Logo of amjphAmerican Journal of Public Health Web SiteAmerican Public Health Association Web SiteSubmissionsSubscriptionsAbout Us
Am J Public Health. 2007 January; 97(1): 157–162.
PMCID: PMC1716235

Maternal Obesity in Early Pregnancy and Risk of Spontaneous and Elective Preterm Deliveries: A Retrospective Cohort Study


Objectives. We sought to determine the association between maternal body mass index and risk of preterm delivery.

Methods. We assessed 187 290 women in Scotland and estimated adjusted odds ratios for spontaneous and elective preterm deliveries among overweight, obese, and morbidly obese women relative to normal-weight women.

Results. Among nulliparous women, the risk of requiring an elective preterm delivery increased with increasing BMI, whereas the risk of spontaneous preterm labor decreased. Morbidly obese nulliparous women were at increased risk of all-cause preterm deliveries, neonatal death, and delivery of an infant weighing less than 1000 g who survived to 1 year of age (a proxy for severe long-term disability). By contrast, obesity and elective preterm delivery were only weakly associated among multiparous women.

Conclusions. Obese nulliparous women are at increased risk of elective preterm deliveries. This in turn leads to an increased risk of perinatal mortality and is likely to lead to increased risks of long-term disability among surviving offspring.

The association between maternal obesity and pregnancy outcomes is complex. Maternal obesity is known to be associated with increased rates of complications in late pregnancy such as stillbirth, cesarean delivery, gestational diabetes, and shoulder dystocia.13 However, a low body mass index (BMI) is associated with an increased risk of preterm delivery, and some studies have shown that BMIs above the “normal” range are protective against spontaneous preterm births.46 A large-scale retrospective cohort study demonstrated an interaction between BMI and parity: obese nulliparous women were at increased risk of extreme preterm deliveries and neonatal death, whereas obese multiparous women were not at increased risk of these outcomes.7 The reasons for these complex patterns of association are unclear.

Preterm deliveries can occur as a result of preterm labor or can be elective procedures. Preeclampsia is the reason for elective preterm deliveries in more than 40% of cases.4 It is well recognized that obese women are at increased risk of preeclampsia8 and that nulliparous women are at higher risk of preeclampsia than multiparous women. We hypothesized that the higher background risk of preeclampsia among nulliparous women might lead to a stronger association between obesity and elective preterm deliveries and might therefore explain the association between obesity and extreme preterm deliveries among these women.

Our aim was to determine the association between maternal obesity in early pregnancy and risk of preterm delivery, with attention given to type of delivery (spontaneous vs elective), parity (nulliparous vs multiparous), and the most important negative consequences of prematurity. In assessing consequences of prematurity, we examined both neonatal death and long-term survival of extremely low-birthweight (ELBW) infants. Because ELBW infants have a 40% to 45% risk of severe neurodevelopmental delays in childhood,9 we used ELBW as a proxy measure of severe long-term morbidity.


Data Sources and Patient Selection

The Scottish Morbidity Record (SMR2) collects information on clinical and demographic characteristics and outcomes for all women discharged from Scottish maternity hospitals. The registry is subjected to regular quality assurance checks, and its data have been more than 99% complete since the late 1970s.10 In addition, the Scottish Stillbirth and Infant Death Enquiry (SSBIDE), a national register, routinely classifies all perinatal deaths in Scotland.11

All women presenting for prenatal care in the west of Scotland are offered biochemical screening, using maternal serum α-fetoprotein and human chorionic gonadotrophin, to assess their risk of having a fetus affected by Down syndrome or a structural fetal abnormality.12 Maternal weight is recorded at the time of sampling for biochemical screening to allow for weight correction of analytes. This process corrects levels of these proteins for the effect of maternal size and improves prediction of Down syndrome risk. The laboratory information management system of the West of Scotland Regional Genetics Service (Institute of Medical Genetics) contains a database including this maternal information along with biochemical screening results. The General Registrar’s Office maintains computerized birth and death registration records.

