OBJECTIVE: To examine the maternal and foetal risks of adverse pregnancy outcome in relation to maternal obesity, expressed as body mass index (BMI, kg/m2) in a large unselected geographical population.
DESIGN: Retrospective analysis of data from a validated maternity database system which includes all but one of the maternity units in the North West Thames Region. A comparison of pregnancy outcomes was made on the basis of maternal BMI at booking.
SUBJECTS: A total of 287 213 completed singleton pregnancies were studied including 176 923 (61.6%) normal weight (BMI 20–24.9), 79 014 (27.5%) moderately obese (BMI 25–29.9) and 31 276 (10.9%) very obese (BMI≥30) women.
MEASUREMENTS: Ante-natal complications, intervention in labour, maternal morbidity and neonatal outcome were examined and data presented as raw frequencies and adjusted odds ratios with 99% confidence intervals following logistic regression analysis to account for confounding variables.
RESULTS: Compared to women with normal BMI, the following outcomes were significantly more common in obese pregnant women (odds ratio (99% confidence interval) for BMI 25–30 and BMI≥30 respectively): gestational diabetes mellitus (1.68 (1.53–1.84), 3.6 (3.25–3.98)); proteinuric pre-eclampsia (1.44 (1.28–1.62), 2.14 (1.85–2.47)); induction of labour (2.14 (1.85–2.47), 1.70 (1.64–1.76)); delivery by emergency caesarian section (1.30 (1.25–1.34), 1.83 (1.74–1.93)); postpartum haemorrhage (1.16 (1.12–1.21), 1.39 (1.32–1.46)); genital tract infection (1.24 (1.09–1.41), 1.30 (1.07–1.56)); urinary tract infection (1.17 (1.04–1.33), 1.39 (1.18–1.63)); wound infection (1.27 (1.09–1.48), 2.24 (1.91–2.64)); birthweight above the 90th centile (1.57 (1.50–1.64), 2.36 (2.23–2.50)), and intrauterine death (1.10 (0.94–1.28), 1.40 (1.14–1.71)). However, delivery before 32 weeks' gestation (0.73 (0.65–0.82), 0.81 (0.69–0.95)) and breastfeeding at discharge (0.86 (0.84–0.88), 0.58 (0.56–0.60)) were significantly less likely in the overweight groups. In all cases, increasing maternal BMI was associated with increased magnitude of risk.
CONCLUSION: Maternal obesity carries significant risks for the mother and foetus. The risk increases with the degree of obesity and persists after accounting for other confounding demographic factors. The basis of many of the complications is likely to be related to the altered metabolic state associated with morbid obesity.
Obesity is a significant health problem in the Western World with up to 40% of women in the UK being overweight.1 Several studies have reported that maternal obesity is associated with increased risk of adverse pregnancy outcomes including gestational diabetes, infectious morbidity, postpartum haemorrhage, delivery of large-for-dates babies and, more recently, stillbirth.2,3,4,5,6,7,8,9,10,11,12 However, many of the previous studies have lacked accurate quantification of such risks. This study develops other work especially regarding ante-natal complications of pregnancy. The aim of this study was to examine pregnancy outcome in obese women compared to those of normal weight by reviewing a large number of singleton pregnancies using a validated database.13 We aimed to test the hypothesis that obesity, as determined by maternal body mass index (BMI), is associated with adverse outcomes for mother and baby, and to quantify this risk after allowing for possible confounding factors.
Materials and methods
Data were derived from the St Mary's Maternity Information System database (SMMIS), including data from 1989 to 1997 inclusive. This is a clinical database recording maternity information from NHS hospitals within the geographical boundaries of the North West Thames Region and contains data on more than 80% of all deliveries in a region with a population of more than three and a half million. The database has been previously validated for commonly recorded variables including intrauterine death.13
BMI was calculated as weight at antenatal booking (kg), divided by height (m) squared. The subjects were divided into groups according to maternal BMI at booking using body mass groupings recommended by Abrams and Parker14 (normal, BMI 20–<25; moderately obese, BMI 25–<30, and very obese, BMI>30). In this study, women with BMI<20 were considered underweight and were excluded from further analysis (n=38 182).
The raw frequencies of the various outcomes of pregnancy in the maternal BMI groups were calculated and multiple logistic regression models were then constructed to examine the magnitude and significance of the independent effect of BMI. For each outcome variable, BMI (normal, moderately obese, very obese), maternal age, ethnic group, parity, history of hypertension and history of diabetes mellitus were included in all the models. All variables were categorical in the analytic models used. In addition, different factors of specific importance for each particular outcome were included as appropriate (see tables). The results are presented as frequencies by BMI group and odds ratios with 99% confidence intervals after adjustment for possible confounding factors. In order to correct for the influence of gestational age and foetal sex, birthweight was expressed as a delta value (the number of standard deviations by which the observed birthweight differed from the expected mean for sex and gestation). SAS Version 6.12 statistical analysis software was used. For the purposes of this study, pre-eclampsia was defined and documented as new onset hypertension and proteinuria, since the measurement of blood pressure alone may over represent the true prevalence, particularly if an inappropriate sized blood pressure cuff is used in obese subjects.
