Transforming Growth Factor Beta-3 and Environmental Factors and Cleft Lip with/without Cleft Palate

Abstract

To identify the interactions among two loci (C641A and G15572-) of transforming growth factor beta 3 (TGFβ3), and exposures in pregnancy with cleft lip with/without cleft palate (CL/P), a hospital-based case–control study was conducted. Associations among offspring polymorphisms of TGFβ3 C641A and G15572-, paternal smoking, paternal high-risk drinking, maternal passive smoking, and maternal multivitamin supplement with CL/P were analyzed by logistic regression analysis, and the results showed that maternal passive smoking exposures and maternal multivitamin use were associated with the risk of CL/P but offspring polymorphisms of TGFβ3 C641A and G15572-, paternal smoking, and paternal high-risk drinking were not. Interactions among these variables were analyzed using the multifactor dimensionality reduction method, and the results showed that the two-factor model, including maternal passive smoking and TGFβ3 C641A, among all models evaluated had the best ability to predict CL/P risk with a maximum cross-validation consistency (9/10) and a maximum average testing accuracy (0.5892; p = 0.0010). These findings suggested that maternal passive smoking exposure is a risk factor for CL/P, whereas maternal multivitamin supplement is a protective factor. The polymorphism of TGFβ3 C641A participates in interaction effect for CL/P with environmental exposures, although the polymorphism was not associated with CL/P in single-locus analysis, and synergistic effect of TGFβ3 C641A and maternal passive smoking could provide a new tool for identifying high-risk individuals of CL/P and also an additional evidence that CL/P is determined by both genetic and environmental factors.

Introduction

N onsyndromic oral clefts (cleft lip only [MIM 119530], cleft lip with cleft palate [MIM 119530], and cleft palate only [MIM 119540]) together represent one of the most frequently observed congenital anomalies and represent a major public health burden in terms of both medical costs and emotional burden to patients and their families (Mitchell et al., 2002). Often cleft lip only and cleft lip with cleft palate are grouped together into cleft lip with/without cleft palate (CL/P), based on both embryologic and epidemiologic similarities (Christensen and Fogh-Andersen, 1994; Wilkie and Morriss-Kay, 2001). CL/P is a complex trait, and the pathogenetic mechanisms determined by genetic and environmental factors (Jugessur and Murray, 2005) remain unknown. Epidemiological studies have suggested that a proportion of CL/P may be associated with maternal exposures during the critical early period of pregnancy; these exposures include prescription drugs (Pradat et al., 2003; Dolk et al., 2008), cigarette smoking (Bille et al., 2007; Honein et al., 2007), alcohol (Bille et al., 2007; DeRoo et al., 2008), and deficiency of nutritional folic acid (van Rooij et al., 2004; Wilcox et al., 2007). In addition, many linkage and association studies of CL/P have been published, and some potential candidate loci and candidate genes have been identified. However, these studies often involve independent (i.e., main) effects of factors.

Transforming growth factor beta 3 (TGFβ3), located at 14q24 (Lidral and Moreno, 2005), has a number of important roles in embryonic development, including facial development, by regulating cell proliferation, differentiation, apoptosis, and chemotaxis (Massagué et al., 2000). In addition, some studies (Mitchell et al., 2001; Sato et al., 2001; Kim et al., 2003; Vieira et al., 2003) have demonstrated that TGFβ3 gene mutations are associated with CL/P in humans, although the others (Beaty et al., 2001; Li et al., 2006) show inconsistent results.

Two (641C/A and 15572G/-) polymorphisms in TGFβ3 were analyzed in this case–control study, and the interaction effects between these two loci and some common environmental exposures on the development of CL/P in a Han Chinese population were explored by the multifactor dimensionality reduction (MDR) method, which has been shown to be effective for detecting gene–environmental interactions in case–control studies with relatively small samples.

Materials and Methods

Study population

The subjects studied included 200 nonsyndromic CL/P patients who were aged between 6 months and 15 years (mean age: 4.9 ± 6.23 years), and 200 phenotypically normal individuals as the controls. All subjects were genetically unrelated ethnic Han Chinese. The selection criterion for patients was that they had to have CL/P with no other malformations. An individual with any other anomalies, such as neural tube defects, monogenic traits (e.g., Van der Woude's), syndromes (e.g., trisomy), and sequences (e.g., Pierre Robin), besides CL/P were excluded. All the cases included in the study were diagnosed through clinical examinations by well-trained dentists and oral surgeons and were evaluated by at least one dysmorphologist who confirmed their nonsyndromic status. Members of the control group were from the outpatients with trauma and the inpatients with bone fracture treated in the same hospital. Selection criteria for controls were no evidence of any other serious illness, especially hereditary diseases, and no family history of CL/P. The control group was matched to the case group by sex, age distribution, and socioeconomic status.