We used a probability-based matching approach13 with maternal identifiers to link information from the SMR2, the SSBIDE, the Institute of Medical Genetics prenatal screening database, and the General Registrar’s Office database of birth certificates. We used offspring identifiers contained in the birth certificates used to link biochemical, pregnancy, and perinatal mortality data to the death certificate registry, allowing us to identify deaths among offspring. We excluded multiple births, stillbirths, and births occurring outside 22 to 43 weeks of gestation.

Births in the cohort assessed here occurred between November 1991 and December 2001. The cohort was defined as women who (1) had a record in the prenatal screening database (in which maternal weight was recorded), (2) could be linked to an SMR2 record, (3) had given birth to a singleton infant weighing more than 400 g, and (4) had given birth between 22 and 43 weeks of gestation. In addition to excluding stillbirths and perinatal deaths because of fetal abnormalities, we excluded women with missing data.


Several outcomes were examined: preterm delivery, spontaneous preterm delivery, elective preterm delivery, neonatal death, delivery of an ELBW infant, delivery of an ELBW infant surviving to 1 year of age, and preeclampsia. A preterm delivery was defined as a birth occurring before 37 weeks of gestation, and a term delivery was defined as a birth occurring at or after 37 weeks of gestation. A spontaneous delivery was defined as a vaginal birth or a birth in which the woman was documented as having been in labor at the time of delivery but the labor was not documented as having been induced and was therefore presumed to be spontaneous. An elective delivery was defined as a birth in which the woman did not experience spontaneous labor (i.e., an induced vaginal birth or cesarean birth without a documented duration of labor).

Infants weighing between 400 g and 1000 g were classified as ELBW infants. Infants recorded as having been live born but not as having died (according to either the SSBIDE database or the General Registrar’s Office death certificate database) in the first year of life were defined as surviving to 1 year of age. Preeclampsia was defined according to International Classification of Diseases, Ninth Revision, diagnostic codes in relation to post-delivery hospital discharge.14

Maternal age, parity, postcode of residence, and all outcome data were obtained solely from the SMR2. Data on maternal weight were obtained solely from the biochemical database. When possible, maternal height and smoking data were obtained from the SMR2; in instances in which this information was missing, the biochemical database was used. Smoking status (defined as the smoking status of the woman at the time of her first prenatal care visit) was determined as recorded in the patient’s case record. Maternal age was classified as the age of the mother at the time of delivery. Maternal weight was defined as that recorded at the time of Down syndrome screening. BMI (defined as weight in kilograms divided by height in meters squared) was categorized as lean (less than 20 kg/m2), normal (20–24.9 kg/m2), overweight (25–29.9 kg/m2), obese (30–34.9 kg/m2), and morbidly obese (35 kg/m2 or above).

Postcode of residence was used to calculate Carstairs socioeconomic deprivation values (higher values indicated greater deprivation). Deprivation classifications were based on 1991 census data on car ownership, unemployment, overcrowding, and social class within postcode sectors containing, on average, approximately 1600 residents.15 Since the early 1990s, gestational age has been confirmed (in the first half of pregnancy) using ultrasound in more than 95% of pregnancies in the United Kingdom.16 Gestational age at birth was defined as completed weeks of gestation on the basis of the estimated date of delivery from each woman’s clinical record, and standard national criteria exist for using menstrual and ultrasound data to estimate date of delivery. However, the specific means employed in a given record are not specified. Birthweight was categorized into gender-specific and gestational age–specific percentiles derived from the study cohort.

Statistical Analysis

We summarized continuous variables (age, height, and BMI) using medians and inter-quartile ranges, and we compared groups using the Kruskal–Wallis test. We made univariate comparisons of dichotomous data categories using the χ2 test or the Fisher exact test. All continuous variables were categorized. The level of statistical significance was set at P< .05 (2-sided). Logistic regression analyses were used to calculate adjusted odds ratios (ORs).17 Independent variables were BMI, age, height, deprivation category, smoking and marital status, and numbers of previous spontaneous early pregnancy losses and therapeutic abortions.

