Five large case-control studies1–5 and one cohort study6 have reported that maternal obesity before pregnancy is associated with an increased risk of neural tube defects, and one study1 also suggested that maternal obesity may be associated with other types of central nervous system (CNS) birth defects. Maternal diabetes mellitus before pregnancy (pregestational diabetes) is an established risk factor for congenital malformations, including CNS birth defects.7–11 Poor glycemic control during the first trimester, as compared with later in pregnancy, is associated with the highest risk.12–15 Diabetes first diagnosed during pregnancy (gestational diabetes) may also be associated with some types of CNS malformations, although the evidence is less conclusive.16–19
Obesity and diabetes demonstrate similar metabolic abnormalities, including insulin resistance and hyperinsulinemia, and the increased risk of birth defects associated with each condition may share a common underlying metabolic disorder.2,20 We evaluated whether gestational diabetes influenced associations between maternal obesity and CNS birth defects, including neural tube defects (anencephaly and spina bifida), isolated hydrocephaly, and holoprosencephaly.
Because gestational diabetes is diagnosed later in pregnancy than when most birth defects occur, we considered it a potential marker for an underlying metabolic abnormality that occurs during early embryologic development.
The population-based Texas Birth Defects Monitoring Program of the Texas Department of Health actively ascertained all infants and fetuses with birth defects that were included in our study. Live births (with defects diagnosed at up to 1 year of age), late fetal deaths, and elective terminations of any gestational age were eligible for inclusion.21 Case mothers had pregnancies affected by holoprosencephaly, anencephaly, spina bifida, or isolated hydrocephalus. Diagnoses were based on the International Classification of Diseases, 9th Revision (ICD-9) and given hierarchical preference in the phenotypic order previously stated. Chromosomal abnormalities and syndromes of known etiology were excluded. Control mothers gave birth to live infants without structural birth malformations.
Our study, the Texas Interview for Prevention of Central Nervous System Birth Defects, was designed as an add-on component to the National Birth Defects Prevention Study. The national study, sponsored by the Centers for Disease Control and Prevention (CDC), is an 8-state case-control study that examines risk factors for 35 major birth defects; the study design and questionnaire development have been described previously.22 The Texas study evaluates only CNS birth defects. Women participating in the Texas study agreed to answer the telephone interview questionnaire for the national study and also agreed to complete a second add-on interview.
Geographic areas eligible for the Texas CNS study were phased-in, starting with births on January 1, 1997, from 6 of the 11 public health regions in Texas. As the Texas Birth Defects Monitoring Program expanded its surveillance, the Texas CNS study was expanded to include additional public health regions, and eligibility became statewide on July 1, 2000.
Controls were live births randomly selected from the same hospitals as cases. The number of controls from each hospital was based on the proportion contributed by that hospital to the total live births for a geographic area during the previous year. Our current study analyzes data only from the national study telephone interview questionnaire, not the add-on questionnaire.
Study participants had births or pregnancy terminations occurring from January 1997 through June 2001. Interviews were conducted a mean of 21 months (range, 7–36 months) after estimated conception date for cases and a mean of 20 months (range, 9–35 months) after estimated conception date for controls. Participation was approximately 60% for both cases and controls; 477 case mothers and 497 control mothers completed the 60-minute telephone interview.
Maternal prepregnancy body mass index (BMI) is given in kilograms/meter2 based on self-reported prepregnancy maternal weight and height. We used the analytic categories of BMI currently recommended by the National Heart, Lung, and Blood Institute and the World Health Organization (<18.5, 18.5–24.9, 25.0–29.9, 30.0–34.9, 35.0–39.9, and ≥40.0 kg/m2).23 Maternal obesity was defined as a prepregnancy BMI of ≥30.0 kg/m2. As a result of missing height or weight variables, we calculated BMI for only 410 (86%) case mothers and 416 (84%) control mothers. BMI was available for 42 (86%) holoprosencephaly, 98 (82%) anencephaly, 158 (86%) spina bifida, and 112 (84%) hydrocephaly cases. Missing BMI varied by maternal ethnicity and language of interview: white (7%), Hispanic who interviewed in English (18%), Hispanic who interviewed in Spanish (38%), and other ethnic groups combined (14%).
