I Med Genet 1994;31:187-192

Genetic study of indirect inguinal hernia

Yaoqin Gong, Changshun Shao, Qian Sun, Binxi Chen, Yuan Jiang, Chenhong Guo,Jianjun Wei, Yishou Guo

AbstractWe performed a genetic analysis of 280families with congenital indirect inguinalhernia ascertained in Shandong province.The multifactorial threshold model andsegregation analysis were applied to thesefamilies to investigate the mode of inherit-ance ofcongenital indirect inguinal hernia.Our results indicate that congenital in-direct inguinal hernia is not compatiblewith a multifactorial threshold model,and the frequent vertical transmission andhigh segregation ratio suggest autosomaldominant inheritance with incompletepenetrance and sex influence. Throughfurther pedigree analysis of the multiplecase families with at least two closelyrelated affected members, we notedpreferential paternal transmission ofthe disease gene, which might suggest therole of genomic imprinting in the aetio-logy of this condition.

(JT Med Genet 1994;31:187-192)

Department ofMedical Genetics,Shandong MedicalUniversity, Jinan,Shandong, P R China250012Y GongC ShaoQ SunB ChenY JiangC GuoJ WeiY Guo

Correspondence toDr Gong.Received 14 June 1993Revised version accepted forpublication 11 October 1993

Inguinal hernia (IH) is a common develop-mental anomaly; its incidence has been esti-mated to be 6 to 12-5% in the different generalpopulations studied.' IH can be classified intodirect inguinal hernia (DIH) and indirectinguinal hernia (IIH), the latter being far morecommon than the former. IIH is further sub-divided into congenital and acquired IIH.Congenital IIH is generally considered to re-sult from a congenital weakness in the internalring and a persistent processus vaginalis andcan be detected during the first months oryears after birth.Although there has been considerable evid-

ence suggesting the role of genetic factors inthe aetiology of IH,2' its mode of inheritanceremains controversial. Hypotheses proposedinclude (1) autosomal dominant inheritancewith incomplete penetrance,34 (2) autosomaldominant inheritance with sex influence,5 (3)X linked dominant inheritance,6 and (4) poly-genic inheritance.7-9

In a recent survey ofmajor genetic disease inthe general population of Shandong pro-vince,'0 we identified 280 families with con-genital IIH. We applied the multifactorialthreshold (MFT) model and segregation ana-lysis to these data to investigate the mode ofinheritance in congenital IIH. Our results in-dicate that congenital IIH is not compatiblewith an MFT model, and the frequent verticaltransmission and high segregation ratiosuggest autosomal dominant inheritance withincomplete penetrance and sex influence infamilial IIH. Through further pedigree ana-

lysis of multiple case families with at least twoclosely related affected members, we noted thepreferential paternal transmission of the dis-ease gene, suggesting the role of genomicimprinting in the aetiology of this condition.

Materials and methodsFAMILIESThe cases used in this study were identified inthe project "Survey of Major Genetic Diseasesin Shandong Province" which was carried outin the five year period from 1985 to 1990.10The survey sites were located in 11 prefecturesand covered a population of about one million.The work for IIH was carried out in a twostage procedure. During the first stage thepatients with IIH were registered by thetrained village or community doctors. Eachpatient identified in the first step was regardedas a proband (index patient). In the secondstage, a thorough clinical examination wasconducted on these index patients and theiraffected first degree relatives, then the pedi-gree was drawn. Among 392 index patients, 52index patients with acquired IIH wereexcluded from further analysis. The 340 pro-bands (319 male, 21 female) had all beenoperated on by 5 years. These probands werefrom 280 families. The hernia occurred on theright side in 138 probands, on the left side in84 probands, and on both sides in the other118 probands.

GENETIC ANALYSISThe observed characteristics of familial ag-gregation of IIH were compared with theexpected values based on an MFT model todiscriminate between multifactorial and men-delian inheritance."-" These characteristicsinclude correlations of the proband's sex to therecurrence risk of relatives, and the severity ofthe proband's defect to the recurrence risk ofrelatives. These correlations were evaluated inthe first degree relatives of different types ofprobands.

