The obesity epidemic among young women has prompted research into obesity-related complications related to childbearing. Today, a number of observational studies link prepregnancy obesity to severe pregnancy complications, such as preeclampsia, preterm birth, and stillbirth,1–4 but less is known about severe complications for the offspring in the first year of life and later. Previous studies suggest an increased neonatal or infant mortality in offspring of obese women,1–3,5,6 but increased risks of preterm birth,6,7 birth asphyxia,2,4 and congenital malformations8–10 may partly explain these associations.
An increased risk of infant death in very preterm infants born to obese mothers has been reported in a British study.11,12 Also, a U.S. study based on data from the 1960s found an obesity-related increased risk of perinatal mortality, which was mainly explained by an increased risk of early preterm birth with intrauterine infection in obese women.13 A large study based on contemporary data from Sweden did not support these findings,14 and at present it is unclear whether an excess mortality among preterm infants is associated with maternal obesity.
In a previous study, we investigated the association between prepregnancy obesity and spontaneous preterm birth in the Danish National Birth Cohort.7 We found that obesity was associated with risk of preterm birth caused by preterm premature rupture of membranes (PROM) (hazard ratio 1.5, 95% confidence intervals 1.2–1.7), but not with preterm labor (hazard ratio 0.9, 0.8–1.1). Intrauterine infections play a key role in the cause of preterm PROM,15,16 and the excess risk found in obese women may be due to a higher susceptibility to inflammation or infection in this group. Such an interpretation is complicated by the fact that clinical subtypes of preterm birth have causation overlaps and that infection is also involved in the cause of preterm labor. However, for the fetus, the clinical manifestation of preterm PROM and preterm labor varies considerably. Whether survival in the offspring of obese women differs according to subtypes of preterm birth is not known, and insight into these relations may help us understand previous findings related to maternal obesity and neonatal mortality. We, therefore, used this large prospective cohort study to examine the association between prepregnancy obesity and neonatal mortality, with focus upon different subtypes of preterm birth.
MATERIALS AND METHODS
The Danish National Birth Cohort is a nationwide study of 100,000 pregnant women and their offspring. Approximately 50% of all general practitioners participated in the recruitment, which took place from 1996 to 2002. Pregnant women were approached at the first antenatal visit, and approximately 60% of the women accepted the invitation to participate in four computer-assisted telephone interviews, twice during pregnancy and twice after delivery.
In this cohort, 92,672 liveborn singletons were identified in the Civil Registration System where all Danish citizens are registered at birth and receive a personal identifier. The child’s personal identifier was linked to the Birth Register, which included information about birth weight, Apgar score after 5 minutes, congenital malformations, and gestational age at birth. The estimate of gestational age was based on the estimate reported at birth, which in more than 90% of cases is checked and adjusted according to early ultrasonography.17 All infants born alive before 37 completed weeks of gestation were identified, and information from the National Discharge Register was used to categorize preterm infants into three subtypes of preterm birth. Initially, spontaneous preterm births with preterm PROM were defined by the presence of the diagnosis O42 (International Classification of Diseases, 10th Revision). The remaining births were categorized as either induced preterm deliveries or spontaneous births without preterm PROM, based on whether induction or cesarean delivery before the onset of labor had been carried out. The prevalence of preterm PROM was 23% of all preterm births, which is in accordance with that reported in several large population-based studies.18,19 However, we knew that Danish obstetricians are reluctant to use the preterm PROM diagnosis in births with a short time span between spontaneous rupture of membranes and the onset of contractions, and therefore, we validated the preterm PROM diagnosis in 689 preterm births from three hospitals, for which independent data about the event that initiated birth were available. From two of these hospitals, we also had information about the exact time of events (n=360). Considering all preterm births with a preterm PROM diagnosis, the sensitivity was 50%, and the positive predictive value was 92%. For preterm PROM with more than 12 hours between rupture of membranes and onset of contractions, the sensitivity was high (92%), and the positive predictive value was 92%. Thus, the identified preterm PROM group in this study most probably represents a large proportion of the severe preterm PROM cases and does only include a few preterm PROM cases that may share causes with preterm labor.