In analyses of birth outcomes for which the same women may have been included 2 or more times as a result of successive pregnancies, we estimated odds ratios using logistic regressions involving robust standard errors and clustering with maternal identifiers. We assessed interaction terms using the Wald test, as is appropriate for clustered data.17 We used Stata Version 8.2 (Stata Corp, College Station, Tex) to conduct all statistical analyses.


The linked database contained 227490 records of singleton births. Data for height were missing in 6270 cases (2.8%) and data for weight in 24835 cases (10.9%); in 26171 (11.5%) records, either or both of these values were missing. Among the remaining 201319 records, we excluded 206 (0.1%) deaths because of fetal abnormalities and 893 (0.4%) stillbirths because of other causes, leaving 200220 records. Of this total, birthweight data were missing or birthweights were less than 400 g in 57 cases (0.03%), and data on gestational age were missing or gestational age was outside 22 to 43 weeks in 62 cases (0.03%). Among the remaining 200104 records, maternal age was missing in 3 cases (less than 0.01%), parity was missing in 23 cases (0.01%), deprivation category was missing in 347 cases (0.2%), and smoking status was missing in 12487 cases (6.2%). Overall, 1 or more of these values were missing in 12814 records (6.4%), leaving a study sample of 187290 singleton births.

Table 1 [triangle] presents maternal characteristics and basic outcome data broken down by term delivery, spontaneous preterm delivery, and elective preterm delivery. All of the factors assessed varied among these 3 categories, although the highly statistically significant differences in maternal height actually reflected very small differences in mean height and the 3 groups had identical median values. Among women with preterm deliveries, elective delivery was associated with a reduced risk of neonatal death (relative risk [RR] = 0.72; 95% confidence interval [CI] = 0.55, 0.94; P= .02) and no overall increased risk of delivering an ELBW infant (RR = 1.06; 95% CI = 0.88, 1.28; P= .51). However, it was associated with an increased risk of delivering an ELBW infant who survived to 1 year of age (RR = 1.92; 95% CI = 1.49, 2.47; P< .001).

Maternal Characteristics and Pregnancy Outcome Data, by Type of Delivery: Cohort of Scottish Women, 1991–2001

In the case of all adverse outcomes, nulliparous women were at greater risk than multiparous women (Table 2 [triangle]). There were statistically significant interactions between nulliparity and BMI for overall preterm deliveries, spontaneous preterm deliveries, and elective preterm deliveries (all Ps<.001). There was a nonsignificant trend toward interactions between BMI and nulliparity for preeclampsia (P=.12), delivery of an ELBW infant (P=.06), and neonatal death (P=.23). All further analyses of outcomes related to BMI involved stratification according to parity.

Associations Between Parity and Pregnancy Outcomes: Cohort of Scottish Women, 1991–2001

Risk of spontaneous preterm deliveries decreased with increasing BMI, and this protective effect of increasing BMI was stronger among multiparous women (Figure 1a [triangle]; Tables 3 [triangle] and 4 [triangle]). By contrast, risk of elective preterm deliveries increased with increasing BMI, and the association was stronger among nulliparous women (Figure 1b [triangle]; Tables 3 [triangle] and 4 [triangle]). Therefore, the net effect of BMI on key outcomes associated with prematurity differed according to parity status. Among nulliparous women, a BMI of 35 or above was associated with increased risks of overall preterm birth, neonatal death, and delivery of an ELBW infant still alive at 1 year of age (Table 3 [triangle]). By contrast, multiparous women with a BMI of 35 or above were not at increased risk of any of these outcomes (Table 4 [triangle]).