Interviews

After informed consent was obtained from each participant, mothers of cases and controls were interviewed in the hospital with a questionnaire. The information includes demographics (parents' age, schooling, socioeconomic status, and children's age and sex), family history of oral clefts, as well as parents' tobacco and alcohol consumption, mother's multivitamin supplement, and passive smoking.

For smoking, a woman was asked if she and her child's natural father had smoked during the 3 months before conception, and also if they had smoked in the first trimester of pregnancy. The average number of cigarettes smoked per day was calculated separately when smoking was reported. To assess passive smoke exposures, a woman was asked whether anyone smoked inside her home or near her at work during the same 6 months.

Reports of paternal and maternal alcohol consumption were collected for the 3-month period before conception and the first trimester of pregnancy. Participants were asked to enumerate the number of times they drank per month, the typical number of alcohol units they drank per occasion, and the kinds of drinks (beer, wine, or liquor) consumed when drinking was reported.

Maternal multivitamin supplementation was reported separately for the 3-month period before conception and the first trimester of pregnancy. Mothers who supplemented with vitamins were asked how long they used them in the period and how often in a week. At the end of the interview, an approximately 5 mL blood sample was collected from each subject.

DNA extraction and genotyping

Genomic DNA was extracted from EDTA-anticoagulated peripheral blood leukocytes by the salting-out method. Cases and controls were genotyped for the TGFβ3 markers (641C/A and 15572G/-) by a polymerase chain reaction (PCR)–restriction fragment length polymerase assay. The PCR reactions were performed in a total volume of 25 μL containing 100 ng genomic DNA, 20 pM of each primer, 0.2 mM dNTPs, 20 mM Tris–HCl (pH 8.8), 10 mM KCl, 10 mM (NH₄)₂SO₄, 2 mM MgSO₄, 0.1% Triton X-100, and 1 unit of Taq polymerase (New England BioLabs, Ipswich, MA). The sequences for the primers and conditions of amplification are listed in Table 1. The PCR products were incubated for 6 h with the corresponding digestion enzyme (Table 1). The digested products were then analyzed on 8% polyacrylamide gels and visualized by a silver staining protocol.

Table 1.

Primers and Conditions to Amplify Polymorphisms of Transforming Growth Factor Beta-3 Gene

Loci	Primer forward	Primer reverse	Product size (bp)	Annealing temperature (°C)	Digestion enzyme
641C/A	5′-CTGGTTTTCCTCCCTCCTTC-3′	5′-CTCCCAGCTCCAGTTCAGAC-3′	449	59	PshA
15572G/-	5′-CTTGAAGGTCCACTCTGAGC-3′	5′-GGGCAAAGCAGAATTACTCA-3′	499	58	BseD

Statistical analysis

Using the SPSS 11.5 statistical software package, the Pearson's χ ² test was used to test for differences between the cases and the control subjects for the categorical variables of sex, socioeconomic status, paternal schooling, and maternal schooling. The Wilcoxon rank sum test was used to test the difference of child's age, paternal age, and maternal age between groups. Unconditional logistic regression analysis was conducted to assess the independent main effects of the single-nucleotide polymorphisms C641A and 15572G/- of TGFβ3 and environmental exposures, including parents' tobacco and alcohol consumption, mother's multivitamin supplement, and passive smoking. Maternal smoking and maternal alcohol consumption were excluded from the analysis because none of mothers in this study reported smoking or drinking.

Potential interactions among TGFβ3 C641A and 15572G/- and environmental exposures were assessed using the open-source MDR software package (Hahn et al., 2003) available from www.epistasis.org/.

In all analyses, environmental exposures of interest were considered as dichotomous variables. The strategies used to dichotomize these variables are described in the following paragraphs.

Paternal smoking

It is hard to classify fathers who stopped smoking “when their wives became pregnant,” because their own recognition of pregnancy may or may not have preceded the crucial time of lip and/or palate formation. Therefore, we chose to define a smoker as a man who smoked at least one cigarette per day in either the 3 months before conception or the first trimester of pregnancy. Individuals who had never smoked or occasionally smoked were considered nonsmokers.