Information on maternal BMI was available for 325 395 completed singleton pregnancies including 79 014 (24.3%) moderately obese and 31 276 (9.6%) very obese. The distribution of BMI at booking is shown in Figure 1. Demographic characteristics of the maternal BMI groups are shown in Table 1.
The pregnancy outcome results by BMI group are summarised in Tables 2, 3, 4 and 5, with the proportion (%) of each risk factor or outcome being presented for each of the three maternal BMI groups. Odds ratios have been adjusted for the factors listed at the foot of the tables.
A history of diabetes mellitus and hypertension before the index pregnancy were more frequent in both the moderately and very raised BMI groups compared with women of normal BMI (Table 1). The odds of development of gestational diabetes and preeclampsia were increased in those with a raised BMI, even after allowing for demographic characteristics and histories of the above conditions (Table 2).
Induction of labour and delivery by caesarean section were both more common in obese women (Table 3). The frequency of both elective and emergency caesarean section was almost twice as high for very obese women as it was for women of normal BMI, which may have in part been related to the increased rate of induction of labour. The risk of postpartum haemorrhage rose with increasing BMI, and was about 30% more frequent in women with a moderately raised BMI and about 70% more frequent for women with very raised BMI compared with women of normal BMI. Although the assessment of postpartum haemorrhage is notoriously difficult the same increase in risk was demonstrated for blood loss of >1000 ml which would be considered clinically significant. All infectious morbidities were more prevalent in overweight women (Table 4).
Delta birthweight was positively associated with increasing maternal BMI at all gestational ages and the mean birthweight was significantly increased for the moderately and very obese groups compared to the normal BMI group (t=39.2, P<0.0001 and t=53.1, P<0.0001 respectively, Figure 2). Similarly, the prevalence of large for gestational age babies (birthweight>90th centile) was almost twice as high in the very obese compared to the normal group (Table 5), whereas obese women were at reduced risk of delivering a small for gestational age baby even after accounting for pregnancy complications such as pre-eclampsia. Table 5 also shows that the risk of stillbirth is increased in women with raised BMI and is significantly increased in those women with the highest BMI. The odds ratios adjusted for additional confounding variables reveal that raised BMI in itself conveys an increased risk of stillbirth.
This study has demonstrated that many adverse outcomes of pregnancy are associated with maternal obesity and has provided quantification of these risks. We have confirmed an increase in the previously reported complications of pregnancy in obese women such as gestational diabetes, induction of labour and wound infection. The risk of pre-eclampsia is positively associated with a raised BMI. This is an important observation because Reports on Confidential Enquiries into Maternal Death in the United Kingdom15 show that hypertensive disease remains among the commonest causes of maternal death and has shown no great improvement over the last 20 y. Changes in thromboembolic disease associated with obesity were not significant; even in this large study there were not many events. The indications for thromboemblolic preventative therapy vary across the region and with time through the study period.
The increase in caesarean section rate among obese women persists even allowing for confounding variables such as abnormal presentation of the foetus and medical complications of pregnancy. The caesarean section rate for the very obese group was over 20% compared with nearer 10% for the normal-weight group in this series. The relevance of the raised caesarean section rate in this group is considerable because of their increased risk of associated complications, such as infectious morbidity. The causes of this increased operative delivery rate cannot be determined but include increased planned elective caesarean section for predicted macrosomia, maternal request or obstetrician request. The increase in emergency caesarean sections may in part have been a consequence of the increased rate of large for gestational age infants leading to disproportion during labour or it is possible that uterine contractility may be suboptimal in a subgroup of obese women, or there may be increased fat deposition in the soft tissues of the pelvis. The increased risk of postpartum haemorrhage in obese women, even after accounting for such predisposing factors as caesarean section may be explained by more bleeding from the relatively larger area of implantation of the placenta usually associated with a large for gestational age foetus.
