ARTICLE |
a Arkansas Center for Birth Defects Research and Prevention
b Department of Pediatrics, College of Medicine, University of Arkansas for Medical Sciences, Little Rock, Arkansas
c Arkansas Children's Hospital Research Institute, Little Rock, Arkansas
d National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention, Atlanta, Georgia
| ABSTRACT |
|---|
|
|
|---|
METHODS. Using data from population-based birth defect registries, the National Birth Defects Prevention Study enrolled infants with nonsyndromic congenital heart defects (case subjects) and infants without congenital heart defects or any other birth defect (control subjects). Small for gestational age was defined as birth weight below the 10th percentile for gestational age and gender. Association between congenital heart defects and small for gestational age was examined by conditional logistic regression adjusting for maternal covariates related to fetal growth.
RESULTS. Live-born singleton infants with congenital heart defects (case subjects, n = 3395) and live-born singleton infants with no birth defect (control subjects, n = 3924) were included in this study. Case subjects had lower birth weights compared with control subjects. Small for gestational age was observed among 15.2% of case subjects and among only 7.8% of control subjects. Congenital heart defect infants were significantly more likely to be small for gestational age than control infants.
CONCLUSIONS. Infants with congenital heart defects are approximately twice as likely to be small for gestational age as control subjects. Small for gestational age status may affect clinical management decisions, therapeutic response, and prognosis of neonates with congenital heart defects. Although the etiology of growth retardation among infants with congenital heart defects is uncertain, further exploration may uncover a common pathogenesis or causal relationship between congenital heart defects and small for gestational age.
Key Words: small for gestational age cardiac disease congenital anomalies
Abbreviations: CHDcongenital heart defect NBDPSNational Birth Defects Prevention Study ORodds ratio CIconfidence interval
Congenital heart defects (CHDs) are the most prevalent birth defects and frequently require multiple hospitalizations and surgical procedures. Prevalence estimates range from 8 to 11 per 1000 live births.1,2 Thus, in the United States,
32000 to 44000 infants are born with a cardiac defect each year.1 Many infants will require corrective or palliative surgery and extensive hospitalizations during the first year of life.3,4 Medical and surgical outcomes are dependent on the complexity of the cardiac lesion, and other infant characteristics, such as lung development, prematurity, and birth weight.5,6 After corrective or palliative surgery, multiple studies have documented failure to grow in infants with CHDs because of increased metabolic demands and decreased caloric intake.6 The degree of preoperative growth failure has been associated with longer time on the ventilator, difficulty feeding postoperatively, a higher risk of infection, longer hospitalizations, and poor postoperative catch-up growth.59
Although the association between CHDs and low birth weight or small for gestational age has been described in clinical and pathologic case series and in case-control studies, few studies have been population based and/or controlled for maternal and fetal determinants of growth.1015 The estimated risk of small for gestational age in large population-based studies has varied by the type of cardiac defects. Published reports indicate that infants with CHDs are
1.8 to 3.6 times more likely to be small for gestational age than infants without CHDs.11,16
The etiology of most nonsyndromic CHDs is unknown but likely involves a complex interaction between multiple environmental exposures and genetic susceptibilities. Similarly, the etiology of small for gestational age is complex. The temporal relationship between CHDs and small for gestational age cannot be discerned from case series or case-control studies. It is possible that both CHDs and small for gestational age occur independently but share common risk factors. For example, maternal periconceptional smoking is the leading cause of small for gestational age in developed countries.1721 Similarly, some case-control studies provide evidence indicating that maternal periconceptional smoking increases the occurrence of CHDs.16,2225 Elevated total maternal homocysteine levels have been shown to be associated with an increased risk of small for gestational age26 in infants. Our group has also reported a higher incidence of CHDs associated with high maternal homocysteine levels, suggesting a common pathway25,27 related to maternal nutrition. Alternatively, it is possible that small for gestational age in infants with CHDs may be attributed to abnormal fetal hemodynamics.28,29 The objective of our study was to investigate the association between CHDs and small for gestational age in a large population-based sample while controlling for multiple maternal and fetal covariates.
| METHODS |
|---|
|
|
|---|
For the investigation reported herein, infants born with CHDs were included if they met the following criteria: (1) the infant had no known single gene disorder or chromosomal abnormality; (2) the diagnosis of a CHD occurred before the child was 1 year of age and was based on an echocardiogram, heart catheterization, surgical or autopsy report; (3) the mother spoke English or Spanish; (4) the infant was not placed for adoption or in foster care; (5) they were singleton births; and (6) they had no extracardiac anomalies. Control subjects were infants with no birth defects randomly selected from birth certificates or hospital discharge listings in the same states as case subjects. During the study period, the birth population covered by the 8 states was
482000 births per year. Approximately 18% of eligible case mothers and 21% of control mothers refused to participate in the NBDPS study during this time period.