In the study population, we included 85,375 liveborn singletons of mothers who participated in the first pregnancy interview at approximately 16 weeks of gestation and gave information about height and prepregnancy weight (available for 98.3%). Maternal body mass index (BMI) was calculated as prepregnancy weight in kilograms divided by height in meters squared and categorized according to the World Health Organization’s definitions as underweight (BMI less than 18.5 kg/m2), normal weight (BMI of 18.5 or more but less than 25), overweight (BMI of 25 or more but less than 30), and obese (BMI of 30 or more).20 From the pregnancy interview, we also obtained information on parity, mother’s age, waiting time to pregnancy, infertility treatment, smoking, alcohol intake in the first part of pregnancy, and sociooccupational status. The categorization of these and other variables are displayed in Table 1.
Pregnancy complications associated with obesity, defined as preeclampsia, hypertension, and diabetes, were identified by International Classification of Diseases, 10th Revision, codes in the National Discharge Register. We expected some underreporting of gestational diabetes in the National Discharge Register, and for that reason we added self-reported information on the disease from the pregnancy interviews.
The main outcome was neonatal mortality, defined as death of a liveborn infant within the first 28 days of life. Information about mortality was obtained from the Birth Register.
Cox regression models were used to analyze neonatal mortality. In these analyses, the relative risk of neonatal death was estimated as a mortality rate ratio (hazard ratio). All infants were followed up from birth to death or censored after 28 days of life. All models were stratified on gestational week at first pregnancy interview to improve comparability.
We first described the association between the mother’s prepregnancy BMI and the child’s survival by computing crude and adjusted hazard ratios relative to normal-weight mothers. In the adjusted analysis, we considered two models. In the first model, we corrected for these maternal factors: age, parity, cigarette smoking in pregnancy, and sociooccupational status. To investigate whether neonatal factors may be in the causal pathway between BMI and neonatal mortality, we extended the adjustment in a second model to also include gestational age in days at delivery as a continuous variable, Apgar score at 5 minutes (three categories), and the presence of malformations. Variable selection was made a priori, based on knowledge about risk factors for neonatal mortality and on the numbers of neonatal deaths in the study.
In preterm infants, we investigated the association between BMI and neonatal mortality after controlling for maternal and neonatal factors as described above. Moreover, we investigated whether this association was modified by the subtype of preterm birth. Due to small numbers in this analysis, we restricted the adjustment to neonatal factors.
In supplementary analyses, we evaluated different analytic strategies. We investigated whether the association between BMI and neonatal mortality was modified by parity or by the presence of malformations by adding interaction terms to the model. Effect modification was assessed by computing ratios of hazard ratios. No effect modification was present if this ratio did not differ significantly from 1. We also repeated all analyses after exclusion of infants with malformations. The role of fetal growth restriction was examined by adding to the models standardized birth weight for gestational age (z-score), defined according to the reference curve given by Marsal et al.21 Because obesity-related risks of adverse pregnancy outcomes may be explained by obesity-related diseases,1–3 and subfertility is also associated with both obesity and infant mortality,22 we repeated all analyses after the exclusion of infants born to women with obesity-related diseases and with exclusion of infants born to women who received infertility treatment or had a waiting time to pregnancy of more than 1 year. A correction for within-cluster correlation (robust standard errors) was applied in all reported analyses, because 5,985 women contributed more than one pregnancy to the study. A two-tailed significance level of .05 was used in all statistical tests, and estimates are presented with 95% confidence intervals. All analyses were made using STATA 9.0 (StataCorp, College Station, TX). The study was approved by the Scientific Ethics Committee for Copenhagen and Frederiksberg on behalf of all the committees in Denmark, by the Danish National Board of Health, and by the Danish Data Protection Board.
Eight percent of infants were born to mothers who were obese before pregnancy, 19% were born to overweight mothers, 68% were born to normal-weight mothers, and 5% were born to underweight mothers. Compared with normal-weight mothers, mothers who were overweight or obese tended to be younger, smoked more, drank less alcohol, and had a lower social status (Table 1). Subfertility was more frequent among overweight and obese women, as were gestational diabetes and hypertensive disorders during pregnancy. Infants of obese women were more often preterm, had slightly more malformations, and had lower Apgar scores.