Classical segregation analysis was per-formed on 341 nuclear families. Accordingto the parents' mating type, we dividednuclear families into three groups: (1) oneparent affected (U x A matings), (2) bothparents normal with positive family history,that is, one parent has a first or second degreerelative with IIH (U x U(f) matings), and(3) both parents normal (U x U matings). Weused the following distributions for the expec-tation of families of size s with only one affec-ted sib (1) and with more than one affected sib(2).

187

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from

Gong, Shao, Sun, Chen, Jiang, Guo, Wei, Guo

FmIly2

Family A

_ Al

Family B Family C Family DFm>

Famil

Figure I Families with multiple cases of IIH.

A

E

F G HK

Figure 2 Illustration of transmission types. Familiesand B, vertical transmission through two generations.Families C, D, and E, vertical transmission throughthree generations. Families F, G, and H, verticaltransmission through four generations. Family I, verttransmission through five generations. Families _7 andnon-vertical transmission.

P(r = 1) ={sp[x + (1 -x)qs-l]}/{sptx +(1 -x) [1 -(1 -p)S]}

P(r > 1) = {(1-x) [I1-(I - p7r)s-spirqs-ID{Sp11X + (1x)[l(1 p7 )S],j

where s is the sibship size, r the numberaffected sibs, p the segregation frequencyequal to 1 - p, x the frequency of spora

cases, and the probability of ascertainmeThe values of p and x were estimated un

incomplete selection (with n = 0 78) bymaximum likelihood (ML) method.14

Table 1 Percentages of different types offamilies among the multiple case families

Transmission No of generations Family type No offamilies

Vertical 2 A 23B 15

38 (48 7)3 C 3

D 7E 2

12 (15 4)4 F 1

G 1

H 24 (5.1)

5 I 1 (1-3)Non-vertical J 15

K 823 (29 5)

Table 2 Risk to first degree relatives by proband's sex*

Proband's sex Total Ist degree relatives Affected 1st degree relatives Incidence (%

Male 1030 64 6 21

Female 72 5 6-94

* p > 0.05.

Table 3 Risk to first degree relatives by IIH type of probands*

IIH type of proband Total 1st degree relatives Affected 1st degree relatives Incidence

Bilateral 253 16 0.0632Unilateral

Right 509 32 0-0629Left 340 21 0 0618Total 849 53 0-0624

*p > 0.05.

716 this method, the fit of a parametric value, sayp, to the data is tested by Xp2= (Up)2/Kpp,

O with 1 df, where Up is the ML score and; Kpp its variance. The iteration procedure

will lead to the best estimate of the para-y E meter.14

ANALYSIS OF PARENTAL ORIGIN OF THE DISEASE

GENEThe criteria for the multiple case families used

* to analyse the parental origin of the IIH gene1 were: (1) there were at least two affected mem-, bers; (2) affected members were distributed inl at least two sibships; and (3) the relationship; between them was up to third degree. The

illustration of how these criteria were used areshown in fig 1. Using these criteria, families B,

s A C, and D in fig 1 have been included in ourstudy, while family A and E have beenexcluded.~ical Ac

IK, According to the above criteria, a total of128 families was identified, of which 70 fami-lies were selected from the 280 families, theother 58 families coming from previously pub-lished studies,'5 which were ascertained

(1) through 58 probands who underwent opera-tions in Jinan Children's Hospital during 1989to 1991. Phenotypically normal carriers who

(2) had carried and transmitted the disease gene tothe next generation were identified by pedigree

of analysis. The sex of each carrier and affectedq parent was recorded.

idicmnt.der Resultsthe PEDIGREE ANALYSESrith Of the 280 families, there was a positive his-

tory in 78 families (27 85%). In 202 families(72 14%), the index patient was the onlyknown affected family member. The illustra-

) tion of different family types is shown in fig 2.In 55 of the 78 families with a positive history,the trait was transmitted vertically (table 1).Direct male to male transmission was noted in29 families. The vertical and male to maletransmission of the trait in these familiessuggests that familial IIH is inherited in auto-somal dominant fashion.