Relationship between maternal body mass index (BMI) in early pregnancy in nulliparous and multiparous women and the proportion of spontaneous preterm births (a) and elective preterm births (b).
Body Mass Index (BMI) in Early Pregnancy and Outcome of First Pregnancy: Cohort of Scottish Women (n = 84 701), 1991–2001
Body Mass Index (BMI) in Early Pregnancy and Outcomes of Multiparous Women: Cohort of Scottish Women (n = 102 589), 1991–2001

Among nulliparous women with a BMI of 35 or above who had had an elective preterm delivery, 40.2% (49 of 122) had been diagnosed with preeclampsia; the corresponding percentage for multiparous women was 18.0% (25 of 139; P<.001). Preeclampsia had been diagnosed in 4917 (2.6%) women in the cohort overall. Among nulliparous women, neither neonatal death (adjusted OR=1.23; 95% CI=0.57, 2.66; P=.59) nor delivery of an ELBW infant who survived until 1 year of age (adjusted OR=1.75; 95% CI=0.81, 3.77; P=.15) was associated with obesity (BMI of 35 or above) after adjustment for gestational age at delivery. In the same group, adjustment for preeclampsia resulted in attenuation of the associations between obesity and elective pre-term delivery (adjusted OR=1.43; 95% CI=1.16, 1.75; P=.001), neonatal death (adjusted OR=2.43; 95% CI=1.35, 4.41; P=.003), and delivery of an ELBW infant who survived until 1 year of age (adjusted OR=2.52; 95% CI=1.40, 4.52; P=.002).

Of the original 227 490 records, 38 795 (17.1%) were excluded as a result of missing data for BMI, maternal age, parity, deprivation category, or smoking status. The rates of prematurity (5.84%) and low birthweight (5.65%) in this group were slightly higher than (but similar to) those of the study population. Among the group with missing data, 12 814 (33.0%) had a BMI recorded. We compared the relation between BMI (expressed as a continuous variable) and risk of prematurity in the group with missing data and the study population. The odds ratio for spontaneous preterm delivery associated with a 1-unit increase in BMI was 0.96 in both the group with missing data (95% CI = 0.94, 0.98; P < .001) and the study population (95% CI = 0.95, 0.96; P < .001). Odds ratios for elective preterm delivery were 1.03 (95% CI = 1.01, 1.05; P = .008) in the group with missing data and 1.02 (95% CI = 1.01, 1.04; P < .001) in the study population.


The main finding of this study is that obesity in early pregnancy is associated with an increased risk of elective preterm delivery. By contrast, obesity was associated with a decreased risk of spontaneous preterm delivery. The net effect of obesity depends, therefore, on the balance between these 2 outcomes. We found that morbidly obese nulliparous women had a more than 2-times greater risk of elective preterm delivery but only a 20% lower risk of spontaneous pre-term delivery. The net effect was that these women were at increased risk of all-cause prematurity, neonatal death, and delivery of an ELBW infant who survived to 1 year of age. These data indicate that morbidly obese women who are planning to conceive should be encouraged to lose weight before their first birth and that rising rates of morbid obesity in the prepregnant population are likely to lead to increased rates of severe morbidity and neonatal death related to prematurity.

This is the first study, to our knowledge, to demonstrate an increased risk of elective pre-term delivery among obese women. It has previously been shown that obese women are at lower risk of spontaneous preterm birth.46 Two recent studies analyzing the relation between BMI and elective preterm delivery did not demonstrate an overall association.4,18 The probable explanation for this apparent discrepancy is that data on nulliparous and multiparous women were pooled. In addition, both cohorts included fewer than 3000 women. The cohort used in our study was more than 50-times larger than the cohorts from these previous studies, and the highly statistically significant results indicate that the associations described are very unlikely to be chance findings.

Moreover, it is biologically plausible that such associations would be observed. Forty percent of morbidly obese nulliparous women who had had an elective preterm delivery had been diagnosed with preeclampsia, compared with only 2.6% of the remainder of the study population. Many previous studies have shown that preeclampsia risk increases with increasing BMI, and this effect is thought to be mediated by the cardiovascular influences of insulin resistance and dyslipidemia.19 We found that increasing BMI was associated with comparably increased relative risks of preeclampsia in nulliparous and multiparous women (Tables 3 [triangle] and 4 [triangle]). However, overall rates of preeclampsia were 3.9% and 1.6%, respectively, in these 2 groups (Table 2 [triangle]). The stronger association between obesity and elective preterm delivery among nulliparous women was probably because of these women’s higher background risk of preeclampsia.

