Hypospadias Risk and Polymorphism in SRD5A2 and CYP17 Genes: Case-Control Study Among Indian Children

Hypospadias Risk and Polymorphism in SRD5A2 and CYP17Genes: Case-Control Study Among Indian Children

Ratika Samtani, Minu Bajpai, Kapil Vashisht, P. K. Ghosh and K. N. Saraswathy*From the Department of Anthropology, University of Delhi (RS, PKG, KNS) and Department of Pediatric Surgery, All India Institute of MedicalSciences, Ansari Nagar (MB, KV), Delhi, India

Abbreviations

and Acronyms

DHT � dihydrotestosterone

Submitted for publication October 11, 2010.* Correspondence: Department of Anthropol-

ogy, University of Delhi, Delhi 110007, India (tele-phone: 91-11-64586216; FAX: 91-11-27666614;e-mail: [email protected]).

Purpose: Hypospadias is a common congenital error of genital development, thefrequency of which is increasing. As androgens have a significant role in thedevelopment of the male urethra, we sought to investigate the association be-tween 2 functional polymorphisms, CYP17-A1/A2 and SRD5A2-V89L, which areinvolved in the biosynthesis of testosterone and dihydrotestosterone, respec-tively, in relation to hypospadias.Materials and Methods: We examined DNA samples of 80 cases and 100 controlsfor SRD5A2-V89L and CYP17-A1/A2 gene polymorphisms. Information pertain-ing to family history, preoperative position of the urethral meatus and parentaloccupations along with maternal reproductive profile were collected for cases andcontrols.Results: Genotyping of 80 cases and 100 controls revealed a significant associa-tion between V89L polymorphism and hypospadias (OR 2.4, 95% CI 1.2–4.6,p �0.05). When analyzing the risk of hypospadias based on grade, genotypicdistribution of SRD5A2-LL genotype differed significantly between severe formsand controls, with an odds ratio of 3.6 (95% CI 1.2–10.0, p � 0.02). Of affectedchildren 71.25% had parents from a rural background, with agriculture as theprimary occupation. A statistically significant association was observed for theLL genotype (OR 4.6, 95% CI 1.7–12.29, p �0.05) between children with parentshaving an agricultural background (likely exposed to pesticides) and controlswith no such exposure. CYP17-A1/A2 genotypes did not show any significantresults.Conclusions: V89L polymorphism of the SRD5A2 gene is a strong determinant ofhypospadias risk among children of Indian origin. However, our results suggestthat the presence of leucine allele, especially among agriculturalists, may in-crease the propensity of having a child with hypospadias.

Key Words: agriculture; hypospadias; polymorphism, genetic;

2334 www.jurology.com

HYPOSPADIAS is a common congenitalerror of genital development. The in-cidence of hypospadias is 1 per 125 to300 live male births and has report-edly increased during the last few de-cades.1,2 A greater frequency of hy-pospadias is uniformly seen in whitescompared to other races.3,4 This con-

0022-5347/11/1856-2334/0THE JOURNAL OF UROLOGY®

© 2011 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RES

which the urethral meatus is locatedon the ventral side of the penis prox-imal to the tip of the glans, from thebalanopreputial sulcus to the perinealarea. Patients born with severe hy-pospadias and penile curvature mayhave ambiguous genitalia in the new-born period, making an immediate ac-

Vol. 185, 2334-2339, June 2011Printed in U.S.A.

EARCH, INC. DOI:10.1016/j.juro.2011.02.043

ROLE OF SRD5A2 AND CYP17 GENE POLYMORPHISMS IN HYPOSPADIAS 2335

Hypospadias may be classified as simple (glanu-lar or penile) or severe (scrotal or perineal) based onthe anatomical location of the urethral meatus. It isregarded as a complex disorder caused by geneticand environmental influences, either alone or incombination.6 Formation of the male urethra duringthe first trimester of gestation is fully androgendependent and, therefore, it appears more reason-able to explain hypospadias as an abnormality of theandrogen metabolic pathway.7 Regarding this met-abolic pathway, formation of testosterone and itsconversion to the more potent DHT are crucial ininducing the development of the internal/externalgenitalia, urethra and prostate. This mechanismalso binds to the androgen receptor, forming a com-plex (androgen receptor-DHT) that translocates tothe nucleus and transactivates target genes. Thus,genetic polymorphisms in genes controlling andro-gen action and biosynthesis of testosterone and di-hydrotestosterone are likely to be important in theetiology of hypospadias.