The variables of gestational age and birth outcome were obtained from medical records. All other variables, including the presence or absence of diabetes, as well as weight and height before pregnancy, were obtained from the telephone interviews. Diabetes was classified as pregestational (type 1 or type 2 diabetes present before pregnancy) or gestational (first diagnosed during pregnancy).
Odds ratios (ORs) and 95% confidence intervals (95% CIs) were used to estimate relative risks. If there were fewer than 5 combined cases or controls, or if there were no cases or controls in a category, the ORs were not calculated. We used multivariate logistic regression to control for potential confounding and to evaluate potential interaction.
Table 1 presents demographic factors and birth outcomes for cases and controls. Of the 477 cases, 120 were anencephaly, 184 spina bifida, 49 holoprosencephaly, and 124 isolated hydrocephalus. Most mothers were white or Hispanic. Among Hispanic women, Spanish as the language of interview was more likely for mothers of infants with anencephaly (36%) or spina bifida (26%) compared with control mothers (19%). Mothers of infants with holoprosencephaly or hydrocephaly were more likely to have a lower educational level as compared with control mothers.
Maternal diabetes and associated risk for types of CNS birth defects are presented in Table 2. Type 1 or type 2 diabetes before pregnancy (pregestational diabetes) was strongly associated with holoprosencephaly (adjusted OR = 47; 95% CI = 9.5–230) and isolated hydrocephaly (12; 2.9–47). Risk was not elevated for anencephaly (0.8; 0.1–9.1) or spina bifida (0.4; 0–4.4). Because pregestational diabetes is an established risk factor for CNS birth defects and our focus was on diabetes diagnosed during pregnancy (gestational diabetes), cases of pregestational diabetes were excluded from subsequent analyses.
The risk associated with gestational diabetes was increased only for holoprosencephaly (adjusted OR = 2.9; CI = 1.0–8.4) (Table 2). Mothers of infants with anencephaly were less likely to have gestational diabetes compared with control mothers (0.3; 0.1–1.2).
Associations for maternal BMI categories and types of CNS birth defects are shown in Table 3. Elevated risks were observed for maternal obesity (in the 3 categories of BMI ≥30 kg/m2) and all types of CNS birth defects. Risks were increased most consistently in the 2 highest BMI categories of maternal obesity for spina bifida and hydrocephaly. Underweight women (BMI <18.5 kg/m2) had a decreased risk for spina bifida (OR = 0.3, 95% CI = 0.1–1.1). Being underweight did not appear to be associated with risk for anencephaly, isolated hydrocephaly, or holoprosencephaly. Maternal obesity (BMI ≥30 kg/m2) was associated with an increased risk of neural tube defects and isolated hydrocephaly for all maternal ethnic groups (data not shown).
Table 4 shows the association of maternal obesity (BMI ≥30 kg/m2) and risk for CNS birth defects. After adjusting for potential confounders, maternal obesity (BMI ≥30 kg/m2) was associated with increased risks of delivering offspring with anencephaly (OR = 2.3; CI = 1.2–4.3), spina bifida (2.8; 1.7–4.5), or isolated hydrocephaly (2.7; 1.5–5.0).