Paternal alcohol consumption

Similarly, paternal alcohol consumption was categorized as yes/no according to whether or not high-risk drinking in either the 3 months before conception or the first trimester of pregnancy occurred. High-risk drinking is defined as drinking 40 g/day or more of pure alcohol on average (WHO, 2004a). Level of drinking was estimated on the basis of the quantity, frequency of consumption, and alcoholic content of different kinds of drinks (wine, beer, liquor).

Maternal supplement of multivitamin

If the mother reported taking a multivitamin either the 3 months before conception or the first trimester of pregnancy, we classified her as a multivitamin user in our analyses.

Maternal passive smoking

A mother was considered exposed to passive smoking if she reported someone smoking inside her home or near her at work and exposure for at least 15 min/day for more than 1 day/week (WHO, 2004b).

Results

Characteristics of the study population

As reported in a previous paper (Jianyan et al., 2010), there were no significant differences between the cases and controls for child's age, child's sex, and socioeconomic status. Maternal age and paternal age were separately similar between the two groups. Educational level was lower for the cases' mothers or fathers compared with the controls.

The main effects of gene and environmental exposures

The effects of TGFβ3 genotypes 641C/A and 15572G/-, paternal smoking, paternal high-risk drinking, maternal multivitamin supplement, and maternal passive smoking were analyzed by unconditional logistic regression. Maternal schooling and paternal schooling were considered as potential confounding factors when testing for association between CL/P and exposures. Adjusted odds ratios and 95% confidence intervals were then computed. No significant association in relation to genotype or allele distributions for TGFβ3 641C/A and 15572G/- was found between patients and controls, and the results are shown in Table 2. As reported in a previous paper (Jianyan et al., 2010), frequencies of paternal smoking and paternal high-risk drinking in CL/P cases are not significantly different when compared with controls. Maternal multivitamin supplement was associated with a significantly decreased risk of CL/P, whereas maternal passive smoking was associated with an increased risk of CL/P.

Table 2.

The Main Effect of Transforming Growth Factor Beta-3 Gene

Variables	Number of cases	Number of controls	Odds ratios ^a	95% confidence intervals
C641A
Genotypes
CC	193	196	1.000
CA	7	4	0.563	0.162, 1.953
Alleles
C	393	396	1.000
A	7	4	0.567	0.165, 1.953
15572G/-
Genotypes
GG	172	176	1.000
G-	28	24	0.838	0.467, 1.503
Alleles
G	372	376	1.000
-	28	24	0.848	0.483, 1.490

Estimated odds ratios from unconditional logistic regression adjusted for maternal schooling and paternal schooling.

Potential interactions between gene and environmental exposures

MDR software was used to identify which combination of factors is the best model to predict the high-risk subgroup of subjects. Table 3 summarizes the results of MDR analysis for all number of factors evaluated (n = 1, 2, 3, 4, 5, 6). The best model of each order is shown along with its cross-validation consistency, average testing accuracy, and significance level as determined by permutation testing. The model including only maternal passive smoking was the best single-factor model for predicting CL/P risk, with a cross-validation consistency of 8/10 and an average testing accuracy of 0.5650. Among two-factor models, the combination of maternal passive smoking and TGFβ3 C641A proved to be the strongest, with a cross-validation consistency of 9/10 and an average testing accuracy of 0.5892. When three factors were considered, the combination of maternal passive smoking, maternal multivitamin supplement, and paternal high-risk drinking was the strongest model, with a cross-validation consistency of 7/10 and an average testing accuracy of 0.5700. A model containing maternal passive smoking, maternal multivitamin supplement, paternal smoking, and paternal high-risk drinking was the strongest among all four-factor models, with a cross-validation consistency of 9/10 and an average testing accuracy of 0.5625. When five factors were considered, the combination of maternal passive smoking, maternal multivitamin supplement, paternal smoking, paternal high-risk drinking, and TGFβ3 G15572- was the strongest model, with a cross-validation consistency of 8/10 and an average testing accuracy of 0.5517. When all six factors were considered, the cross-validation consistency and the average testing accuracy of the model were, respectively, 7/10 and 0.5418. Among all models, the two-factor model, including maternal passive smoking and TGFβ3 C641A, had a maximum cross-validation consistency and a maximum average testing accuracy (p = 0.0010). Therefore, we chose the two-factor model as the best model.

Table 3.