The biosynthetic conversion of cholesterol to tes-tosterone involves several discrete steps, in whichCYP17 gene encodes an enzyme with 17�-hydroxy-lase and 17,20-lyase activities, which is a rate limit-ing step in testosterone biosynthesis.8 17�-Hydroxylasehas a major role in hydroxylating pregnenolone andprogesterone, which are converted to C19 steroidprecursors of testosterone and estrogen by 17,20-lyase activity. A transition from T to C has beendescribed in the 59 untranslated regions, which cre-ates an additional Sp-1 type (CCACC box) motif andan MspA1 I restriction enzyme site. Although theeffect of this base change on gene expression is un-known, the variant C allele (designated A2) mayresult in increased transcriptional activity and thusincreased biosynthesis of testosterone.9 The A2 al-lele has been extensively studied in relation to prostatecancer risk in men,10,11 and polycystic ovarian syn-drome and breast cancer risk in women.12,13 How-ever, the role of A2 allele in the studied anomalies isdebatable. To our knowledge there have been nostudies to date examining the possible effect of thispolymorphism on isolated hypospadias.

The type II steroid 5�-reductase enzyme encodedby SRD5A2 gene on chromosome 2 irreversibly me-tabolizes the conversion from testosterone to DHT.14

A major functional polymorphism of the SRD5A2gene, V89L, is caused by a G to C transversion thatresults in the substitution of valine for leucine atcodon 89. According to Makridakis et al, the distri-bution of valine allele of V89L polymorphism, knownto increase enzymatic activity (increased DHT lev-els), parallels prostate cancer risk between differentethnic groups.15 However, a lower prostate cancerrisk observed among Chinese and Japanese men due

DHT levels) is attributed to the protective effect ofthe high frequency of the leucine allele found amongthese groups.15 A direct association of V89L poly-morphism with isolated hypospadias has been re-ported previously.6,16 However, according to Thai etal, this association needs to be considered with cau-tion due to the use of unmatched population basedcases and controls in their study.6

Few studies have been conducted to date toincrease our understanding of the various causesof hypospadias. To our knowledge this is the firststudy to elucidate hypospadias risk among chil-dren of Indian origin (North Indian, white popu-lation), where the anomaly seems to be increasing.We sought to investigate the association betweenthe 2 genetic polymorphisms, SRD5A2-V89L andCYP17-A1/A2, and isolated cases of hypospadiasamong children of Indian origin via a case-controlstudy. In particular we tested hypotheses on theroles of CYP17-A2 allele (known to increase tes-tosterone levels) and SRD5A2-V89L allele (knownto decrease DHT levels) in the development ofhypospadias.

PATIENTS AND METHODS

A total of 80 boys with hypospadias 4 to 10 years old wereprospectively entered in the study. All cases were diag-nosed at the Department of Pediatric Surgery, All IndiaInstitute of Medical Sciences and scheduled to undergorepair between February 2008 and October 2010. All pa-tients were 46,XY. Preoperative position of the urethralmeatus was noted and phenotype was graded as mild,medium or severe. Patients with undescended testis, in-tersex conditions or known endocrine abnormalities wereexcluded from the study.

A total of 100 age, geography (North India), ethnicity(white) and language (Indo-European) matched controlsfrom the general North Indian population without hypos-padias or any history of genital abnormalities were col-lected via house-to-house surveys. Demographic and clin-ical data pertaining to reproductive profile of the mother,occupation of the parents, and family history of genitalabnormalities in male blood relatives were recorded forcases and controls. Intravenous blood samples of 3 mlwere collected in ethylenediaminetetraacetic acid coatedtubes after receiving informed written consent from theparents.