Classification of Cardiac Defects
Each CHD case was reviewed by 1 of 4 NBDPS case classifiers31 and described as "simple" or "associated" based on the complexity of the lesion. Cardiac defects were classified into 4 categories based on the anatomic lesion: (1) conotruncal, including transposition of the great arteries, tetralogy of Fallot, truncus arteriosus, double outlet right ventricle, malaligned ventricular septal defect, and interrupted aortic arch type B; (2) septal, including atrial, ventricular, and atrioventricular septal defects; (3) right-sided obstructive, including pulmonary valve stenosis, pulmonary atresia, and tricuspid atresia; and (4) left-sided obstructive, including aortic valve stenosis, hypoplastic left heart syndrome and variants, coarctation of the aorta, and interrupted aortic arch types A and C lesions. Complex CHD lesions, which included total and partial pulmonary venous connection, double inlet left ventricle, and other lesions with
3 cardiac defects, were excluded from this analysis.
Maternal and Infant Covariates
As part of the NBDPS, mothers of case and control subjects completed an extensive interview regarding multiple preconceptional, periconceptional, and pregnancy exposures. A detailed list of exposures included in these interviews has been published previously.30 Maternal factors collected for the NBDPS and known to be associated with small for gestational age included maternal socioeconomic status, age, race/ethnicity,32 gestational weight gain, prepregnancy BMI, parity, preexisting diabetes, hypertension, infections such as cytomegalovirus or rubella smoking, alcohol consumption, and use of marijuana or cocaine.3134 Infant variables studied included infant gender33 and race/ethnicity.35
Growth Parameters
Birth weight was obtained from the birth certificate and added to the NBDPS clinical database. Small for gestational age was defined as a birth weight less than the 10th percentile for a given gestational age and gender based on a standardized birth weight distribution of US live births.36,37 Gestational age was obtained from the clinical database; if this was missing, we used a calculated variable using expected date of delivery.
Statistical Methods
Maternal and infant characteristics measured on an interval scale were compared between CHD case and control subjects using the Mann-Whitney/Wilcoxon rank sum test. The frequency of small for gestational age was computed for normal control subjects, CHD case subjects, and each CHD subtype independently. Multiple linear regression was used to model birth weight as a function of case-control status and maternal and infant covariates. Adjusted means were computed from the fitted model. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the association between small for gestational age and CHDs were estimated using conditional logistic regression adjusting for infant and maternal factors and grouping on mother's state of residence. Covariates for inclusion were determined based on results of the bivariate analyses and evidence published previously.
| RESULTS |
|---|
|
|
|---|
In this NBDPS cohort, the vast majority 2796 (82.4%) had a simple cardiac lesion. Septal heart defects were the most common malformation (44.4%), followed by conotruncal (22.4%), right-sided obstructive (16.8%), and left-sided obstructive (16.3%) defects.
Selected infant and maternal characteristics of case and control subjects are presented in Table 1. There were no significant differences between case and control subjects for maternal education, total household income, parity, or cocaine or alcohol use. Maternal hypertension during the index pregnancy was more frequent among case subjects than control subjects (OR: 1.24; 95% CI: 1.061.45). Case subjects were also more likely to smoke during the pregnancy than control subjects (OR: 1.22; 95% CI: 1.041.43). Mean prepregnancy weight is shown in Table 1 to be higher for case subjects than control subjects (68.0 vs 66.4 kg; P = .014), and case subjects gained significantly less weight during pregnancy than control subjects (13.7 vs 14.5 kg; P = .002). Infants with CHDs were >3 times more likely to be premature than control subjects (OR: 3.16; 95% CI: 2.693.70) and less likely to be female (OR: 0.85; 95% CI: 0.780.94). The mean gestational age for case subjects was 38 weeks, and the mean gestational age for control subjects was 39 weeks. After controlling for gestational age and infant and maternal risk factors, CHD case subjects had lower mean birth weights (3039.56 g; SE: 35.8) compared with control subjects (3128.42 g; SE: 35.5; P < .0001).
|
|
|
| DISCUSSION |
|---|
|
|
|---|
Being small for gestational age is usually defined as <10% for a given gestational age, and, thus, one may expect more control infants to be small for gestational age than found in our study sample. However, we excluded twins and higher-order multiples and infants with syndromes, which may explain why only 7.8% of the control subjects were small for gestational age.
