Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Perinatol. Author manuscript; available in PMC 2009 July 7.
Published in final edited form as:
PMCID: PMC2706423

Maternal and fetal variation in genes of cholesterol metabolism is associated with preterm delivery



To examine the contribution of variants in fetal and maternal cholesterol metabolism genes in preterm delivery (PTD).

Study Design

A total of 40 single-nucleotide polymorphisms (SNPs) in 16 genes related to cholesterol metabolism were examined for 414 preterm infants (gestational ages 22 to 36 weeks; comprising 305 singletons and 109 twins) and at least 1 parent. Fetal effects were assessed using the transmission disequilibrium test (TDT) for each SNP, followed by a log linear model-based approach to utilize families with missing parental genotypes for those SNPs showing significance under TDT. Genetic variant effects were examined for a role in PTD, gestational age and birth weight. Maternal effects were estimated using a log linear model-based approach.


Among singleton gestations, suggestive association (P<0.01 without adjusting for multiple comparisons) was found between birth weight and fetal DHCR7 gene/SNP combinations (rs1630498, P=0.002 and rs2002064, P=0.003). Among all gestations, suggestive associations were found between PTD and fetal HMGCR (rs2303152, P=0.002) and APOA1 (rs 5070, P=0.004). The result for HMGCR was further supported by the log linear model-based test in the single births (P=0.007) and in all births (P=0.006). New associations (APOE and ABCA1) were observed when birth weight was normalized for gestational age suggesting independent effects of variants on birth weight separate from effects on PTD. Testing for maternally mediated genetic effects has identified suggestive association between ABCA1 (rs4149313, P=0.004) and decreased gestational age.


Variants in maternal and fetal genes for cholesterol metabolism were associated with PTD and decreased birth weight or gestational age in this study. Genetic markers may serve as one mechanism to identify high-risk mothers and fetuses for targeted nutritional treatment and/or prevention of low birth weight or PTD.

Keywords: prematurity, cholesterol, preterm delivery, birth weight


An estimated 12% of all births in the United States are preterm (less than 37 completed weeks gestation), with an increase of over 20% in the past decade.1 Morbidities associated with preterm delivery (PTD) contribute to high costs of hospitalization and lead to long-term disability.2,3 The cause of PTD can be identified in some cases, however the etiology in the majority of spontaneous PTD remains unknown.4,5 Previous research has concentrated on the identification of medical, social and environmental variables;5 however genetic predisposition toward PTD has been shown to be one of the strongest risk factors.6 Women are more likely to deliver a preterm infant if they have had a prior PTD, if they themselves were born prematurely or if they have a sister who delivered a preterm child.6,7 PTD is a common, complex and heterogeneous disease; therefore, it is unlikely that any one gene contributes to the majority of PTDs. Rather, the etiology may involve multiple genes acting independently or jointly, with interactions between genes and environment.8-10 Additionally, whether the mother or the fetus (or both) confers genetic risks of PTD is unknown. Fetal genotype and its effect on the initiation of labor may be as or more important than the contribution of the maternal genotype,9 and interactions between the maternal and fetal genetic backgrounds could further complicate etiologic analysis.

Complex traits such as PTD require both genetic and environmental investigations. Advances in technology and sample collection have afforded the opportunity to carry out powerful gene-only tests11 but ultimately these will need to be correlated with environmental covariates as well. In exploratory studies, identification of gene effects can offer gene pathways and specific environmental variables to be examined in more comprehensive and better-powered secondary studies.

Five pathways have been suggested to influence predisposition to PTD: ‘(1) intrauterine infection/inflammation; (2) maternal-fetal hypothalamic-pituitary adrenal axis activation; (3) uteroplacental vascular pathology; (4) pathologic uterine contraction; (5) susceptibility to environmental toxins’.12 While previous genetic studies have focused on genes related to infection/inflammation, coagulation, structural proteins and various signaling molecules (vascular endothelial growth factor (VEGF), β-2 adrenergic receptor), genes regulating nutritional status have not been as extensively studied.