Genomic DNA was extracted from blood using DNAextraction kits (Qiagen, Valencia, California). DNA se-quencing was performed for the exon 5 region of theSRD5A2 gene for severe forms of hypospadias.17 For theSRD5A2 gene a modification of the assay of Vilchis et alwas used to genotype study subjects for the V89L poly-morphism.18 Instead of metaphor agarose, visualization ofthe fragments was accomplished in a 12% polyacrylamidegel after staining with ethidium bromide. Polymorphismsin CYP17 gene were detected using polymerase chainreaction followed by restriction enzyme digestion, as de-

ROLE OF SRD5A2 AND CYP17 GENE POLYMORPHISMS IN HYPOSPADIAS2336

Odds ratio as a measure of the relative risk of hypos-padias and 95% CI were calculated by standard method.19

The prevalence of polymorphism was compared betweenpatients and controls using chi-square testing, with p �0.05considered statistically significant. Multivariate analysiswas performed using SPSS®, version 15.0.

RESULTS

Demographic data for the 2 groups are outlined intable 1. Significant differences were observed be-tween cases and controls regarding education statusof the father, diet and agricultural practices (p �0.05).Cases were categorized into 3 groups—mild, me-dium and severe phenotypes (table 2). Sequencing ofthe exon 5 of the SRD5A2 gene among the mostsevere phenotypes did not yield any mutations.

The allelic distributions for both gene polymor-phisms (SRD5A2-V89L and CYP17-A1/A2) were inHardy-Weinberg equilibrium for cases and controls(p �0.05). Cases and controls differed significantlywith respect to SRD5A2 gene polymorphism, with leu-cine allele being overrepresented among cases vs con-trols (p � 0.01, table 3). Compared to VV genotype, theORs for VL and LL genotypes exhibited 2.05 (95% CI1.0–4.1, p � 0.04) and 3.6 (95% CI 1.5–8.8, p � 0.004)fold increased risk, respectively (table 2). Although theA2 allele of CYP17 gene in the homozygous state dem-onstrated a 2.1-fold increased risk, the association wasnot statistically significant.

Table 1. Parent demographics

Characteristic No. Cases (%) No. Controls (%) p Value

Maternal age at first birth (yrs): 0.160Younger than 20 13 (16.25) 13 (13)21–30 66 (82.5) 80 (80)Older than 30 1 (1.25) 7 (7)

Maternal age at menarche (yrs): 0.142Younger than 14 17 (21.25) 31 (31)14 or Older 63 (78.75) 69 (69)

Birth wt of affected child (gm): 0.092Less than 2,500 15 (18.75) 10 (10)2,500 or More (normal) 65 (81.25) 90 (90)

Maternal education level: 0.064Low 29 (36.25) 21 (21)Medium 24 (30) 33 (33)High 27 (33.75) 46 (46)

Paternal education level: �0.05Low 42 (52.5) 9 (9)Medium 22 (27.5) 10 (10)High 16 (20) 81 (81)

Folic acid supplements inpregnancy:

0.484

Yes 39 (48.75) 54 (54)No 41 (51.25) 46 (46)

Diet: �0.05Vegetarian 51 (63.75) 38 (38)Nonvegetarian 29 (36.25) 62 (62)

Parental occupation:Agricultural 57 (71.25) 0 (0)

When analyzing the risk of hypospadias with re-spect to phenotypic severity, we observed a signifi-cant difference in genotypic distribution of SRD5A2-LLand CYP17-A2/A2 genotypes between cases of me-dium phenotype and controls (table 4). Regardingsevere forms, CYP17 gene polymorphism seemed tohave no role. However, genotypic distribution for LLand VL plus LL genotypes was statistically signifi-cant between severe forms and controls (table 4).

A majority of children (71.25%) with hypospadiashad parents from rural backgrounds with agricultureas the primary occupation. Although our study did notestimate pesticide levels among cases, their role in theetiology of hypospadias cannot be ruled out. Comparedto VV genotype, odds ratios for LL and VL plus LLgenotypes exhibited increased risk in children of par-ents with an agricultural background (likely exposedto pesticides) compared to controls with no such expo-sure (table 5). To avoid the occupation specific distri-bution bias of V89L polymorphism, 82 ethnicity, geogra-phy and language matched individuals from agriculturalbackgrounds were screened and considered as agricul-tural controls. Controls with and without an agricul-tural background did not differ significantly regardingV89L polymorphism (Chi-square � 1.44, p � 0.486).When multivariate analysis was performed ad-justing for diet, occupation and parental educa-tion, a significant risk of hypospadias was foundwith genotypes of V89L polymorphism (VL OR3.01, 95% CI 1.87– 4.8, p � 0.00; LL OR 1.9, 95%CI 1.05–3.4, p � 0.03; VL plus LL OR 2.57, 95% CI1.79 –3.68, p � 0.000).