Foetal macrosomia is more common in the obese non-diabetic mother compared to the lean mother with gest-ational diabetes.16 The original Pedersen hypothesis17 suggested that increased glucose concentrations in the diabetic mother led to foetal hyperglycaemia and hyperinsulinaemia causing increased foetal growth. Obesity is associated with maternal insulin resistance and foetal hyperinsulinaemia even in the absence of maternal diabetes.18 Insulin resistant individuals have higher fasting plasma triglyceride levels and greater leucine turnover.19,20 Amino acids are insulin secretagogues and an increased flux on amino acids could stimulate foetal hyperinsulinaemia. Triglycerides are energy rich and placental lipases can cleave triglyceride and transfer free fatty acids to the foetus.21 The combination of an increased energy flux to the foetus and foetal hyperinsulinaemia may explain the increased frequency of large for gestational age infants seen in the obese non-diabetic women in this study.
Our study has also shown a small but significant increase in foetal death related to a raised maternal BMI, having made allowance for medical complications. The combination of rapid foetal growth induced by the endogenous hyperinsulinaemia in obese women and the functional limitations of the placenta to transfer sufficient oxygen to meet the requirements of the foetus, may lead to hypoxia and death in some cases. Unfortunately, because of known inconsistency in the recording of precise neonatal data, it has not been possible to determine whether there is also an increase in neonatal morbidity, but Lucas et al22 have reported that the relative risk of neonatal death is greater in preterm infants born to obese mothers than to thin women, and suggested that this may be secondary to the altered metabolic milieu in obesity reducing the infant's ability to adapt to postnatal life. Obese women in this study had a reduced risk of anaemia and placental abruption. The reason for the increased susceptibility to infectious disease is not described by these epidemiological data and may be a topic for further research.
Our findings are in agreement with, or have developed, previous studies. A study of 10 000 women in 1985 related ante-partum complications to varying degrees of maternal obesity.5 The authors found a significant (P<0.05) increase in hypertension, diabetes, thrombophlebitis and urinary tract infection. The largest study to date, from Sweden, investigated 167 750 women.11 Pre-pregnancy weight was related to risk of adverse pregnancy outcomes. The authors found higher maternal weight before pregnancy was associated with increased risk of late foetal death although it protected against the delivery of small gestational age infant. The Swedish workers had better outcome data in terms of foetal death but less information on antenatal complications as we have demonstrated in this study.
There are potential criticisms of this study. First, the proportion of women who booked after 20 weeks gestation was only about 8% for the normal group and about 13% and 14% respectively for the moderately and very obese groups. Since BMI was calculated using weight at booking, and weight increases with advancing pregnancy, late bookers may potentially have artificially contributed to an increase in the numbers in the raised BMI groups. This effect is however likely to be of small magnitude and, if present, would actually tend to reduce the effects attributed to obesity. Indeed, the significance of results is not altered if the analyses are repeated after adjusting for gestation at booking. Another major study11 used maternal memory of pre-pregnancy weight; we aimed to use a simple measure applicable in the ante-natal clinic to predict adverse pregnancy outcome. Our study did not include weight gain; as in other studies many women did not have weight measured twice, therefore weight gain may bias results. BMI is a better indicator of body composition than weight alone, being a more sensitive indicator of obesity in shorter women, more predictive of gestational diabetes and no less predictive than weight alone for any other outcomes.23
No information on social class was available for this study, however social class and obesity are related.24 Some of the outcome variables studied here also have an association with social class, particularly low birth weight,25 and breastfeeding.26 It is noteworthy that in this instance, raised maternal BMI has been shown to be protective against low birth weight, but the complex relationship between social class and pregnancy outcome in conjunction with a raised BMI cannot be determined from this study. None of the centres in this study have a practice where obese women receive more diagnostic scrutiny or receive differential diagnostic practice. Whilst this is a possibility, we do not believe it has altered these results.
Greater understanding is needed of the pathophysiological link between obesity and the various adverse outcomes of pregnancy we have described before effective and safe management strategies can be devised. At present one can only advise that it would be sensible to attempt to achieve nearer normal weight before conception.
Keil U, Kuulasmaa K . WHO MONICA Project: risk factors Int J Epidemiol 1989 18 (Suppl 1): S46–55.
Edwards LE, Dickes WF, Alton IR, Hakanson EY . Pregnancy in the massively obese: course, outcome and obesity prognosis of the infant Am J Obstet Gynecol 1978 131: 479–483.
Naeye RL . Weight gain and the outcome of pregnancy Am J Obstet Gynecol 1979 135: 3–9.
Gross T, Sokol RJ, King KC . Obesity in pregnancy: risks and outcome Obstet Gynecol 1980 56: 446–450.
Garbaciak JA, Richter M, Miller S, Barton JJ . Maternal weight and pregnancy complications Am J Obstet Gynecol 1985 152: 238–243.
Johnson SR, Kolberg BH, Varner MW, Railsback LD . Maternal obesity and pregnancy Surg Gynecol Obstet 1987 164: 431–437.