Maternal nutritional states and plasma cholesterol levels are thought to contribute to risk of placental insufficiency, intrauterine growth retardation (IUGR), low birth weight and PTD.13-16 We examined both the maternal and fetal effects of common allelic variants in various cholesterol metabolism genes to examine their effect on the outcome of PTD. Single-nucleotide polymorphisms (SNPs) within seven lipid-related genes have been shown to account for most of the interindividual genetic variation in cholesterol levels.17 Candidate gene selection in this study was based on the polymorphisms identified by Knoblauch et al.17 and other genes involved in cholesterol synthesis and metabolism pathways.


DNA samples

This study selected preterm neonates (delivery before 37 completed weeks of gestation) admitted to the Neonatal Intensive Care Unit of University of Iowa Children’s Hospital. Babies were either born locally or transferred from other units within the first 28 days of life. Gestational age was determined by the first day of the last menstrual period; obstetrical judgment was made as to whether the last menstrual period date was certain. If uncertain, ultrasound measurements were made to determine gestational age. Samples were collected into a repository designed to supply DNA and limited epidemiologic data to examine complications of newborn infants. All families provided signed informed consent (IRB199911068). A total of 414 preterm babies (gestational ages 22 to 36 weeks; mean 31.25±3.7 weeks) along with one or both parents were enrolled in this study (253 fathers, 351 mothers). Only singleton births and twin gestation pregnancies were included (305 single births, 109 twins); the indication for the PTD was not available for most cases as the study uses a neonatal registry lacking this information. DNA was extracted from cord blood for the infants and from venous blood or buccal swabs for the parents. Demographic characteristics of the sample population are shown in Table 1. Maternal ethnicity/race was self-identified.

Table 1
Demographic characteristics of the study population


Genotyping was performed for all SNP markers using the Applied Biosystems (Foster City, CA, USA) TaqMan chemistry. Pre-designed or validated SNP genotyping assays were ordered from Applied Biosystems. These assays utilize primers to amplify the region containing the SNP of interest and two TaqMan Minor Groove Binder probes; each specific to one of the variant alleles at the site labeled with distinct reporter dyes, FAM and VIC. The list of genes and SNPS are in Table 2. SNPs were selected for known functional impact, high heterozygosity to enhance power and as tagging SNPs in haplotype blocks.

Table 2
List of genes and SNPs

Standard conditions supplied by Applied Biosystems were used for all reactions. Following thermocycling, fluorescence levels of the FAM and VIC dyes were measured and genotypes scored by inspection and entered in a laboratory database (Progeny) for generation of datasets for analysis. Mendelian inheritance inconsistencies were evaluated by retyping. A total of 97% of samples provided useable genotypes.

Statistical analysis

There are a variety of approaches to examine genetic effects on either the fetus or mother as the risk case. In this study, we took advantage of the samples available (fetus and parents) to use the transmission disequilibrium test (TDT) approach18 to search for fetal effects. This approach is a case-only approach and relies on transmission distortion of parental alleles to measure effect and does not require controls (the nontransmitted alleles from the parents serving, in effect, as the controls). To measure maternal effects, we used the log-linear methods where spousal genotypes served as controls for allele effects.

Testing effect of the fetus

Analysis was carried out for singleton gestations separately, then for singleton and twin gestations combined (both twins were included in the combined analyses). Three separate outcome variables were used for analysis: PTD as a dichotomous trait, and gestational age and birth weight as continuous outcomes. TDTs18 were utilized to detect allelic associations revealed by transmission of alleles from the parents to the fetus. Alleles at each marker were tested for association with the three outcome variables using the family-based association test program.19-21 When potentially significant results were observed (P<0.10), markers were further analyzed using a log linear model-based approach to utilize families with a missing parental genotype.22 Haplotype analysis was also completed for genes that had more than one SNP genotyped. In this study genetics effects only were examined and no controlling for confounding variables was possible, as data were incomplete within the original repository of samples.

Multiple testing

In this study, multiple tests were conducted with each of the 40 SNPs (two ways of grouping samples: single pregnancies only and single pregnancy with twin pregnancy; and three phenotypic outcomes: PTD, gestational age and birth weight). Given the multiple testing, for statistical significance at an α level of 0.05, the most conservative Bonferroni correction could be applied; (that is P-values of 0.0002 or less would be considered significant evidence of association under the most conservative model). The use of this correction introduces the possibility of Type-II error. Since this study is both hypothesis testing and hypothesis generating (given the study sample size and other limitations), less stringent values are also of interest.