Summary of Results for Cleft Lip with/without Cleft Palate Risk Prediction from Multifactor Dimensionality Reduction Analysis

Number of factors	Best model	Cross-validation consistency	Average testing accuracy	Permutation test P-value
One factor	Maternal passive smoking	8/10	0.5650	0.0107
Two factors	Maternal passive smoking	9/10	0.5892	0.0010^a
	TGFβ3 C641A
Three factors	Maternal passive smoking	7/10	0.5700	0.0010
	Maternal multivitamin supplement
	Paternal high-risk drinking
Four factors	Maternal passive smoking	9/10	0.5625	0.0010
	Maternal multivitamin supplement
	Paternal smoking
	Paternal high-risk drinking
Five factors	Maternal passive smoking	8/10	0.5517	0.0107
	Maternal multivitamin supplement
	Paternal smoking
	Paternal high-risk drinking
	TGFβ3 G15572-
Six factors	Maternal passive smoking	7/10	0.5418	0.0107
	Maternal multivitamin supplement
	Paternal smoking
	Paternal high-risk drinking
	TGFβ3 C641A
	TGFβ3 G15572-

The overall best MDR model.

TGF, transforming growth factor; MDR, multifactor dimensionality reduction.

The distribution of CL/P and controls for the set comprising maternal passive smoking status and the TGFβ3 C641A polymorphism, is depicted in Figure 1. The threshold for high and low risk was 1 (200/200). Note that pattern of high-risk (dark shading) and low-risk (light shading) combinations is nonlinear across the four two-factor cells. This is evidence for interaction as suggested by the interaction information analysis.

FIG. 1.

The identified best model. In each cell, the left bar represents the distribution of cases, and the right bar the distribution of controls. Note that the ratio of the total number of CL/P to the total number of controls in the database did not exceed the threshold of 1; the boxes are labeled as low risk, or else labeled as high risk. High-risk cells are indicated by dark shading, and low-risk cells by light shading. Genotypes 0 and 1 denote CC and CA, respectively, for TGFβ3 C641A. Passive smoking 0 and 1 denote unexposure and exposure, respectively. CL/P, cleft lip with/without cleft palate; TGF, transforming growth factor.

Discussion

This study evaluated the interactions among genetic variation at TGFβ3 C641A and G15572-, and several common environmental exposures (parents' tobacco and alcohol consumption, mother's multivitamin supplement, and mother's passive smoking) during the 3 months before conception to the first trimester of pregnancy and CL/P.

No significant evidence of association was found between TGFβ3 C641A or G15572- and CL/P. This result is consistent with that reported by Beaty et al. (2001) and Li Zhi-ping et al. (Li et al., 2006), but conflicts with majority of previous reports (Mitchell et al., 2001; Sato et al., 2001; Kim et al., 2003; Vieira et al., 2003). This discrepancy likely reflects the complex etiology of the CL/P malformation. Indeed, the numerous genetic and environmental factors involved probably contribute differently to the development of the malformation in distinct populations. In addition, differences in experimental design may also account for some of the observed discrepancies.

Potential interactions among paternal smoking, paternal high-risk drinking, maternal passive smoking, maternal multivitamin supplement, and genotypes of TGFβ3 C641A and G15572- were explored by MDR analysis in our study. The two-factor model, including maternal passive smoking and TGFβ3 C641A, was the strongest model overall because it had the highest level of average testing accuracy and showed good cross-validation consistency. The MDR two-factor model indicated that maternal passive smoking and TGFβ3 C641A were a high-risk combination of factors but did not specify whether there was a synergistic relationship. Figure 1 interprets the nature of the interactions in these multifactor models. In Figure 1, we observed that although maternal passive smoking was a risk factor for CL/P, in some cases, it depends on the genotypes carried by the studied individuals. The subjects with the CA genotype of TGFβ3 could have higher CL/P risk, although their mother reported no passive smoking exposures. The interaction model provides additional evidence that CL/P is a complex disorder, and continued investigation of the combined effects of offspring genotypes and environmental teratogens is needed.