Areas of Future Study

Among nulliparous women, obesity was not associated with risk of either neonatal death or delivery of an ELBW infant who survived to 1 year of age after adjustment for gestational age at delivery. This finding suggests that the association between obesity and these clinically important outcomes is mediated by obesity’s association with prematurity. Adjustment for preeclampsia resulted in marked, but not complete, attenuation of the associations observed between morbid obesity and elective preterm delivery, neonatal death, and delivery of an ELBW infant who survived to 1 year of age. The persistence of positive associations between morbid obesity and these outcomes after adjustment for preeclampsia may reflect errors in preeclampsia diagnoses, or, alternatively, other complications of pregnancy may be associated with obesity and may lead to an increased risk of these outcomes. This issue requires further study.

Many studies addressing factors associated with preterm labor lack either the data or the statistical power necessary to address the important consequences of prematurity. In addition to neonatal deaths, the record linkages used in the present study allowed us to identify long-term survivors whose birthweights were below 1000 g. Follow-up studies of these survivors in childhood demonstrated that 40% to 45% had severe neurodevelopmental impairments,9 as mentioned earlier, and this finding led to our designation of ELBW as a proxy for severe long-term morbidity. We demonstrated that morbidly obese nulliparous women were at increased risk of both perinatal mortality and perinatal outcomes likely to lead to severe morbidity. This underlines the clinical significance of the association with preterm delivery described here. Ideally, future studies will analyze risks of long-term severe morbidity directly rather than use a proxy measure.

The overall rate of prematurity in our study was relatively low, at 5.4%. This result is consistent with the findings of other European studies.7 By contrast, previous US studies have reported overall prematurity rates of 10% to 15%.4,18 However, these cohorts included 40% to 60% African American women and involved similarly high percentages of women who were unmarried or living in households with incomes below the poverty level. The present data are applicable to a relatively low-risk population. However, as observed in our comparisons of nulliparous and multiparous women, associations of birth outcomes with BMI depend on the relative balance of background risks of spontaneous and elective pre-term deliveries. Among nulliparous women at high risk of spontaneous preterm delivery, an increased BMI may be associated with a reduced overall risk of prematurity. Again, this is an issue for further study.


As is the case with any large-scale study in which routinely collected data are used, our study involved a number of weaknesses. The SMR2 database does not routinely collect data on maternal weight, and we were able to obtain this information only by linking records to a prenatal screening database. As a result, the population studied was selected on the basis of women having accepted screening for congenital abnormalities. However, 81% of women in the west of Scotland undergo serum screening,12 and thus, the study included most women seeking prenatal care.

Because maternal weight was used to adjust prenatal screening results, the value recorded was that from early pregnancy. As a result, we lacked data on prepregnancy weight and weight gain during pregnancy. However, our primary aim was to determine the probable effects of rising obesity rates in the general population on negative consequences of prematurity. BMI in early pregnancy is a good proxy for prepregnancy BMI, given that relatively little weight gain will have occurred between these intervals. Finally, approximately 17% of eligible women were excluded because of missing data, raising the possibility that our study population was biased. However, associations between BMI and spontaneous and elective preterm deliveries were similar when women with missing data on other maternal variables were compared with the study population.

Our results show that maternal obesity is associated with an increased risk of elective preterm delivery. The association is stronger among nulliparous women, probably as a result of their increased risk of preeclampsia, and here it led to an overall association between obesity and preterm birth in this group. Obese nulliparous women are at increased risk of the serious negative consequences associated with preterm births.


This research was funded by project grants from the Foundation for the Study of Infant Deaths (grants 245 and 249).

Human Participant Protection
The record linkage for this study was approved by the Privacy Advisory Committee, National Health Service for Scotland.