We observed 230 neonatal deaths among liveborn infants. Compared with normal-weight mothers, the crude neonatal mortality was elevated in infants of overweight and obese mothers, and adjustment for maternal and neonatal risk factors had only little effect on the findings (adjusted hazard ratios 1.7, 95% CI 1.2–2.5, and 1.6, 95% CI 1.0–2.4, respectively) (Table 2). Among preterm infants, a similar increase in mortality related to high maternal BMI was observed, although the findings related to maternal obesity were not statistically significant (Table 2). The association between a high BMI and neonatal mortality was independent of parity or the presence of malformations (P for no interaction=.56 and .63, respectively).
In the analysis of preterm infants categorized according to subtypes, we found that maternal overweight and obesity were associated with increased risks of neonatal mortality in infants born after preterm PROM (Table 3). Relative to infants of normal-weight women with preterm PROM, infants of overweight and obese women faced a more than tripled and an almost sixfold increase in the risk of neonatal mortality in the adjusted analyses. In contrast, there was no association between high BMI and neonatal mortality in infants born after spontaneous preterm birth without preterm PROM. Thus, the associations between overweight and obesity and neonatal death were significantly different in infants born after spontaneous preterm birth with and without preterm PROM (P=.03 and P=.001, respectively). The ratio of the hazard ratios (reflecting risk of neonatal death in preterm births with preterm PROM relative to no preterm PROM) were 4.1 (95% CI 1.2–14.4) for overweight and 7.5 (95% CI 2.2–25.3) for obese mothers. Also, the association between maternal obesity and neonatal mortality in infants born after preterm PROM was stronger than the corresponding association in infants born after induced preterm deliveries (ratio of hazard ratio 6.1, 95% CI 1.55–23.6), P=.01). We repeated the analysis of preterm subtypes in a population of infants without malformations (Table 4). This restriction limited the number of neonatal deaths in the analyses to 87, but the results were similar to those based on all preterm infants. In particular, the interaction between obesity and subtype of preterm birth persisted (P=.004).
The analyses were repeated after exclusion of infants born to women with obesity-related diseases or infants born to subfertile women. The hazard ratios according to categories of prepregnancy BMI (Table 2) were in both cases very similar, the largest change being 8%. Larger changes (up to 32%) were seen for the more imprecise estimates in the analysis of prepregnancy BMI and subtypes of preterm birth (Tables 3 and 4), but the conclusions remained the same.
When birth weight standardized for gestational age was included in the models as a continuous covariate, it had no effect on the association between BMI and neonatal mortality. However, it was notable that all neonatal deaths in preterm infants of obese mothers happened in neonates born before 34 weeks, which was not seen in the other BMI groups (Table 5). Also, deceased infants born before 34 weeks of gestation of overweight and obese women were less growth restricted than those born to overweight and obese mothers that survived the neonatal period. A similar pattern was not observed among infants born to normal-weight and underweight women. However, these tendencies were not statistically significant.
Neonatal mortality was increased in infants of overweight and obese mothers, which is in accord with the results from other studies,1–3,5,6,11,13 but in preterm infants, the increase was of a somewhat weaker magnitude than that reported by Lucas et al.11 This may reflect true differences between the two studies in BMI thresholds or degrees of preterm birth or may be due to chance.
We took into account maternal risk factors and diseases, but the excess risk related to high BMI persisted. Short gestation, birth complications leading to asphyxia, and congenital malformations may be part of the causal pathway between maternal obesity and neonatal death. Controlling for these important neonatal predictors of survival may, therefore, be considered overadjustment, but the findings were also unchanged when we included this information.
Growth restriction in very preterm infants was associated with obesity in the British study,11 but we did not find this in our study. On the contrary, among overweight and obese women, very preterm infants who died in the neonatal period seemed to be larger for a given gestational age than the surviving preterm infants. This suggests that growth restriction does not explain the excess mortality related to maternal obesity, and if the association is causal, the underlying mechanisms remain speculative. Prepregnancy obesity has been associated with a range of metabolic, inflammatory, and vascular abnormalities during pregnancy, even in pregnant women without clinical disease.23,24 Perhaps endothelial dysfunction and inflammatory up-regulation due to obesity accelerate the timing and severity of some pregnancy complications that may lead to the birth of more vulnerable infants among obese women. It may be a coincidence, but all deceased preterm infants of obese mothers in our study were born before 34 weeks of gestation.