TEST OF THE MFT MODELIn this data set, the sex ratios in probands andaffected subjects were 15 32 (319:21) and 13 36(374:28), respectively. According to the MFTmodel, when the incidence is higher in one sexthan in the other, the less affected sex would beexpected to show a higher recurrence risk intheir first degree relatives compared to themore affected sex. However, as is shown intable 2, the incidence of I IH in the first degreerelatives of female probands was not signific-antly higher than that of male probands(p> 005).Another prediction with a similar theoretical

basis is that the more severely affected pro-bands would be expected to show a higherrisk in their first degree relatives compared tothe less severely affected probands. In thisstudy, the incidence of IIH in the first degree

188

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from

Genetic study of indirect inguinal hernia

Table 4 Distributions of affected males (r) among male sibs (s)

UXU UXU (f) UXA

s/r 1 2 Total slr 1 2 3 Total s/r 1 2 Total

1 114 114 1 44 44 1 21 212 49 5 54 2 20 9 29 2 5 2 73 27 2 29 3 12 2 14 3 4 44 10 1 11 4 5 1 6 4 1 1 25 2 2 5 2 1 3 5 1 1 2Total 202 8 210 83 11 2 96 32 4 36

relatives of bilateral probands was not signific-antly higher than that of unilateral probands(table 3). A similar result was observed in thecomparison of probands with right IIH andleft IIH.

SEGREGATION ANALYSISThe above analyses suggest that dominantinheritance may be more likely in these fami-lies. As noted in table 4, we have 36 nuclearfamilies with one parent affected and 306 fami-lies with both parents unaffected, of which 96had a positive family history; thus completepenetrance can be ruled out for the dominanttransmission model. As most cases of IIH aremales, segregation analyses were performed onmale sibships only. The results of segregationanalyses for the three groups are summarisedin table 5.

Since the proportion of sporadic cases, ifany, may be unique for the families with oneparent affected, p and x were simultaneouslyestimated from the data. The iteration proced-ure resulted in the final estimates of p = 0-225and x = 0 013, showing that among non-spora-dic cases the segregation frequency is 0-225. Ifthese families were considered as representingdominant inheritance, the penetrance is 0 225/0-5 = 0.45.For the sibships of UXU (f), p = 0-225 and

x = 0 013 have an excellent fit to the data, withXp2 = 0-032 and xx2 = 0041. This is consistentwith the speculation that familial IIH showsdominant inheritance with incomplete pene-trace.For those families with a negative family

history, the proportion of sporadic cases is farfrom zero (Xx2= 13 28, p<001). The best es-timate of x is 0 599 for this group of families.

PARENTAL ORIGIN OF THE IIH GENEA total of 159 sibships with affected or carrierparents was included in the 128 pedigreesselected. In 58 of 60 sibships the father was theaffected parent, while the mother was affectedin only two sibships. In 66 of 99 sibships thephenotypically normal father carried andtransmitted the disease gene; in the other 33sibships the disease genes were transmitted

Table 5 Segregation analyses of IIHfamilies

Group No offamilies p x XP2 xx2

UxU 210 0-225 0 12 26 13-28210 0-225 0-599 0-04 0

U x U(f) 96 0-5 0 23-12 31-0296 0-225 0-013 0-03 0-04

UxA 36 0 5 0 7-54 7-2736 0 225 0-013 0 0

Table 6 Number of sibships with affected or carrierparent

Parent Affected Carrier Total

Father 58 66 124Mother 2 33 35Total 60 99 159

Table 7 Number of sibships with affected and carriergrandparents

Grandparent Affected Carrier Total

Grandfather 32 3 35Grandmother 2 1 3Total 34 4 38

Table 8 The parental origin of the IIH gene bypedigree analysis

Group Male affected Female affected

Total 177 11Paternal 140 (79-1%) 6 (54-5%)