Table 5 presents stratified analysis results for the interaction of the joint effects of maternal obesity (BMI ≥30 kg/m2) and gestational diabetes (after excluding pregestational diabetes). Obesity alone increased risk for anencephaly (adjusted OR = 2.0; CI = 1.1–3.6), spina bifida (2.6; 1.6–4.3), or hydrocephaly (2.3; 1.3–4.2). Gestational diabetes alone did not show increased risk for any CNS birth defects. Based on results for single exposures, expected odds ratios with multiplicative interaction for the joint exposures of maternal obesity and gestational diabetes are 2.1 for spina bifida, 1.4 for holoprosencephaly, and 4.4 for hydrocephaly. The results for the joint effects supported multiplicative interaction for spina bifida (adjusted OR = 4.5; 95% CI = 1.5–13) and holoprosencephaly (6.5; 1.3–31). The other CNS birth defects did not show evidence of interaction.
Obese women in our study had an increased risk of delivering offspring with isolated hydrocephaly. Only 1 other study has examined the association between maternal obesity and hydrocephaly.24 That study of 14 hydrocephaly cases found a 1.5-fold increased risk, possibly as a result of chance. Similar to previous studies, we observed that maternal obesity is associated with an increased risk for neural tube defects, that the risk with maternal obesity is greater for spina bifida than for anencephaly, and that the risk for spina bifida is decreased in underweight women.1–6 We also confirmed the previously reported findings that pregestational diabetes is associated with increased risk for holoprosencephaly and isolated hydrocephaly and that gestational diabetes is related to an increased risk of holoprosencephaly.7,9,16,17,25
For spina bifida and holoprosencephaly, the odds ratios for the joint effects of maternal diabetes and maternal obesity were consistent with multiplicative interaction, supporting the hypothesis that these risk factors operate through a shared causal mechanism. Hydrocephaly showed lack of evidence for interaction, suggesting independent causal mechanisms for these risk factors. Because of imprecise results, caution is indicated when considering these interpretations. Lack of evidence for interaction with anencephaly may be the result of bias. For anencephaly, 2.5% of mothers reported gestational diabetes compared with 5.9% or more for controls and other cases. This disparity is probably the result of early pregnancy terminations for anencephaly: 40% before 21 weeks gestation for anencephaly compared with 5% or less for other CNS birth defects.
The interactions we observed with gestational diabetes and maternal obesity are consistent with findings from a recent prospective study of 22,951 pregnant women.26 Maternal obesity (defined in that study as a BMI ≥28 kg/m2) and maternal diabetes of all types (including gestational diabetes) were not associated with increased risks for birth defects when examined as solitary exposures. However, mothers with both obesity and diabetes demonstrated a 3-fold increased risk for all birth defects combined compared with nonobese, nondiabetic mothers. Moore et al.26 interpreted these findings as suggestive of a synergistic effect for obesity and diabetes in the pathogenesis of birth defects. The sample size in that study was too small to evaluate individual types of congenital malformations such as CNS birth defects.
Potential limitations need to be considered when interpreting our findings. The 60% response rate for both case and control mothers raises the possibility that control mothers represent a self-selected population. For control mothers, 13.9% reported a BMI of ≥30 kg/m2 as compared with 26.3% for case mothers. Most (71%) control mothers were 29 years of age or less. The 1998 Behavioral Risk Factor Surveillance System for the United States found that among women 18 to 29 years of age, 12% reported a BMI ≥30 kg/m2, as did 17% of 30 to 39 year olds.27 Thus, obesity percentages of control mothers were reasonably similar to a comparable age group nationally, making selection bias unlikely as an explanation for our findings. Missing BMI variables were common for Hispanic mothers, especially those who interviewed in Spanish. However, the percent of cases (14%) with missing BMI values did not differ from controls (16%). Consequently, missing BMI is also an unlikely explanation for our findings.