The observed increased neonatal mortality in infants of women with a high prepregnancy BMI and preterm PROM may be consistent with the findings of Naeye,13 who linked an increased perinatal mortality in offspring of obese mothers to early preterm birth related to chorioamnionitis. The interpretation is, however, complicated by the fact that infection is thought also to be involved in the cause of preterm labor. In this connection, it is important to consider that the effect of infection on the fetus may be much more severe when the fetal membranes have ruptured, especially with long time spans between rupture of membranes and labor. Perhaps it is at this stage that maternal obesity may impair the survival of the fetus, either through biologic mechanisms or as a genetic marker. Regrettably, the current data do not allow us to investigate these hypotheses. We evaluated the cause of death in all deceased preterm infants by means of the Register of Cause of Death and death certificates, but apart from congenital malformations, most deaths were stated to be a direct consequence of prematurity, such as respiratory and cardiovascular collapse. Better and more detailed clinical data are necessary to allow a thorough investigation of cause of death in preterm infants.
We relied on self-reported information about prepregnancy BMI and potential confounders. Because this information was collected prospectively, we expected any misreporting to be nondifferential, which would most likely lead to an underestimation of the true effect. Ascertainment of pregnancy outcomes, malformations, and infant mortality was primarily based on national registers. We found a high consistency between various sources of information and believe that any misclassification was minor or nondifferential, which would most likely lead to an underestimation of the true effect of overweight and obesity.
The categorization of preterm birth into subtypes was also based on register information, and our validation study indicated that we reliably identified a large proportion of the severe preterm PROM cases. However, it is problematic that published frequencies of preterm PROM vary from 7.1 to 51.2%,15,19,25 probably because the official definition of preterm PROM does not include a minimal time interval between spontaneous rupture of membranes and the onset of labor,26 nor does it contain a proper definition of being in labor. This impairs the use of subtypes of preterm birth as outcomes or predictors, but alternatives are lacking. The replication of our previous findings of heterogeneity in the association between prepregnancy obesity and subtypes of spontaneous preterm birth,7 but now related to neonatal mortality, justifies our choice27 and also adds credibility to the findings. Obese women may have more severe types of preterm PROM that leave the infant more susceptible to neonatal complications and death. However, this hypothesis needs further investigation.
Despite the large size of the Danish National Birth Cohort, the number of neonatal deaths was small, and the new findings related to preterm PROM need to be replicated in a larger data set with more detailed information. However, the results were statistically significant and insensitive to adjustment for a variety of other explanatory factors.
The study population was white, with a low risk of preterm birth and neonatal mortality, and the public health implications of the findings for this groups are limited. Thus, the absolute risk difference between infants of normal-weight and obese mothers were less than 2 per 1,000 liveborn infants. However, in the subpopulation of preterm infants born after preterm PROM, the risk difference was as high as 56 infants per 1,000 liveborn, and the findings may be relevant for clinicians concerned with the management of preterm birth and the care for preterm infants. Whether the findings apply to other ethnic populations or to populations with higher risks of obesity, preterm birth, and neonatal mortality needs to be examined.
The obesity epidemic in women of childbearing age is now present in many parts of the world, and it is important to consider the effect of such serious change on neonatal survival, which is one of the most sensitive indicators of maternal health. It has recently been suggested that the incidence of spontaneous preterm births is also increasing.28 These trends underline the potential importance of the findings reported in the present study.
1. Baeten JM, Bukusi EA, Lambe M. Pregnancy complications and outcomes among overweight and obese nulliparous women. Am J Public Health 2001;91:436–40.
2. Cedergren MI. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol 2004;103:219–224.
3. Cnattingius S, Bergstrom R, Lipworth L, Kramer MS. Prepregnancy weight and the risk of adverse pregnancy outcomes. N Engl J Med 1998;338:147–52.
4. Sebire NJ, Jolly M, Harris JP, Wadsworth J, Joffe M, Beard RW, et al. Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London. Int J Obes Relat Metab Disord 2001;25:1175–82.
5. Lu GC, Rouse DJ, DuBard M, Cliver S, Kimberlin D, Hauth JC. The effect of the increasing prevalence of maternal obesity on perinatal morbidity. Am J Obstet Gynecol 2001;185:845–9.
6. Smith GC, Shah I, Pell JP, Crossley JA, Dobbie R. Maternal obesity in early pregnancy and risk of spontaneous and elective preterm deliveries: a retrospective cohort study. Am J Public Health 2007;97:157–62.