Testing effect of the mother

Analyses of maternal effects were based on singleton gestations. We used a log-linear approach23 to study the maternally mediated effects on PTD. This approach uses ‘triads’, consisting of an affected offspring and the two parents, as the analysis unit and it tests maternally mediated genetic effects based on the symmetry assumption of allele counts between the mothers and the fathers in the source population, as defined by Schaid.24 This approach places no assumption of the underlying genetic model, that is, it allows for different relative risks corresponding to mother carrying one and carrying two copies of a susceptibility allele relative to no copies. We carried out likelihood ratio tests (LRT) of the maternal genetic effects and obtained the maximum likelihood estimators of the genetic relative risk. The expectation maximization algorithm25 was applied to fully utilize families with missing parental genotypes.

We applied a quantitative polytomous logistic regression approach26 to test for linkage and association between maternal genotypes and the gestational age. This approach allows for analysis of families with missing genotypes and does not require a normality assumption of the quantitative trait. Simulation study has shown robustness and powerfulness of this approach in detecting maternal-mediated effects both in the presence or absence of offspring effects.26

Adjustments for gestational age and birth weight

To adjust for the strong association of birth weight with gestational age birth weight was reanalyzed, in two ways: centered (gestational age- and gender-specific means subtracted) and normalized (gestational age- and gender-specific means subtracted and divided by the s.d.s.).


Fetal effect

The results are shown in Table 3 for all SNPs with P<0.10. The full dataset is available in Supplementary material online at and summarized in Figure 1.

Figure 1
Single pregnancies. Key to significant SNPs. 1, HMGCoARrs2303152; 2, DHCR7rs1630498; 3, LIPCrs1973028; 4, LIPCrs6088.
Table 3
Log linear model-based approach for fetal genetic effects

When singleton gestations were considered alone, association analysis yielded suggestive association (P<0.01 without correcting for multiple comparisons) for the DHCR7 gene/SNP combination when decreased birth weight was the outcome of interest (rs1630498, P=0.002 and rs2002064, P=0.003) (Figure 1). When single gestations were considered, and PTD was the primary outcome examined, SNP rs2303152 for the HMGCR gene had a P-value of 0.015. Analysis using a log linear model-based approach to allow inclusion of incomplete triads for this SNP yielded a even lower P-value of 0.007 (Table 3). Haplotype analyses for these genes with suggestive single SNP association yielded weaker association: HMGCR P=0.08 for PTD and DHCR7 (7-dehydrocholesterol reductase) P=0.09 for gestational age and P=0.03 for birth weight; in addition, hepatic triglyceride lipase (LIPC) had P=0.04 for association with PTD.

When all pregnancies were considered, the only suggestive associations were between PTD and the APOA1 gene/SNP combination (rs5070, P=0.004) and HMGCR (rs2303152, P=0.002) (Figure 2). Haplotype analyses for these genes and all other cholesterol metabolism genes typed for two or more SNPs showed no association. When we used centered or normalized birth weights (results were very similar for both) the results from Table 3 that were positive for prematurity, gestational age or birth weight (raw) became nonsignificant, while two additional SNPs (that were not significant with any of the other phenotypes) became significant. One was in apolipoprotein E (APOE, rs405509) with P=0.02 for both singleton and all pregnancies. The other was in ABCA1 (rs2066716) with P=0.05 for singleton and P=0.03 for all pregnancies.

Figure 2
Single and twin pregnancies. Key to significant SNPs: 1, HMGCoARrs2303152; 2, DHCR7rs2002064; 3, APOArs662799; 4, APOA1rs5070; 5, LIPCrs1968685.

Maternal effect

The results for maternal effect with P<0.05 are shown in Table 4 and summarized in Figure 3. We observed suggestive association (P<0.01) between gestational age and ABCA1 (rs4149313, P=0.004). Tests with individual P-values less than 0.05 were also observed for APOE (rs7412, P=0.022), LCAT (rs1109166, P=0.026) and LIPC (rs6083, P=0.032) when using gestational age as the outcome and for DHCR24 (rs2274941, P=0.025), when using PTD as the outcome (Table 4, Figure 3). We excluded analysis of HMGCAR (rs2303152) secondary to low-allele frequency.