DISCUSSION

The enzyme 5�-reductase type 2 encoded by SRD5A2gene converts testosterone to the more potent dihy-drotestosterone and is required in the genitourinarytract for normal development of the male externalgenitalia. The frequency distribution of the 3 pheno-types of hypospadias based on severity suggests that

Table 2. Frequency distribution of hypospadias cases basedon phenotypic severity

Phenotypic Variability No. CasesFrequency

Distribution (%)

Mild phenotype: 8.51Coronal 7Glanular 1

Middle phenotype: 35.1Mid penile 13Distal penile 20

Severe phenotype: 56.38Penoscrotal 36Proximal penile 8Scrotal 9

our study population was more prone to exhibit

ROLE OF SRD5A2 AND CYP17 GENE POLYMORPHISMS IN HYPOSPADIAS 2337

severe hypospadias, which is often associated withmale pseudohermaphroditism (table 2). Our find-ings suggest that the SRD5A2-V89L polymorphismis a strong determinant of hypospadias risk amongthe studied population. This assumption was fur-ther ascertained by multivariate analysis, whichwas performed after adjusting for diet, and parentaloccupation and education.

In this study the risk of hypospadias was furtherincreased in the presence of leucine allele in ho-mozygous form (table 3). Our results are in agree-ment with previous reports suggesting that inade-quate levels of DHT (due to high frequency of Lallele) cause hypospadias of varying degrees.6,15,20

Similar findings were reported by Wang et al in aChinese population (Mongoloid background), wherea higher incidence of leucine allele was found amonghypospadias cases vs controls.16 The frequency of Lallele varies across populations of diverse ethnici-ties, ranging from 18.3% among Senegalese to 57.1%among Chinese. It can be ascertained that althoughdistribution of the valine allele of V89L polymor-phism parallels prostate cancer risk, populationswith a high incidence of leucine allele similarly mayhave a greater risk of hypospadias or abnormal gen-italia formation.

The etiology of hypospadias is often assumed to bemultifactorial, implicating a gene-environment in-

Table 3. Genotypic and allelic distribution of SRD5A2 and CYP

Genotype Cases Controls

No. SRD5A2-V89L (%):V/V 19 (23.75) 43 (43.0V/L 40 (50.0) 44 (44.0L/L 21 (26.25) 13 (13.0V/L � L/L 61 (76.25) 57 (57.0

% L allele frequency 51.25 35.0No. CYP17-A1/A2 (%):

A1/A1 21 (26.25) 33 (34.38A1/A2 40 (50.0) 49 (51.04A2/A2 19 (23.75) 14 (14.58A1/A2 � A2/A2 59 (73.75) 63 (65.62

% A2 allele frequency 48.75 40.10

Table 4. Distribution of SRD5A2 and CYP17 genotypes by hypo

Middle Phenotype

% Crude OR 95% CI p V

SRD5A2-V89L:V/V 22.22 1.00V/L 51.88 2.2 0.8–6.4 0L/L 25.9 3.8 1.1–13.5 0V/L � L/L 77.78 2.6 0.9–7.1 0

CYP17-A1/A2:A1/A1 17.4 1.00A1/A2 55.1 2.7 0.8–8.7 0A2/A2 27.5 4.7 1.2–18.24 0

teraction in the development of this anomaly. A par-adox exists where 71.25% of the individuals withhypospadias in our study had parents with agricul-ture as a primary occupation. A number of humanepidemiological studies have examined the risk ofhypospadias in the offspring of parents based onregional agricultural practices and pesticide expo-sure.21,22 When we compared parents working inagriculture (likely exposed to pesticides) to controls(no exposure), a statistically significant associationwas observed between LL genotype and hypospadiasin children (table 5), which suggests that the risk ofhypospadias further increases in individuals carry-ing the L allele who are likely exposed to pesticides(endocrine disrupting chemicals that are antiandro-genic in nature).