7. Nohr EA, Bech BH, Vaeth M, Rasmussen KM, Henriksen TB, Olsen J. Obesity, gestational weight gain and preterm birth: a study within the Danish National Birth Cohort. Paediatr Perinat Epidemiol 2007; 21:5–14.
8. Cedergren MI, Kallen BA. Maternal obesity and infant heart defects. Obes Res 2003;11:1065–71.
9. Waller DK, Mills JL, Simpson JL, Cunningham GC, Conley MR, Lassman MR, et al. Are obese women at higher risk for producing malformed offspring? Am J Obstet Gynecol 1994;170:541–8.
10. Watkins ML, Rasmussen SA, Honein MA, Botto LD, Moore CA. Maternal obesity and risk for birth defects. Pediatrics 2003;111:1152–8.
11. Lucas A, Morley R, Cole TJ, Bamford MF, Boon A, Crowle P, et al. Maternal fatness and viability of preterm infants. Br Med J (Clin Res Ed) 1988;296:1495–7.
12. Lucas A, Gore SM, Cole TJ, Bamford MF, Dossetor JF, Barr I, et al. Multicentre trial on feeding low birth weight infants: effects of diet on early growth. Arch Dis Child 1984;59:722–30.
13. Naeye RL. Maternal body weight and pregnancy outcome. Am J Clin Nutr 1990;52:273–279.
14. Johansson S, Montgomery SM, Ekbom A, Olausson PO, Granath F, Norman M, et al. Preterm delivery, level of care, and infant death in Sweden: a population-based study. Pediatrics 2004;113:1230–5.
15. Furman B, Shoham-Vardi I, Bashiri A, Erez O, Mazor M. Clinical significance and outcome of preterm prelabor rupture of membranes: population-based study. Eur J Obstet Gynecol Reprod Biol 2000;92:209–16.
16. Goldenberg RL, Culhane JF. Infection as a cause of preterm birth. Clin Perinatol 2003;30:677–700.
17. Jorgensen FS. Ultrasonography of pregnant women in Denmark 1999–2000. Description of the development since 1980–1990 [in Danish]. Ugeskr Laeger 2003;165:4409–15.
18. Buchmayer SM, Sparen P, Cnattingius S: Previous pregnancy loss: risks related to severity of preterm delivery. Am J Obstet Gynecol 2004;191:1225–31.
19. Savitz DA, Blackmore CA, Thorp JM. Epidemiologic characteristics of preterm delivery: etiologic heterogeneity. Am J Obstet Gynecol 1991;164:467–71.
20. Obesity: preventing and managing the global epidemic. Report on a WHO Consultation. Technical Report Series, No 894. Geneva (Switzerland): WHO; 2000. p. 1–253.
21. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatrica 1996;85:843–8.
22. Basso O, Olsen J. Subfecundity and neonatal mortality: longitudinal study within the Danish national birth cohort. BMJ 2005;330:393–4.
23. Ramsay JE, Ferrell WR, Crawford L, Wallace AM, Greer IA, Sattar N. Maternal obesity is associated with dysregulation of metabolic, vascular, and inflammatory pathways. J Clin Endocrinol Metab 2002;87:4231–7.
24. Retnakaran R, Hanley AJ, Raif N, Connelly PW, Sermer M, Zinman B. C-reactive protein and gestational diabetes: the central role of maternal obesity. J Clin Endocrinol Metab 2003;88:3507–12.
25. Berkowitz GS, Blackmore-Prince C, Lapinski RH, Savitz DA. Risk factors for preterm birth subtypes. Epidemiology 1998;9:279–85.
26. Lee T, Silver H. Etiology and epidemiology of preterm premature rupture of the membranes. Clin Perinatol 2001;28:721–34.
27. Savitz DA, Dole N, Herring AH, Kaczor D, Murphy J, Siega-Riz AM, et al. Should spontaneous and medically indicated preterm births be separated for studying aetiology? Paediatr Perinat Epidemiol 2005;19:97–105.
© 2007 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
28. Langhoff-Roos J, Kesmodel U, Jacobsson B, Rasmussen S, Vogel I. Spontaneous preterm delivery in primiparous women at low risk in Denmark: population based study. BMJ 2006;332:937–9.