Reasons for an association between maternal obesity and CNS birth defects are unknown. For neural tube defects, any exposure effect occurs early in pregnancy because neural tube formation is complete by day 28 in normal human embryos.28 Obese women may have some metabolic abnormality or nutritional deficiency that disrupts development of the embryo early in pregnancy. The reasons that maternal diabetes is associated with an increased risk for CNS birth defects are also unknown. As postulated for maternal obesity, the increased risk may be the result of a metabolic abnormality early in pregnancy. This hypothesis is supported by the observation that poor glycemic control early in pregnancy for mothers with pregestational or gestational diabetes increases the risk for some birth defects.7,29
Although gestational diabetes occurs too late in pregnancy to be considered a direct risk factor, it may be associated with other preexisting factors that are directly associated with the development of birth defects. A recent study that evaluated the association of gestational diabetes and chromosomal anomalies in offspring supports the hypothesis that gestational diabetes may be a marker for other underlying metabolic abnormalities.30 Moore et al. suggested that women who develop gestational diabetes might have underlying biochemical abnormalities related to the development of chromosomal defects early in pregnancy. Another recent study2 that evaluated postpartum hyperinsulinemia, present both with obesity and diabetes, and risk for neural tube defects also provides support for this hypothesis. Because hyperinsulinemia is associated with impaired glucose tolerance and diabetes, Hendricks et al.2 suggested that the increased risk for birth defects associated with hyperinsulinemia might be the result of an adverse event such as hyperglycemia at a critical time in early pregnancy.
Obesity and diabetes have metabolic abnormalities in common, including insulin resistance and hyperinsulinemia.20 Insulin resistance typically occurs with obesity, although individuals with normal weight may also have insulin resistance. There is evidence that an increased amount of centrally distributed fat may be more strongly related to insulin resistance than obesity per se. Also, the ability of pancreatic islet β-cells to respond to insulin resistance varies among individuals. Perhaps mothers who develop gestational diabetes are particularly susceptible to the adverse effects of obesity during pregnancy.
Obesity and the associated risks of diabetes are rapidly increasing in the United States. The Behavioral Risk Factor Surveillance System reported that from 1991 to 1998 the prevalence of obesity (BMI ≥30 kg/m2) in women 18 years of age and older increased from 12% to 18%.27 CNS birth defects are severe and often lethal malformations that affect more than 1 of every 1000 pregnancies in the United States.31–35 Our findings support the hypothesis that an increased risk for CNS birth defects is associated with maternal obesity and diabetes, including gestational diabetes. Our observations also support the hypothesis that maternal diabetes and obesity share a common mechanism for the risk for neural tube defects and perhaps holoprosencephaly. Although the underlying mechanisms for these relationships remain unclear, our findings provide support for the need for preventive measures to reduce the increased obesity and diabetes in the United States.
1. Waller DK, Mills JL, Simpson JL, et al. Are obese women at higher risk for producing malformed offspring? Am J Obstet Gynecol
2. Hendricks KA, Nuno OM, Suarez L, et al. Effects of hyperinsulinemia and obesity on risk of neural tube defects among Mexican Americans. Epidemiology
3. Watkins ML, Scanlon KS, Mulinare J, et al. Is maternal obesity a risk factor for anencephaly and spina bifida? Epidemiology
4. Werler MM, Louik C, Shapiro S, et al. Prepregnant weight in relation to risk of neural tube defects. JAMA
5. Shaw GM, Velie EM, Schaffer D. Risk of neural tube defect-affected pregnancies among obese women. JAMA
6. Kallen K. Maternal smoking, body mass index, and neural tube defects. Am J Epidemiol