Figure 3
Maternal genetic effects in singleton pregnancies.
Table 4
Maternal effects on prematurity and gestational age


Elucidating the genetic contribution to PTD may afford opportunities to improve the obstetrical management of women at high risk for delivering a child prematurely and identify pathophysiologic mechanisms that are new and/or amenable to interventions and prevention. Among these interventions may be the alteration of nutritional status based on the maternal and fetal genetic profiles.

Maternal nutrition, fetal growth and PTD

The relationship between maternal nutrition and PTD has not been thoroughly studied. Maternal and fetal nutrition are altered by fetal absorption, endocrine status and placental function. Inadequate nutrition may result in pathologic processes marked by slowed growth and IUGR.27,28 IUGR frequently complicates PTD and results in higher rates of neonatal morbidity and mortality. Approximately 30% of neonates born as a result of spontaneous PTD do not achieve the 10th percentile of their individual growth potential.29Additionally, many epidemiologic studies have linked low birth weight to increased risk of adult cardiovascular, endocrine and metabolic diseases, independent of gestational age.16,30

Maternal overnutrition has been shown to retard placental and fetal growth and increase fetal and neonatal mortality in rats, pigs and sheep.31 Human studies examining obesity as a risk for PTD are conflicting12,32,33 however, there is no question that the rates of obesity in the United States and other countries are rapidly increasing. If such an association exists, it will be important to determine the genetic and physiologic factors that contribute to PTD. Normally, increased levels of lipid and lipoprotein components become available to ensure a steady source of energy for the fetus during the third trimester. Atherosclerosis of placental spiral arteries or a physiologic failure to elevate maternal lipid levels may occur in overweight or obese individuals and contribute to placental insufficiency. Placental insufficiency may in turn result in PTD, IUGR and preeclampsia.34-36 Placental lipoprotein lipase (LPL) mRNA expression was shown by Tabano15 to be significantly higher when IUGR complicated the pregnancy.

Role of cholesterol metabolism in PTD

While different allelic variants have been associated with various abnormalities in lipid metabolism, interindividual variation has been demonstrated in all lipid-related genes.37 Knoblauch et al.17 demonstrated that SNPs and SNP-derived haplotypes explained most of the genetic variation in cholesterol levels. SNPs within seven genes: APOE, cholesteryl ester transfer protein (CETP), LIPC, apolipoprotein B (APOB), low-density lipoprotein cholesterol receptor (LDLR), LPL and ATP-binding cassette transporter (ABCA1) accounted for variation in LDL, high-density lipoprotein (HDL) and LDL/HDL ratios.17,38,39 Selection of candidate genes for this study was based on the polymorphisms suggested by Knoblauch and other genes involved in the pathways of cholesterol synthesis and metabolism. Current evidence also points to gene-gene and gene-environment interactions playing a role in plasma lipid concentrations, which were not examined in this study.37,39

Although preliminary, the results presented here suggest that cholesterol metabolism may play a role in PTD and/or birth weight. The HMGCR SNP rs2303152 in the fetus was associated with PTD. This SNP occurs in an intronic region, and thus may not have a functional role. This genetic risk occurs in the fetus, which has the notable implication that maternal cholesterol status may be less important than fetal metabolism of maternal cholesterol. Significance was weaker when the twins were included in the analysis, but twins may have independent mechanisms leading to PTD (uterine size, elaboration of inducing factors by two placentas/fetuses) that could be confounding these results. The methods to analyze which twin gestations in PTD remain under development. Issues include the age at which a twin gestation is considered preterm and whether one or both twins are included in the analysis.

SNPs for the DHCR7 gene were significantly associated with decreasing birth weight. The DHCR7 enzyme catalyzes the final step in the cholesterol synthesis pathway. When we normalized birth weight using gestational age appropriate norms these initial associations became less significant but new associations with APOE and ABCA1 appeared suggesting that some of the effects seen are secondary to the strong association of birth weight with gestational age but that birth weight-specific effects may be present as well. The ABCA1 gene was associated with both maternal and fetal effects and plays a critical role in lipid metabolism by regulating the transfer of cellular cholesterol.