The use of the MDR method in this study is an advantage. Over the past decade, a few potential gene–environment interactions have been evaluated as risk factors for CL/P (Prescott et al., 2001; Shi et al., 2008). However, previously reported associations between gene–environment interactions with CL/P were evaluated using parametric statistical methods, such as logistic regression analysis, stratified analysis, and log-linear analysis. The power of parametric statistical methods is low in detecting interactions, and the parameters cannot be accurately estimated when there are many independent variables while the sample size is not large enough (Moore and Williams, 2002). With the nonparametric and genetic model-free MDR approach, multilocus genotypes/diplotypes and environmental factors can be evaluated together to fit a best prediction model and further be pooled into high- and low-risk groups, effectively reducing the dimensionality from N dimensions to one dimension (Ritchie et al., 2001). Thus, MDR has good power for identifying high-order gene–gene interactions or gene–environment interactions with relatively small samples (Ritchie et al., 2001). However, our study also had some notable limitations. First, we only genotyped two polymorphisms of TGFβ3 and did not cover the gene fully. Many association studies have been reported between single-nucleotide polymorphisms of TGFβ3 and CL/P. C641A was first discovered in Han Chinese populations in 2005 (Hu et al., 2005). Therefore, we selected these two loci for association analysis. Other factors, such as occupational exposures, domestic chemical exposures, drugs use, and certain dietary components, might interact with TGFβ3 or act as potential confounders. Unfortunately, information on these factors in our case–control study was not available. It would be interesting to investigate interactions between TGFβ3 genotypes and these risk factors in future studies. The sample size in this study was relatively small, although the power of MDR is over 80% with a case–control sample dataset with 200 cases and 200 controls (Hahn et al., 2003). This is the possible reason why data of maternal smoking and maternal drinking were not available in the study. It is thus necessary to carry out related studies in larger study designs.

In summary, interaction among TGFβ3 G15572-, maternal passive smoking, and maternal multivitamin supplement was detected by MDR, although studied polymorphisms were not associated with CL/P in single-locus analysis. The result may provide insights into better diagnosis and prevention of CL/P. Avoiding passive smoking exposures in pregnancy as well as maternal multivitamin supplementation is likely to decrease the risk of having a child with CL/P.

Footnotes

Acknowledgments

The authors thank Wu Rui, visiting doctor of the Affiliated Hospital of Guiyang Medical College, and Huang Yuanlu, chief physician of the Baise People's Hospital, for their help in collecting data, and Chen Guichun, Zheng Yueping, Wang Yue, and Huang Yanping of Guizhou Center for Disease Control and Prevention for their help in the experimentation.

Disclosure Statement

No competing financial interests exist.

References

Beaty

T.H.

, Wang

, Hetmanski

J.B.

, Fan

Y.T.

, Zeiger

J.S.

, Liang

K.Y.

, Chiu

Y.F.

, Vanderkolk

C.A.

, Seifert

K.C.

, Wulfsberg

E.A.

, Raymond

, Panny

S.R.

, Mclntosh

2001. A case-control study of nonsyndromic oral clefts in Maryland. Ann Epidemiol, 11:434–442.

Bille

, Olsen

, Vach

, Knudsen

V.K.

, Olsen

S.F.

, Rasmussen

, Murray

J.C.

, Andersen

A.M.

, Christensen

2007. Oral clefts and life style factors—a case-cohort study based on prospective Danish data. Eur J Epidemiol, 22:173–181.

Christensen

, Fogh-Andersen

1994. Etiological subgroups in non-syndromic isolated cleft palate. A genetic-epidemiological study of 52 Danish birth cohorts. Clin Genet, 46:329–335.

DeRoo

L.A.

, Wilcox

A.J.

, Drevon

C.A.

, Lie

R.T.

2008. First-trimester maternal alcohol consumption and the risk of infant oral clefts in Norway: a population-based case-control study. Am J Epidemiol, 168:638–646.

Dolk

, Jentink

, Loane

, Morris

, de Jong-van den Berg

L.T.

2008. Does lamotrigine use in pregnancy increase orofacial cleft risk relative to other malformations? Neurology, 71:714–722.

Hahn

L.W.

, Ritchie

M.D.

, Moore

J.H.

2003. Multifactor dimensionality reduction software for detecting gene-gene and gene-environment interactions. Bioinformatics, 19:376–382.

Honein

M.A.

, Rasmussen

S.A.

, Reefhuis

, Romitti

P.A.

, Lammer

E.J.

, Sun

, Correa

2007. Maternal smoking and environmental tobacco smoke exposure and the risk of orofacial clefts. Epidemiology, 18:226–233.

B.-C.

, C

S.-L.

, Wang

J.-G.

, Wang

G.-L.

, Gao

P.-J.

, Zhu

D.-L.

2005. No association between 3 polymorphisms of transforming growth factor gene and essential hypertension in Chinese. Chin J Cardiol, 33:127–131.