Mitchell MC, Lerner E . A comparison of pregnancy outcome in overweight and normal weight women J Am Coll Nutr 1989 8: 617–624.
Mancuso A, D'Anna R, Leonardi R . Pre pregnancy in the obese patient Eur J Obstet Gynecol Reprod Biol 1991 39: 83–86.
Perlow JH, Morgan MA, Montgomery D, Towers CG, Porto M . Perinatal outcome in pregnancy complicated by massive obesity Am J Obstet Gynecol 1992 167: 958–962.
Bianco AT, Smilen SW, Davis Y, Lopez S, Lapinski R, Lockwood CJ . Pregnancy outcome and weight gain recommendations for the morbidly obese woman Obstet Gynecol 1998 91: 97–102.
Cnattingius S, Bergstrom R, Lipworth L, Kramer MS . Prepregnancy weight and the risk of adverse pregnancy outcomes New Engl J Med 1998 338: 147–152.
De Groot LC . High maternal body weight and pregnancy outcome Nutr Rev 1999 57: 62–64.
Cleary R, Beard RW, Coles J, Devlin HB, Hopkins A, Roberts S, Schumacher D, Wickings HI . The quality of routinely collected maternity data Br J Obstet Gynaecol 1994 101: 1042–1047.
Abrams B, Parker J . Overweight and pregnancy complications Int J Obes 1988 12: 293–303.
HMSO . Report on Confidential Enquiries into Maternal Deaths in the United Kingdom 1991–1993 Department of Health. HMSO: London 1996.
Maresh M, Beard RW, Bray CS, Elkeles RS, Wadsworth J . Factors predisposing to and outcome of gestational diabetes Obstet Gynecol 1989 74: 342–346.
Pedersen J, Bojsen-Moller B, Poulsen H . Blood sugar in newborn infants of diabetic mothers Acta Endocrinol 1954 15: 33–52.
Hoegsberg B, Gruppuso PA, Coustan DR . Hyperinsulinemia in macrosomic infants of non diabetic mothers Diabetes Care 1993 16: 32–36.
Portman OW, Behrman RE, Soltys P . Transfer of free fatty acids across the primate placenta Am J Physiol 1969 216: 143–147.
Robinson S, Coldham N, Gelding SV, Murphy C, Beard RW, Halliday D, Johnston DG . Leucine flux is increased whilst glucose turnover is normal, in pregnancy complicated by gestational diabetes mellitus Diabetologia 1992 35 (Suppl 1): A683.
Thomas CR . Placental transfer of non-esterified fatty acids in normal and diabetic pregnancy Biol Neonate 1987 51: 91–101.
Lucas A, Morley R, Cole TJ, Bamford MF, Boon A, Crowle P, Dossetor JFB, Pearse R . Maternal fatness and viability of preterm infants Br Med J 1998 296: 1495–1497.
Wolfe HM, Zador IE, Gross TL, Martier SS, Sokol RJ . The clinical utility of maternal body mass index in pregnancy Am J Obstet Gynecol 1991 164: 1306–1310.
Stunkyard AJ, Sorensen TIA . Obesity and socioeconomic status—a complex relationship New Engl J Med 1993 329: 1036–1037.
Jonas O, Roder D, Chan A . The association of low socioeconomic status in metropolitan Adelaide with maternal demographic and obstetric characteristics and pregnancy outcome Eur J Epidemiol 1992 8: 708–714.
Kocturk T, Zetterstrom R . The promotion of breastfeeding and maternal attitudes Acta Paediatr Scand 1989 78: 817–823.
We would like to thank the North West Thames Obstetrics and Gynaecology Consultants for allowing us to use the data and all the midwives who were chiefly responsible for collecting the data. We would also like to thank Dr P Elliott, Dr J Chapple and Dr C Patterson, for their comments during the study.
Dedicated to the memory of Dr J Wadsworth.
About this article
Cite this article
Sebire, N., Jolly, M., Harris, J. et al. Maternal obesity and pregnancy outcome: a study of 287 213 pregnancies in London. Int J Obes 25, 1175–1182 (2001). https://doi.org/10.1038/sj.ijo.0801670
The American Journal of Clinical Nutrition (2020)
Defining maternal obesity in studies of birth outcomes: Comparing ICD‐9 codes at delivery and measures on the birth certificate
Paediatric and Perinatal Epidemiology (2020)
Social support modifies the association between pre-pregnancy body mass index and breastfeeding initiation in Brazil
PLOS ONE (2020)
Clinical Science (2020)
A Meta-Analysis of Stressors from the Total Environment Associated with Children’s General Cognitive Ability
International Journal of Environmental Research and Public Health (2020)