We report our results using the Bonferroni correction as the most stringent comparison and with this none reach formal levels of significance. Nonetheless, since this study has limited power and multiple confounders, it is important to view it as hypothesis generating as well. In this setting, the findings of P-values on the order of 0.002 for DHCR7 and HMGCR and 0.02 for APOE and LCAT offer these gene/SNP combinations as ones to consider for replication and also offer gene pathways for consideration.

Limitations of association studies for complex diseases

Significant results found in association studies of complex diseases often fail in replication trials. Reasons for failure include population-specific effects, inadequate power in replication and Type-I errors in the initial finding. Large samples sizes are needed to appropriately evaluate the effect of a genetic variant implicated in a disease.40 This has proven to be the case in studies that have examined the genetic role in both lipid metabolism and PTD.9,37 It has been hypothesized that inadequate consideration of contributory gene-environment interactions may be the cause of these inaccurate observations. Maternal cholesterol status, while based on genetics is likely influenced by environmental components such as diet, tobacco and alcohol use, and physical activity.41 Environmental factors such as infection, cigarette smoking, stress and low socioeconomic status likely add to an individual’s baseline genetic risk of PTD.1,10 In this study, these variables were only inconsistently available, so we have focused initially on genetic factors only.

Future studies

In light of the limitations of the present study, the cholesterol genes should be examined in larger and different populations of individuals with PTD. Additionally, investigation of other genes and SNPs may be useful in further determining the role of cholesterol metabolism genes in PTD. The roles of maternal morbidities, tobacco use, nutrition, cholesterol status, prepregnancy weight and weight gain, and 17-hydroxyprogesterone use during pregnancy are among other critical variables to examine.

These results indicate that the genetic profile of either fetal or maternal cholesterol metabolism may be influential in determining risk of PTD. However, the power of our analysis of the maternal genetic effect is limited by our inability to examine transmission of alleles from the mother’s parents to the mother.42 To more accurately assess this maternal effect, it will be necessary to examine these relationships. Other weaknesses of this study include the use of a repository-based sample set without access to detailed epidemiologic data which limited our ability to differentiate different causes of prematurity and exclude cases arising from indicated vs spontaneous PTD.

It may also be inappropriate to consider preterm deliveries that resulted from cervical or placental causes together with those with spontaneous preterm labor. Future analyses will optimally subdivide different causes and outcomes of prematurity to better characterize the genetic contribution. In light of Tabano’s findings15 linking LPL gene expression and IUGR may be of value to consider babies with IUGR or ‘small for gestational age’ as an outcome category. Our current sample sizes are inadequate to support this subset analysis. Samples from large prospective cohort studies on pregnancy, such as the one in Denmark43 might also be used to provide sufficient data for such stratification. A final concern is that allelic or locus heterogeneity may be large enough to prevent detection of a common allele associated with disease. In these cases, linkage or medical resequencing will be needed to detect genetic effects as has been recently shown by Cohen.44

In the future, genetic markers related to nutrient and cholesterol metabolism may serve as one mechanism to identify high-risk mothers and fetuses. Recent association of low maternal serum cholesterol with preterm labor provides new support for this connection.45 Nutrient-based treatments could be used in these targeted populations to prevent low birth weight or PTD, making further investigation of nutritional pathways key in the investigation of genetic predisposition to PTD.


We thank, in particular, the families that participated in this study. The GCRC nurses at Iowa, Karen Johnson, Gretchen Cress, Nancy Krutzfield, Laura Knosp were essential in contacting, enrolling and sampling families. Kristin Orr and Cara Zimmerman were integral in sample organization. Sandy Daack-Hirsch assisted with consent protocols and Susie McConnell with manuscript preparation. The work was supported by NIH grant HD052953 and March of Dimes grant FY05-126.