Affected father 68 3Carrier father 72 3

Maternal 37 (20 9%) 5 (45-5%)Affected mother 1 0Carrier mother 36 5

from the carrier mother. A striking differenceexists in the sex distribution of the affectedparents and carriers (table 6).Data on the grandparental generation are

shown in table 7. Though not complete, theyshow a similar tendency of distorted sex ratioto that of the parental generation.The results described above suggest that the

affected persons might have inherited the IIHgene more frequently from their father thanfrom their mother. Therefore, the parentalorigin of the IIH gene was determined separ-ately for affected males and females. The re-sults show that the majority of the affectedmales (79-1%, table 8) inherited the IIH genefrom their father (p < 0 01). However, the factthat females rarely develop IIH means thatthere are more families with an affected fatherthan with an affected mother and makes itinvalid to compare the number of affectedfathers with the number of affected mothers todetermine the origin of the IIH gene. Wetherefore compared only the numbers of car-rier fathers and carrier mothers. Of the 108affected males with a carrier parent, 72 had acarrier father and 36 had a carrier mother, theformer being far more than the latter(X2 = 12-00, p < -05). Unfortunately, the femaleaffected sample is too small to be analysed.

DiscussionMost cases of IIH arise from retention orimperfect obliteration of the processus vagina-lis, the embryological outpocketing of perito-neum that precedes testicular descent into thescrotum. The testes originate along the uro-genital ridge in the retroperitoneum and mi-grate caudally during the second trimester ofpregnancy to arrive at the internal inguinal

189

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from

Gong, Shao, Sun, Chen, Jiang, Guo, Wei, Guo

(abdominal) ring at about the sixth month ofintrauterine life. During the last trimester,they proceed through the abdominal wall viathe inguinal canal and descend into the scro-tum, the right slightly later than the left. Theprocessus vaginalis then normally disappearspostnatally except for that portion surround-ing and serving as a covering for the testis.Failure of this obliterative process results incongenital IIH. Acquired IIH probably re-sults from the stress of muscular activity andthe increase in intra-abdominal pressure forc-ing open a previously imperfectly obliteratedprocessus vaginalis. However, it is not clear atthis time whether congenital and acquired IIHare basically different from one another, orwhether they can be considered to be differentmanifestations of the same or a similar defectaffecting different parts of the inguinal region.Therefore, only the probands with congenitalIIH were included in this study.

Previous population studies have suggestedthat congenital IIH is inherited under a multi-factorial threshold model.7"9 In order to assesswhether the multifactorial threshold modelcan explain IIH clustering in this data set, twopredictions were tested.

First, IIH occurs more commonly in malesthan in females. This is because of the defect inthe abdominal wall occasioned by the descent ofthe testes into the scrotum in the male and alsoperhaps because of the protection to the insideof the lower ventral abdominal wall afforded bythe uterus in females. Under the multifactorialthreshold model, when incidence is different inone sex from the other, risk in relatives dependson the sex of the probands. Contrary to thisexpectation, the present series shows thatthe proband's sex was independent of therecurrence risk to the first degree relatives.

Second, the right testis descends later thanthe left and the processus vaginalis is thereforeobliterated later on the right side than on theleft side; hence hernia is more frequent onthe right than on the left side. If IIH wereinherited under a multifactorial thresholdmodel, a higher threshold would be expectedfor those subjects with left IIH, and con-sequently a higher risk would be expected inrelatives of left IIH than in relatives of rightIIH. Similarly, a higher risk would also beexpected in the first degree relatives of bi-lateral probands than in relatives of unilateralprobands. This was not the case for eithercomparison.

Tests of the predictions of the MFT modeldo not support the assumption of multifactor-ial inheritance in IIH; rather, we considerautosomal dominant inheritance with incom-plete penetrance and sex influence more likely,based on the strong evidence provided bypedigree and segregation analysis.