7. Greene MF. Prevention and diagnosis of congenital anomalies in diabetic pregnancies. Clin Perinatol
8. Kucera J. Rate and type of congenital anomalies among offspring of diabetic women. J Reprod Med
9. Becerra JE, Khoury MJ, Cordero JF, et al. Diabetes mellitus during pregnancy and the risks for specific birth defects: a population-based case-control study. Pediatrics
10. Martinez-Frias ML. Epidemiological analyses of outcomes of pregnancy in diabetic mothers: identification of the most characteristic and most frequent congenital anomalies. Am J Med Genet
11. Towner D, Kjos SL, Leung B, et al. Congenital malformations in pregnancies complicated by NIDDM. Diabetes Care
12. Miller E, Hare JW, Cloherty JP, et al. Elevated maternal hemoglobin A1c in early pregnancy and major congenital anomalies in infants of diabetic mothers. N Engl J Med
13. Ylinen K, Aula P, Stenman UH, et al. Risk of minor and major fetal malformations in diabetics with high haemoglobin A1c values in early pregnancy. BMJ
14. Greene MF, Hare JW, Cloherty JP, et al. First-trimester hemoglobin A1 and risk for major malformation and spontaneous abortion in diabetic pregnancy. Teratology
15. Kitzmiller JL, Gavin LA, Gin GD, et al. Preconception care of diabetes. Glycemic control prevents congenital anomalies. JAMA
16. Ramos-Arroyo MA, Rodriguez-Pinilla E, Cordero JF. Maternal diabetes: the risk for specific birth defects. Eur J Epidemiol
17. Martinez-Frias ML, Bermejo E, Rodriguez-Pinilla E, et al. Epidemiological analysis of outcomes of pregnancy in gestational diabetic mothers. Am J Med Genet
18. Aberg A, Westbom L, Kallen B. Congenital malformations among infants whose mothers had gestational diabetes or preexisting diabetes. Early Hum Dev
19. Kalter H. Gestational diabetes mellitus and congenital malformations. Am J Obstet Gynecol
20. Kahn SE, Prigeon RL, Schwartz RS, et al. Obesity, body fat distribution, insulin sensitivity and Islet beta-cell function as explanations for metabolic diversity. J Nutr
21. State-Based Birth Defects Surveillance Programs. Texas Birth Defects Monitoring Division. Teratology
22. Yoon PW, Rasmussen SA, Lynberg MC, et al. The National Birth Defects Prevention Study. Pub Health Rep
. 2001;116(suppl 1):32–40.
23. Kuczmarski RJ, Flegal KM. Criteria for definition of overweight in transition: background and recommendations for the United States. Am J Clin Nutr
24. Watkins ML, Rasmussen SA, Honein MA, et al. Maternal obesity and risk for birth defects. Pediatrics
25. Croen LA, Shaw GM, Lammer EJ. Risk factors for cytogenetically normal holoprosencephaly in California: a population-based case-control study. Am J Med Genet
26. Moore LL, Singer MR, Bradlee ML, et al. A prospective study of the risk of congenital defects associated with maternal obesity and diabetes mellitus. Epidemiology
27. Mokdad AH, Serdula MK, Dietz WH, et al. The spread of the obesity epidemic in the United States, 1991–1998. JAMA
28. Van Allen MI, Kalousek DK, Chernoff GF, et al. Evidence for multi-site closure of the neural tube in humans. Am J Med Genet
29. Bartha JL, Martinez-Del-Fresno P, Comino-Delgado R. Gestational diabetes mellitus diagnosed during early pregnancy. Am J Obstet Gynecol
30. Moore LL, Bradlee ML, Singer MR, et al. Chromosomal anomalies among the offspring of women with gestational diabetes. Am J Epidemiol
31. Botto LD, Moore C, Khoury MJ, et al. Neural tube defects: medical progress. N Engl J Med
32. Hunter AGW. Brain, hydrocephalus. In: Stevenson RE, Hall JG, Goodman RM, eds. Human Malformations and Related Anomalies
, vol 2. Oxford: Oxford University Press; 1993:62–73.
33. Croen LA, Shaw GM, Lammer EJ. Holoprosencephaly: epidemiologic and clinical characteristics of a California population. Am J Med Genet
34. Rasmussen SA, Moore CA, Khoury MJ, et al. Descriptive epidemiology of holoprosencephaly and arhinencephaly in metropolitan Atlanta, 1968–1992. Am J Med Genet
35. Olsen CL, Hughes JP, Youngblood LG, et al. Epidemiology of holoprosencephaly and phenotypic characteristics of affected children: New York State, 1984–1989. Am J Med Genet