1. Mattison DR, Damus K, Fiore E, Petrini J, Alter C. Preterm delivery: a public health perspective. Paediatr Perinat Epidemiol. 2001;15:7–16. [PubMed]
2. Petrou S. The economic consequences of preterm birth during the first 10 years of life. BJOG. 2005;112(s1):10–15. [PubMed]
3. Schmitt SK, Sneed L, Phibbs CS. Costs of newborn care in California: a population-based study. Pediatrics. 2006;117(1):154–160. [PubMed]
4. Ananth CV, Vintzileos AM. Epidemiology of preterm birth and its clinical subtypes. J Matern Fetal Neonatal Med. 2006;19(12):773–782. [PubMed]
5. Slattery MM, Morrison JJ. Preterm delivery. Lancet. 2002;360(9344):1489–1497. [PubMed]
6. Crider KS, Whitehead N, Buus RM. Genetic variation associated with preterm birth: a HuGE review. Genet Med. 2005;7(9):593–604. [PubMed]
7. Porter TF, Fraser AM, Hunter CY, Ward RH, Varner MW. The risk of preterm birth across generations. Obstet Gynecol. 1997;90(1):63–67. [PubMed]
8. Dizon-Townson DS. Preterm labour and delivery: a genetic predisposition. Paediatr Perinat Epidemiol. 2001;(Suppl 2):57–62. [PubMed]
9. Adams KM, Eschenbach DA. A genetic contribution towards preterm delivery. Semin Fetal Neonatal Med. 2004;9(6):445–452. [PubMed]
10. Cotten CM, Ginsburg GS, Goldberg RN, Speer MC. Geonomic analyses: a neonatology perspective. J Pediatr. 2006;148(6):720–726. [PubMed]
11. Christensen K, Murray JC. What genome-wide association studies can do for medicine. N Engl J Med. 2007;356(11):1094–1097. [PubMed]
12. Hao K, Wang X, Niu T, Xu X, Li A, Chang W, et al. A candidate gene association study on preterm delivery: application of high-throughput genotyping technology and advanced statistical methods. Hum Mol Genet. 2004;13(7):683–691. [PubMed]
13. Fall CHD, Yajnik CS, Rao S, Davies AA, Brown N, Farrant HJW. Micronutrients and fetal growth. J Nutr. 2003;133(5 Suppl 2):1747S–1756S. [PubMed]
14. Khoury J, Henriksen T, Christophersen B, Tonstad S. Effect of a cholesterol-lowering diet on maternal, cord, and neonatal lipids, and pregnancy outcome: a randomized clinical trial. Am J Obstet Gynecol. 2005;193(4):1292–1301. [PubMed]
15. Tabano S, Alvino G, Antonazzo P, Grati FR, Miozzo M, Cetin I. Placental LPL gene expression is increased in severe intrauterine growth-restricted pregnancies. Pediatr Res. 2006;59(2):250–253. [PubMed]
16. Wu G, Bazer FW, Cudd TA, Meininger CJ, Spencer TE. Maternal nutrition and fetal development. J Nutr. 2004;134(9):2169–2172. [PubMed]
17. Knoblauch H, Bauerfeind A, Toliat MR, Becker C, Luganskaja T, Günther UP, et al. Haplotypes and SNPs in 13 lipid-relevant genes explain most of the genetic variance in high-density lipoprotein and low-density lipoprotein cholesterol. Hum Mol Genet. 2004;13(10):993–1004. [PubMed]
18. Spielman RS, McGinnis RE, Ewens WJ. Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM) Am J Hum Genet. 1993;52:506–516. [PubMed]
19. Laird NM, Horvath S, Xu X. Implementing a unified approach to family-based tests of association. Genet Epidemiol. 2000;19(Suppl 1):S36–S42. [PubMed]
20. Rabinowitz D, Laird N. A unified approach to adjusting association tests for population admixture with arbitrary pedigree structure and arbitrary missing marker information. Hum Hered. 2000;50(4):211–223. [PubMed]
21. Horvath S, Xu X, Laird NM. The family based association test method: strategies for studying general genotype-phenotype associations. Eur J Hum Genet. 2001;9(4):301–306. [PubMed]
22. Weinberg CR. Allowing for missing parents in genetic studies of case-parent triads. Am J Hum Genet. 1999;64:1186–1193. [PubMed]
23. Weinberg CR, Wilcox AJ, Lie RT. A log-linear approach to case-parent-triad data: assessing effects of disease genes that act either directly or through maternal effects and that may be subject to parental imprinting. Am J Hum Genet. 1998;62(4):969–978. [PubMed]