Jianyan

, Zeqiang

, Yongjuan

, Kaihong

, Bing

, Rongsheng

2010. Analysis of interactions between genetic variants of BMP4 and environmental factors with nonsyndromic cleft lip with or without cleft palate susceptibility. Int J Oral Maxillofac Surg, 39:50–56.

10.

Jugessur

, Murray

J.C.

2005. Orofacial clefting: recent insight into a complex trait. Curr Opin Genet Dev, 15:270–278.

11.

Kim

M.H.

, Kim

H.J.

, Choi

J.Y.

, Nahm

D.S.

2003. Transforming growth factor-beta3 gene SfaN1 polymorphism in Korean nonsyndromic cleft lip and palate patients. J Biochem Mol Biol, 36:533–537.

12.

Z.-P.

, Sun

H.-C.

, Ou

Y.-J.

2006. Association of TGFα and TGFβ3 polymorphisms with nonsyndromic cleft lip and left palate. J Pract Stomatol, 22:667–671.

13.

Lidral

A.C.

, Moreno

L.M.

2005. Progress toward discerning the genetics of cleft lip. Curr Opin Pediatr, 17:731–739.

14.

Massagué

, Blain

S.W.

, Roger

S.L.

2000. TGF-β signaling in growth control, cancer, and heritable disorders. Cell, 103:295–309.

15.

Mitchell

L.E.

, Beaty

T.H.

, Lidral

A.C.

, Munger

R.G.

, Murray

J.C.

, Saal

H.M.

, Wyszynski

D.F.

2002. Guidelines for the design and analysis of studies on nonsyndromic cleft lip and cleft palate in humans: summary report from a workshop of the international consortium for oral clefts genetics. Cleft Palate Craniofac J, 39:93–100.

16.

Mitchell

L.E.

, Murray

J.C.

, O'Brien

, Christensen

2001. Evaluation of two putative susceptibility loci for oral clefts in the Danish population. Am J Epidemiol, 153:1007–1015.

17.

Moore

J.H.

, Williams

S.M.

2002. New strategies for identifying gene-gene interactions in hypertension. Ann Med, 34:88–95.

18.

Pradat

, Robert-Gnansia

, Di Tanna

G.L.

, Rosano

, Lisi

, Mastroiacovo

2003. First trimester exposure to corticosteroids and oral clefts. Birth Defects Res A Clin Mol Teratol, 67:968–970.

19.

Prescott

N.J.

, Winter

R.M.

, Malcolm

2001. Nonsyndromic cleft lip and palate: complex genetics and environmental effects. Ann Hum Genet, 65:505–515.

20.

Ritchie

M.D.

, Hahn

L.W.

, Roodi

, Bailey

L.R.

, Dupont

W.D.

, Parl

F.F.

, Moore

J.H.

2001. Multifactor-dimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer. Am J Hum Genet, 69:138–147.

21.

Sato

, Natsume

, Machido

, Suzuki

, Kawai

2001. Association between transforming growth factor beta 3 and cleft lip and/or palate in the Japanese population. Plast Reconstr Surg, 107:1909–1910.

22.

Shi

, Wehby

G.L.

, Murray

J.C.

2008. Review on genetic variants and maternal smoking in the etiology of oral clefts and other birth defects. Birth Defects Res C Embryo Today, 84:16–29.

23.

van Rooij

I.A.

, Ocké

M.C.

, Straatman

, Zielhuis

G.A.

, Merkus

H.M.

, Steegers-Theunissen

R.P.

2004. Periconceptional folate intake by supplement and food reduces the risk of nonsyndromic cleft lip with or without cleft palate. Prev Med, 39:689–694.

24.

Vieira

A.R.

, Orioli

I.M.

, Castilla

E.E.

, Cooper

M.E.

, Marazita

M.L.

, Murray

J.C.

2003. MSX1 and TGFB3 contribute to clefting in South America. J Dent Res, 82:289–292.

25.

Wilcox

A.J.

, Lie

R.T.

, Solvoll

, Taylor

, McConnaughey

D.R.

, Abyholm

, Vindenes

, Vollset

S.E.

, Drevon

C.A.

2007. Folic acid supplements and risk of facial clefts: national population based case-control study. BMJ, 334:464–469.

26.

Wilkie

A.O.M.

, Morriss-Kay

G.M.

2001. Genetics of craniofacial development and malformation. Natl Rev Genet, 2:458–468.

27.

World Health Organization. 2004a. Global Status Report on Alcohol 2004. Geneva.

28.

World Health Organization. 2004b. Women and the Tobacco Epidemic Challenges for the 21st Century 2001. Geneva.