To confirm whether agricultural community (ingeneral) does not have a high incidence of V89Lpolymorphism, we screened 82 randomly selectedethnicity, geographic and language matched maleswithout hypospadias belonging to agricultural com-munities for V89L genotypes. A nonsignificant (p �0.05)difference was observed between randomly selectedagricultural controls (L allele frequency 37.8%) andcontrols with no agricultural background (35%) re-garding V89L polymorphism. This finding indicatesthat the North Indian population in general (in theabsence of hypospadias) has a lower frequency of

e polymorphisms

Crude Odds Ratio 95% CI p Value

1.00 Reference2.05 1.0–4.1 0.043.6 1.5–8.8 0.0042.4 1.2–4.6 0.007

1.00 Reference1.2 0.6–2.5 0.492.1 0.8–5.1 0.1211.4 0.7–2.8 0.255

s severity

Severe Phenotype

Controls% Crude OR 95% CI p Value

23.9 1.00 43.050.0 2.04 0.8–4.6 0.10 44.026.0 3.6 1.2–10.0 0.02 13.076.1 2.4 1.1–5.2 0.03 57

35.5 1.00 33.040.1 0.75 0.33–1.69 0.53 49.024.4 1.6 0.6–4.3 0.44 14.0

17 gen

))))

))))

spadia

alue

.14

.04

.07

.12

.02

.35

ROLE OF SRD5A2 AND CYP17 GENE POLYMORPHISMS IN HYPOSPADIAS2338

leucine allele compared to hypospadias cases overall(L allele frequency 51.25%) or cases from agricul-tural backgrounds (54.62%). These observationsstrengthen the findings of Thai et al that a slightlyless well functioning enzyme enables other mecha-nisms to intervene in urethral development.6 It canalso be ascertained that when genetic susceptibilityis combined with exposure to pesticides, a thresholdis surpassed, resulting in the manifestation of thisbirth defect.23 No significant association was foundbetween CYP17-A2 allele and hypospadias, which sug-gests that an increase in transcription rate caused bythe A2 allele might not hold any functional signifi-cance in the etiology of hypospadias or disorders re-lated to ambiguous genitalia.

CONCLUSIONS

As this study brings out the possible role of V89Lpolymorphism of SRD5A2 gene in the etiology ofhypospadias, more replicative series involving largescale population screenings for this polymorphism

Table 5. Comparison between parents with and without agricu

Parents in Agriculture

% Crude OR 95% CI p

SRD5A2-V89L:V/V 22.22 1.00V/L 46.29 2.03 0.9–4.5 0L/L 31.49 4.6 1.7–12.3 0V/L � L/L 77.78 2.6 1.2–5.6 0

CYP17-A1/A2:A1/A1 25.93 1.00A1/A2 48.14 1.2 0.5–2.7 0A2/A2 25.93 2.3 0.9–6.2 0A1/A2 � A2/A2 74.07 1.5 0.7–3.1 0

are needed. Improvements in hypospadias surgery

REFERENCES

Environ Health Perspect 2001; 109: 1175. 80: 3689.

and assisted reproductive technologies during thelast 50 years have led to improved reproductive per-formance of hypospadias cases. Since the V89L poly-morphism was associated with hypospadias in ourstudy population, leucine allele is likely to be passedon to subsequent generations, further enhancing therisk of hypospadias. Based on the frequency distri-bution of leucine allele in world populations, it isnoteworthy that leucine allele may not be suffi-ciently deleterious to cause major morbidity or mor-tality. However, its strong association with hypos-padias found in the present population of NorthIndia, which has a relatively common gene pool,raises concern. The further strengthening of the as-sociation among children whose parents work in anagricultural setting (likely exposed to pesticides)makes the situation grim, especially in India, wherethe majority of the population is from an agricul-tural background. Therefore, the presence of leucineallele among individuals practicing farming is likelyto increase the propensity of having a child with

ackground and controls

Parents Not in Agriculture

Controls% Crude OR 95% CI p Value

26.93 1.00 43.057.69 2.09 0.7–5.6 0.15 44.015.38 1.8 0.4–7.4 0.45 13.073.07 2.04 0.7–5.3 0.17 57

26.92 1.00 33.053.85 1.3 0.5–3.7 0.62 49.019.23 1.6 0.4–6.2 0.49 14.073.08 1.4 0.5–3.7 0.49 63

hypospadias.

1. Paulozzi LJ, Erickson JD and Jackson RJ: Hypos-padias trends in two US surveillance systems.Pediatrics 1997; 100: 831.