In families with familial IIH, three import-ant criteria for the model mentioned above aresatisfied. First, vertical and male to male trans-mission of the trait establish the autosomaldominant pattern. Second, some parents ofaffected subjects are not affected, but theyobviously carried and transmitted the deleteri-ous gene. This can be considered as evidence

of reduced penetrance. Third, the finding thataffected males are much more numerous thanaffected females suggests that the trait isinfluenced by sex. This sex influence couldbe attributed to the anatomical differencebetween males and females.

Segregation analyses also support autosomaldominant inheritance of IIH. The estimatedsegregation ratio and penetrance were 0 225and 0 45, respectively. As expected, the pro-portion of sporadic cases was unique for A x Uand U x U (f) matings. In U x U matings,sporadic cases were estimated to account formost cases.

It is interesting to compare IIH with twoother common and extensively investigatedcongenital malformations, cleft lip with orwithout cleft palate and neural tube defects,which also result from incomplete closure ofthe organs involved. Although previous stud-ies have shown that these two conditions havemultifactorial inheritance, S-8 recent analyseshave provided evidence of involvement ofmajor genes.'925 Thus it is plausible to specu-late that morphorgenesis may be determinedby single genes and complicated by environ-mental factors. If the IIH gene has a "suscep-tibility to the environment", it is quite likelythat incomplete penetrance or variable pheno-copies or both may be present in IIH, and thenanalyses will result in non-mendelian estimatesof the transmission probabilities. This shouldnot necessarily be interpreted as evidenceagainst the major gene hypothesis.26

In summary, these data, regardless of theanalytical approach, indicate that there isstrong statistical evidence that familial IIH isinherited in an autosomal dominant fashionwith a reduced penetrance and sex influence.As mentioned above, IIH occurs more fre-

quently in males than in females, so familieswith an affected father and son(s) should beexpected to be more common than those withan affected mother and son(s). In the group ofunaffected carrier parents, however, the carrierfathers are also over-represented although theopposite would be expected. Their inheritancepattern is apparently inconsistent with autoso-mal dominant inheritance. This observed pre-ponderance of paternal transmission calls fornew explanations. We discuss two possibilitieshere.

First, ascertainment bias might be respon-sible as affected paternal relatives may be moreeasily identified since they are more familiarwith each other. However, in the other dis-eases we surveyed by the same methods usedfor the selection of IIH families, such prepon-derance was observed in different sexes fordifferent diseases. For example, an excess ofmaternal transmission was noted in the multi-ple case families with epilepsy, while an excessof paternal transmission occurred in the multi-ple case families with cleft lip with or withoutcleft palate (Gong et al, unpublished data).The fact that male/female ratios of transmit-ters (normal and affected) shift towards dif-ferent sexes in different diseases argues againsta significant bias of ascertainment in oursample.

190

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from

Genetic study of indirect inguinal hernia

Second, sex specific genomic imprintingcould account for the preferential paternaltransmission of the IIH gene. We assume thatthe IIH gene is not expressed or expressed at alow level if it is transmitted from the mother(this kind of imprinting has been referred to asmaternal imprinting27) in contrast to paternalimprinting in which the paternal homologue isinactivated or expressed weakly in the off-spring.Homologous chromosomes may undergo

different modifications when they segregateduring gametogenesis. These modifications ortheir effects may persist through embryogene-sis and distinguish the maternal and paternalalleles or regions until adult stage. Suchimprinted information can apparently result indifferential activity of parental alleles. Inmammals, the first convincing experimentalevidence for genomic imprinting was obtainedin marsupial and mouse X chromosomes.2829 Itwas found later that this phenomenon was notrestricted to the X chromosomes, as autosomesmay also be involved,30 and not restricted tothe marsupials and mice as in a number ofhuman disorders the phenotypic differencesare also related to the parental origin of thedisease gene. For example, a congenital andsevere form of myotonic dystrophy occurs in10 to 20% of myotonic dystrophy familieswhen the gene is transmitted through themother,31 and in 5 to 10% of families where theHuntington's disease gene is transmittedthrough the father, a severe, rigid, juvenileform of the disease occurs.'2 A similar effect ofgenomic imprinting on gene expression hasbeen observed in spinocerebellar ataxia, seiz-ures,34 cerebellar ataxia,35 and Wiedemann-Beckwith syndrome.36 The genes involved inthese diseases are transmitted in a mendelianmanner, but their expression is determined bythe sex of the parent transmitting the gene.The pedigrees of these disorders usually showirregular inheritance patterns.