Peer Reviewed

G. C. S. Smith originated the study and drafted the article. G. C. S. Smith and I. Shah analyzed and interpreted the data. All of the authors contributed to critical revisions of the article.


1. Little RE, Weinberg CR. Risk factors for antepartum and intrapartum stillbirth. Am J Epidemiol. 1993; 137:1177–1189. [PubMed]
2. Brost BC, Goldenberg RL, Mercer BM, et al. The Preterm Prediction Study: association of cesarean delivery with increases in maternal weight and body mass index. Am J Obstet Gynecol. 1997;177:333–337. [PubMed]
3. Cedergren MI. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol. 2004;103:219–224. [PubMed]
4. Hendler I, Goldenberg RL, Mercer BM, et al. The Preterm Prediction Study: association between maternal body mass index and spontaneous and indicated pre-term birth. Am J Obstet Gynecol. 2005;192:882–886. [PubMed]
5. Schieve LA, Cogswell ME, Scanlon KS, et al. Prepregnancy body mass index and pregnancy weight gain: associations with preterm delivery. Obstet Gynecol. 2000;96:194–200. [PubMed]
6. Kramer MS, Coates AL, Michoud MC, Dagenais S, Hamilton EF, Papageorgiou A. Maternal anthropometry and idiopathic preterm labor. Obstet Gynecol. 1995;86: 744–748. [PubMed]
7. Cnattingius S, Bergstrom R, Lipworth L, Kramer MS. Prepregnancy weight and the risk of adverse pregnancy outcomes. N Engl J Med. 1998;338:147–152. [PubMed]
8. O’Brien TE, Ray JG, Chan WS. Maternal body mass index and the risk of preeclampsia: a systematic overview. Epidemiology. 2003;14:368–374. [PubMed]
9. Ohls RK, Ehrenkranz RA, Das A, et al. Neurodevelopmental outcome and growth at 18 to 22 months’ corrected age in extremely low birthweight infants treated with early erythropoietin and iron. Pediatrics. 2004;114:1287–1291. [PubMed]
10. Cole SK. Scottish maternity and neonatal records. In: Chalmers I, McIlwaine GM, eds. Perinatal Audit and Surveillance. London, England: Royal College of Obstetricians and Gynaecologists; 1980:39–51.
11. Scottish Perinatal and Infant Mortality Report 2000. Edinburgh, Scotland: Information and Statistics Division, National Health Service for Scotland; 2001.
12. Crossley JA, Aitken DA, Berry E, Connor JM. Impact of a regional screening programme using maternal serum α fetoprotein (AFP) and human chorionic gonadotrophin (hCG) on the birth incidence of Down’s syndrome in the west of Scotland. J Med Screen. 1994;1:180–183. [PubMed]
13. Kendrick S, Clarke J. The Scottish Record Linkage System. Health Bull (Edinb). 1993;51:72–79. [PubMed]
14. International Classification of Diseases, Ninth Revision. Geneva, Switzerland: World Health Organization; 1980.
15. McLoone P, Boddy FA. Deprivation and mortality in Scotland, 1981 and 1991. BMJ. 1994;309:1465–1470. [PMC free article] [PubMed]
16. Campbell S, Soothill P. Detection and management of intrauterine growth retardation: a British approach. In: Chervenak FA, Isaacson GC, Campbell S, eds. Ultrasound in Obstetrics and Gynecology. Boston, Mass: Little Brown & Co; 1993:1431–1435.
17. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons Inc; 2000.
18. Savitz DA, Dole N, Herring AH, et al. Should spontaneous and medically indicated preterm births be separated for studying aetiology? Paediatr Perinat Epidemiol. 2005;19:97–105. [PubMed]
19. Seely EW, Solomon CG. Insulin resistance and its potential role in pregnancy-induced hypertension. J Clin Endocrinol Metab. 2003;88:2393–2398. [PubMed]

Articles from American Journal of Public Health are provided here courtesy of American Public Health Association
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...