The nature of the etiologic and temporal relationships between CHDs and fetal growth remains elusive, but 2 competing hypotheses seem plausible. Some risk factors for small for gestational age may also be associated with CHDs. It remains unclear whether small for gestational age and CHDs coexist because of common underlying etiologic pathways, such as maternal smoking, or because CHDs independently affect the growth of the fetus.
The critical period for CHD development is between the third and ninth weeks of gestation. Snijders et al40 showed that erythropoietin levels were increased significantly in small for gestational age infants secondary to fetal hypoxia. Although it is possible that altered fetal circulation because of CHDs might result in low birth weight, no consistent hemodynamic pattern with respect to oxygen saturations or hemoperfusion can explain our findings.
Clinical knowledge about the association that we found between CHDs and small for gestational age may have significant implications in both prenatal and postnatal care. Closer monitoring of fetal growth may be warranted after a prenatal diagnosis of CHD. Nutritional programs in pregnant women at risk of small for gestational age have been shown to improve infant outcomes,41 and mothers diagnosed with CHDs by fetal echocardiography may benefit from nutritional interventions. Postnatal clinical management of the infant with CHD and fetal growth restriction is dependant on cardiac pathology. The most common subtype of CHD is septal heart defects, which often require anticongestive medications if there are signs of congestive heart failure. These children are typically more prone to be placed on medications if the child has concomitant growth failure. They are also more likely to be referred for surgical repair of their underlying lesion if they continue to have growth retardation in the face of adequate medical and nutritional therapy. In contrast, infants with cyanotic heart lesions require multiple palliative surgeries in infancy. If this high-risk group of infants is nutritionally deficient secondary to fetal growth restriction, this can lead to prolonged postoperative hospital stays, as well as invasive surgeries, such as gastrostomy tube placement for nutritional rehabilitation before hospital discharge.
Our findings were based on participants in the NBDPS, which is the largest population-based case-control sample of all major cardiovascular malformations ever conducted in the United States.30 This study uses a structured maternal questionnaire to provide detailed information to confirm gestational age and exposure to environmental and behavioral factors. Uniform criteria for clinical confirmation of the cardiac defect and a rigorous review of abstracted medical chart data by an expert panel of clinicians also maximizes homogeneity between cases.
Limitations of our study must be considered. Sample sizes when stratified by prematurity were limited, reducing our ability to detect statistically significant differences in small for gestational age among specific cardiac phenotypes. Echocardiograms were not performed on control subjects, and thus some may have undiagnosed CHDs. We were not able to obtain length and head circumference data on the majority of NBDPS centers; this information would be relevant in future studies, because it would provide "symmetry" data regarding the timing of intrauterine growth retardation. Symmetrical growth retardation is associated with in utero injury early in pregnancy.
Further study of the association between CHD and small for gestational age will hopefully provide insights into basic mechanisms that may lead to both outcomes. Identifying risk factors for both CHDs and small for gestational age will contribute to the evidence base required for the planning and implementation of primary prevention programs to prevent heart defects and optimize pregnancy outcomes. A greater clinical understanding of the association between CHDs and small for gestational age may have significant implications for prenatal care. Closer monitoring of fetal growth and maternal nutrition is warranted after prenatal diagnosis of CHD. These women may be candidates for nutritional monitoring and more frequent and detailed ultrasounds to determine serial fetal weights.42,43 In addition, infants with CHDs who are small for gestational age require substantial nutritional intervention postnatally to curtail growth-related problems and improve clinical outcomes in infancy and early childhood.4447
| ACKNOWLEDGMENTS |
|---|
We thank all staff members of the Centers for Birth Defects Research and Prevention and all of the centers abstractors. Most of all, we thank all of the families of children who have participated in the National Birth Defects Prevention Study.
| FOOTNOTES |
|---|
Address correspondence to Charlotte A. Hobbs, MD, PhD, University of Arkansas for Medical Sciences, College of Medicine, Department of Pediatrics, 1120 Marshall St, Mail Slot 512-40, Little Rock, AR 72201. E-mail: hobbscharlotte{at}uams.edu
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention.
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. R. Menheniott, K. Woodfine, R. Schulz, A. J. Wood, D. Monk, A. S. Giraud, H. S. Baldwin, G. E. Moore, and R. J. Oakey Genomic Imprinting of Dopa decarboxylase in Heart and Reciprocal Allelic Expression with Neighboring Grb10 Mol. Cell. Biol., January 1, 2008; 28(1): 386 - 396. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||