Like the disorders mentioned above, in-direct inguinal hernia (IIH) also shows anunusual mode of inheritance. In some families,IIH follows an autosomal dominant pattern,35while in most families a monogenic mode ofinheritance is not apparent.78 Therefore, weare inclined to speculate that a paternal allele ispreferentially involved in determining the clo-sure of the inguinal canal. An abnormal pater-nal gene would result in the failure of closurealthough the maternal allele there might benormal.

In the mouse there have been numerouscases in which genomic imprinting results inlethality or developmental anomalies, for ex-ample, maternal duplication/paternal defi-ciency for the distal segment of mouse chro-mosome 2 resulted in flat sided, arch backed,hypokinetic newborn; those mice of the reci-procal type had short, square bodies andbroad, flat backs and were notably hyperkine-tic37; and maternal X chromosome disomyresulted in developmental failure of the tro-phoectoderm cell lineage.'8 In man there havebeen few reports of congenital malformationswhose inheritance is affected by genomic

imprinting. One such example is Wiedeman-Beckwith syndrome (WBS), which is charac-terised by macroglossia, gigantism, earlobepits or creases, abdominal wall defects, and anincreased risk for the development of tumours.In WBS the paternally transmitted alleles atthe WBS locus were supposed to be func-tionally inactivated to account for the rarity oftransmitting males.'6

In conclusion the fact that most affectedmales had inherited the IIH gene from theirfather implicates a role of genomic imprintingin the aetiology of the IIH phenotype. Furtherfollow up of the familial IIH families will allowtesting of this hypothesis. Our findings for IIHrepresent a step towards understanding thedevelopmental basis of this common humangenetic disorder.

1 McVay CB. Christopher's textbook of surgery. Philadelphia:Saunders, 1956.

2 Warren RR, Atleson FW. Inheritance of hernia in a familyof Holsterin-Fresian cattle. J Hered 1931;22:345-9.

3 West LS. Two pedigrees showing inherited predispositionto hernia. Jf Hered 1936;27:449-55.

4 Smith MP, Sparkes RS. Familial inguinal hernia. Surgery1965;57:807-12.

5 Weimer BR. Congenital inheritance of inguinal hernia. JHered 1949;40:219-20.

6 Montagu AMF. A case of familial inheritance of obliqueinguinal hernia. J Hered 1942;33:355-6.

7 Sawaguchi S, Matsunaga E, Honna T. A genetic study onindirect inguinal hernia. Jpn J Hum Genet 1975;20:187-95.

8 Czeizel A, Gardonyi J. A family study of congenital inguinalhernia. Am J Med Genet 1979;4:247-54.

9 Zhang SL. Epidemiological survey of genetic diseases ingeneral population of Sichuan, China. Chengdu: CSTUniversity Press, 1990.

10 Guo YS, Gong YQ, Shao CS, Wei JJ, Chen BX, Jiang Y. Asurvey for major hereditary diseases in Shandong pro-vince. Acta Acad Med Shandong 1931;31:271-5.

11 Falconer DS. The inheritance of liability to certain diseases,estimated from the incidence among relatives. Ann HumGenet 1965;29:51-76.

12 Carter CO. Genetics of common disorders. Br Med Bull1969;25:52-7.

13 Reich T, James JW, Morris CA. The use of multiplethresholds in determining the mode of transmission ofsemi-continuous traits. Ann Hum Genet 1972;36:163-84.