24. Schaid DJ, Sommer SS. Genotype relative risks: methods for design and analysis of candidate-gene association studies. Am J Hum Genet. 1993;53(5):1114–1126. [PubMed]
25. Weinberg CR. Allowing for missing parents in genetic studies of case-parent triads. Am J Hum Genet. 1999;64(4):1186–1193. [PubMed]
26. Kistner EO, Infante-Rivard C, Weinberg CR. A method for using incomplete triads to test maternally mediated genetic effects and parent-of-origin effects in relation to a quantitative trait. Am J Epidemiol. 2006;163(3):255–261. [PubMed]
27. Luke B. The evidence linking maternal nutrition and prematurity. J Perinat Med. 2005;33:500–505. [PubMed]
28. Robinson JS, Moore VM, Owens JA, McMillen IC. Origins of fetal growth restriction. Eur J Obstet Gynecol Reprod Biol. 2000;92(1):13–19. [PubMed]
29. Bukowski R, Gahn D, Denning J, Saade G. Impairment of growth in fetuses destined to deliver preterm. Am J Obstet Gynecol. 2001;185:463–467. [PubMed]
30. Barker DJ, Eriksson JG, Forsen T, Osmond C. Fetal origins of adult disease: strength of effects and biological basis. Int J Epidemiol. 2002;31:1235–1239. [PubMed]
31. Wallace JM, Bourke DA, Aitken RP, Milne JS, Hay WW. Placental glucose transport in growth-restricted pregnancies induced by overnourished adolescent sheep. J Physiol. 2003;547:85–94. [PubMed]
32. Castro LC, Avina RL. Maternal obesity and pregnancy outcomes. Curr Opin Obstet Gynecol. 2002;14:601–606. [PubMed]
33. Stotland NE, Cheng YW, Hopkins LM, Caughey AB. Gestational weight gain and adverse neonatal outcome among term infants. Obstet Gynecol. 2006;108:635–643. [PubMed]
34. DeWolf F, Brosens I, Renaer M. Fetal growth retardation and the maternal arterial supply of the human placenta in the absence of sustained hypertension. BJOG. 1980;87:678–685. [PubMed]
35. Kim YM, Bujold E, Chaiworapongsa T, Gomex R, Yoon BH, Thaler HT, et al. Failure of physiologic transformation of the spiral arteries in patients with preterm labor and intact membranes. Am J Obstet Gynecol. 2003;189:1063–1069. [PubMed]
36. Robertsen WB, Brosens I, Dixon G. Maternal uterine vascular lesions in the hypertensive complications of pregnancy. Perspect Nephrol Hypertens. 1976;5:115–127. [PubMed]
37. Corella D, Ordovas JM. Single nucleotide polymorphisms that influence lipid metabolism: interaction with dietary factors. Annu Rev Nutr. 2005;25:341–390. [PubMed]
38. Benn M, Nordestgaard BG, Jensen JS, Grande P, Sillesen H, Tybjaerg-Hansen A. Polymorphism in APOB associated with increased low-density lipoprotein levels in both genders in the general population. J Clin Endocrinol Metab. 2005;90(10):5797–5803. [PubMed]
39. Sorli JV, Corella D, Frances F, Ramirez JB, Gonzales JI, Guillen M, et al. The effect of the APOE polymorphism on HDL-C concentrations depends on the cholesterol ester transfer protein gene variation in a Southern European population. Clin Chim Acta. 2006;366(12):196–203. [PubMed]
40. Hirschhorn JN. Genetic approaches to studying common diseases and complex traits. Pediatr Res. 2005;57:74R–77R. [PubMed]
41. Effects of Omega-3 Fatty Acids on Child and Maternal Health AHRQ Publication No 05-E025-2 August 2005.
42. Mitchell LE, Weinberg CR. Evaluation of offspring and maternal genetic effects on disease risk using a family-based approach: the ‘pent’ design. Am J Epidemiol. 2005;162:676–685. [PubMed]
43. Olsen J, Melbye M, Olsen SF, Sørensen TI, Aaby P, Andersen AM, et al. The Danish National Birth Cohort-its background, structure and aim. Scand J Public Health. 2001;29(4):300–307. [PubMed]
44. Cohen JC, Kiss RS, Pertsemlidis A, Marcel YL, McPherson R, Hobbs HH. Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science. 2004;305(5685):869–872. [PubMed]
45. Edison RJ, Berg K, Remaley A, Kelley R, Rotimi C, Stevenson RE, et al. Adverse birth outcomes among mothers with low serum cholesterol. Pediatrics. 2007 in press. [PubMed]