9.7 10.9 30.56±8 3515 1.9 3.8 28.6 3260 3.0 7.3 28.29±11 3572 1.3 2.9 24.8 3317 3.8 7.6 21.412±17 3544 1.5 3.3 25.9 3317 3.3 5.1 25.118±23 3544 1.6 3.2 24.2 3374 4.4 4.4 28.024±35 3544 1.5 3.2 25.5 3317 4.0 5.6 23.536±47 3544 1.7 3.1 27.2 3317 3.3 4.4 26.7548 3544 2.5 4.5 26.4 3317 6.1 7.0 26.0

Total 3544 1.7 3.5 26.0 3317 4.7 6.1 25.9

Pap7sup DATA

#1997

Black

well

Scien

ceL

td.

Paediatric

and

Perin

atalE

pidemiology,

11,S

up

pl.

1,48±62

58M

.M

.A

dams

etal.

Table 5. Results of logistic regression for the associations between interpregnancy interval and outcome of second pregnancy (lowbirthweight, preterm delivery, and 1st quartile of birthweight), in low-risk women, Georgia

Pregnancy outcome and odds ratio [95% Cl]ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

Low birthweight Preterm delivery 1st quartile of birthweightÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐÐÐÐÐÐ ÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

Interpregnancyinterval White Black White Black White Black(months) women women women women women women

0±2 3.6 [1.7±7.8] 1.8 [0.7±4.4] 1.6 [0.8±3.5] 1.1 [0.4±2.9] 0.9 [0.6±1.4] 1.2 [0.7±2.1]3±5 0.8 [0.4±1.6] 2.5 [1.4±4.3] 1.3 [0.9±2.0] 2.0 [1.2±3.4] 1.4 [1.2±1.7] 1.5 [1.0±2.1]0±5 1.2 [0.7±1.8] 2.3 [1.3±3.9] 1.4 [1.0±2.0] 1.8 [1.1±2.9] 1.3 [1.1±1.6] 1.4 [1.0±1.9]6±8 1.2 [0.7±1.9] 0.8 [0.4±1.6] 1.1 [0.8±1.5] 1.4 [0.8±2.5] 1.2 [1.1±1.5] 1.2 [0.9±1.7]9±11 0.8 [0.5±1.3] 1.0 [0.5±1.9] 0.9 [0.6±1.2] 1.4 [0.9±2.4] 0.9 [0.8±1.1] 0.9 [0.6±1.3]

12±17 1.0 [0.7±1.4] 0.8 [0.5±1.4] 1.0 [0.8±1.3] 1.0 [0.6±1.5] 1.0 [0.9±1.2] 1.1 [0.8±1.4]18±23 1.1 [0.7±1.5] 1.1 [0.6±1.9] 1.0 [0.8±1.2] 0.9 [0.5±1.4] 0.9 [0.8±1.0] 1.3 [1.0±1.7]24±35 1.0 ref 1.0 ref 1.0 ref 1.0 ref 1.0 ref 1.0 ref36±47 1.1 [0.8±1.6] 0.8 [0.5±1.5] 1.0 [0.7±1.2] 0.8 [0.5±1.4] 1.1 [1.0±1.2] 1.2 [0.9±1.6]548 1.7 [1.3±2.3] 1.5 [1.0±2.4] 1.4 [1.1±1.8] 1.3 [0.9±2.0] 1.1 [0.9±1.2] 1.1 [0.9±1.4]

Covariates Maternal age and prenatal Maternal age, education, Quartile of birthweight in 1stcare prenatal care, and paternity pregnancy

P-value for w2 0.0004 0.0007 0.0056 0.0090 0.0043 0.2832test ofsignificance ofinterpregnancyintervalGoodness-of-fit 0.67 0.93 0.94 0.82 0.87 0.60test (P-values)P-value for black± 0.055P5 0.10 P4 0.30 P4 0.50white interaction

Pap7sup DATA

the relationship of interpregnancy interval to infant birthweight and length of gestation among...

Documents