14 Morton NE. Genetic tests under incomplete ascertainment.AmJ7Hum Genet 1959;11:1-16.

15 Sun Q, Gong YQ, Guo CH. Family analysis on congenitalinguinal hernia. Chin J Med Genet 1991;3: 12-13.

16 Carter CO, Evans K. Spina bifida and anencephalus inGreater London. J Med Genet 1973;10:209-34.

17 Carter CO. Genetics of common single malformations. BrMed Bull 1976;32:21-6.

18 Lalouel JM, Morton NE, Jackson J. Neural tube malforma-tions: complex segregation analysis and calculation ofrecurrence risks. 7Med Genet 1979;16:8-13.

19 Marazita ML, Spence MA, Melnick M. Genetic analysis ofcleft lip with or without cleft palate in Danish kindreds.AmJ Hum Genet 1984;19:9-18.

20 Marazita ML, Goldstein AM, Smalley SL, Spence MA.Cleft lip with or without cleft palate: reanalysis of a threegeneration family study from England. Genet Epidemiol1986;3:335-42.

21 Chung CS, Bixler D, Watanabe T, Kiguchi H, Fogh-Andersen P. Segregation analysis of cleft lip with orwithout cleft palate: a comparison of Danish and Japanesedata. AmJ Hum Genet 1986;39:603-1 1.

22 Chung CS, Beechert AM, Lew RE. Test of genetic hetero-geneity of cleft lip with or without cleft palate as related torace and severity. Genet Epidemiol 1989;6:625-31.

23 Hecht JT, Yang P, Michels VV, Buetow KH. Complexsegregation analysis of nonsyndromic cleft lip and palate.AmJ Hum Genet 1991;49:674-81.

24 Demenais F, Le Merrer M, Briard ML, Elston RC. Neuraltube defects in France: segregation analysis. Am J MedGenet 1982;11:287-98.

25 Shaffer LG, Marazita ML, Bodurtha J, Newlin A, NanceWE. Evidence for a major gene in familial anencephaly.Am J Med Genet 1990;36:97-101.

26 Williams WR, Beutow KH. An explanation of non-Mende-lian transmission frequencies in segregation analysis ofqualitative traits. Am J Hum Genet 1986;39:A248.

27 Hall JG. Genomic imprinting: review and relevance tohuman diseases. Am J Hum Genet 1990;46:857-73.

28 Sharman GB. Late DNA replication in the paternallyderived X chromosome of female Kangaroos. Nature1971;230:231-32.

29 Takagi N, Sasaki M. Preferential inactivation of the pater-nally derived X chromosome in the extraembryonic tis-sues of the mouse. Nature 1975;256:640-2.

191

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from

Gong, Shao, Sun, Chen, Jiang, Guo, Wei, Guo

30 Cattanach BM, Kirk M. Differential activity of maternallyand paternally derived chromosome regions in mice.Nature 1985;315:496-8.

31 Harper PS. Congenital myotonic dystrophy in Britain. II.Genetic basis. Arch Dis Child 1975;50:514-21.

32 Reik W. Genomic imprinting: a possible mechanism for theparental origin effect in Huntington's chorea. MedGenet 1988;25:805-8.

33 Zoghbi HY, Pollack MS, Lyons LA, Ferrell RE, DaignerSP, Beaudet AL. Spinocerebellar ataxia: variable age ofonset and linkage to HLA in a large kindred. Ann Neurol1988;23:580-4.

34 Ottman R, Annegers JF, Hauser WA, Kurland LT. Higher

risk of seizures in offspring of mothers than of fatherswith epilepsy. Am I Hum Genet 1988;43:257-64.

35 Harding AE. Genetic aspects of autosomal dominant lateonset cerebellar ataxia. I Med Genet 1981 ;18:436-41.

36 Koufos A, Grundy P, Morgan K, et al. Familial Wiede-mann-Beckwith syndrome and a second Wilms tumorlocus both map to lIpI5.5. Am 37 Hum Genet1989;44:71 1-19.

37 Cattanach BM. Parental origin effects in mice._3 Embryol

Exp Morphol 1986;97(suppl):137-50.38 Shao C, Takagi N. A maternally derived X chromosome is

deleterious to early mouse development. Development1990;1 10:969-75.

192

on February 28, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.31.3.187 on 1 March 1994. D

ownloaded from