2. Myrianthopoulos NC and Chung CS: Congenitalmalformations in singletons: epidemiologic sur-vey. Report from the Collaborative Perinatal Proj-ect. Birth Defects Orig Artic Ser 1974; 10: 1.

3. Chavez GF, Cordero JF and Becerra JE: Leadingmajor congenital malformations among minoritygroups in the United States, 1981–1986. MMWRCDC Surveill Summ 1988; 37: 17.

4. Gallentine ML, Morey AF and Thompson IM:Hypospadias: a contemporary epidemiologic as-sessment. Urology 2001; 57: 788.

5. Baskin LS, Himes K and Colborn T: Hypospadiasand endocrine disruption: is there a connection?

6. Thai HT, Kalbasi M, Lagerstedt K et al: The valineallele of the V89L polymorphism in the five alphareductase gene confers a reduced risk for hypos-padias. J Clin Endocrinol Metab 2005; 90: 6695.

7. Chen T, Li Q, Xu J et al: Mutation screening ofBMP4, BMP7, HOXA4 and HOXB6 genes in Chi-nese patients with hypospadias. Eur J HumGenet 2007; 15: 23.

8. Auchus RJ and Miller WL: Molecular modeling ofhuman P450c17 (17-hydroxylase/17,20-lyase): in-sights into reaction mechanisms and effects ofmutations. Mol Endocrinol 1999; 13: 1169.

9. Morishima A, Grumbach MM and Simpson ER:Aromatase deficiency in male and female siblingscaused by a novel mutation and the physiologicalrole of estrogens. J Clin Endocrinol Metab 1995;

10. Lunn RM, Bell DA, Mohler JL et al: Prostatecancer risk and polymorphism in 17 hydroxylase(CYP17) and steroid reductase (SRD5A2). Carci-nogenesis 1999; 20: 1727.

11. Gsur A, Bernhofer G, Hinteregger S et al: Apolymorphism in the CYP17 gene is associatedwith prostate cancer risk. Int J Cancer 2000; 87:434.

12. Carey AH, Waterworth D, Patel K et al: Polycysticovaries and premature male pattern baldness areassociated with one allele of the steroid metab-olism gene CYP17. Hum Mol Genet 1994; 3: 1873.

13. Feigelson HS, Coetzee GA, Kolonel LN et al: Apolymorphism in the CYP17 gene increases therisk of breast cancer. Cancer Res 1997; 57: 1063.

14. di Salle E, Giudici D, Radice A et al: PNU 157706,

ltural b

Value

.11

.001

.013

.69

.09

a novel dual type I and II 5 alpha-reductase

ROLE OF SRD5A2 AND CYP17 GENE POLYMORPHISMS IN HYPOSPADIAS 2339

inhibitor. J Steroid Biochem Mol Biol 1998; 64:179.

15. Makridakis N, Ross RK, Pike MC et al: A preva-lent missense substitution that modulates activityof prostatic steroid 5 alpha-reductase. CancerRes 1997; 57: 1020.

16. Wang Y, Li Q, Xu J et al: Mutation analysis of fivecandidate genes in Chinese patients with hypos-padias. Eur J Hum Genet 2004; 12: 706.

17. Sasaki G, Ogata T, Ishii T et al: Micropenis andthe 5 alpha reductase-2 (SRD5A2) gene: mutationand V89L polymorphism analysis in 81 Japanese

18. Vilchis F, Hernandez D, Canto P et al: Codon 89polymorphism of the human 5 alpha steroidreductase type 2 gene. Clin Genet 1997; 51:399.

19. Woolf B: On estimating the relation betweenblood group and disease. Am J Hum Genet 1995;19: 251.

20. Silver RI and Russell DW: 5 Alpha-reductase type2 mutations are present in some boys with iso-lated hypospadias. J Urol 1999; 162: 1142.

21. Fernandez MF, Olmos B, Granada A et al:

icals and prenatal risk factors for crypto-chordism and hypospadias: a nested case-control study. Environ Health Perspect, suppl.,2007; 115: 8.

22. Pierik FH, Burdorf A, Deddens JA et al: Mater-nal and paternal risk factors for cryptorchidismand hypospadias: a case-control study in new-born boys. Environ Health Perspect 2004; 112:1570.

23. Manson JM and Carr MC: Molecular epidemi-ology of hypospadias: review of genetic andenvironmental risk factors. Birth Defects Res A