Birth Prevalence, Mutation Rate, Sex Ratio,Parents’ Age, and Ethnicity in Apert Syndrome

Marie M. Tolarova,1* John A. Harris,2 Doris E. Ordway,1 and Karin Vargervik1

1Center for Craniofacial Anomalies, Department of Growth and Development, School of Dentistry, University ofCalifornia, San Francisco, California

2March of Dimes Birth Defects Foundation, California Birth Defects Monitoring Program, Emeryville, California

Apert syndrome was studied to determinebirth prevalence, mutation rate, sex ratio,parents’ age, and ethnicity among 2,493,331live births registered in the California BirthDefects Monitoring Program (CBDMP) from1983 through 1993; 31 affected infants wereidentified. The sample was completed withan additional 22 cases from the Center forCraniofacial Anomalies (CCA), University ofCalifornia, San Francisco, for a total of 53affected children. Birth prevalence, calcu-lated from the CBDMP subsample, was 12.4cases per million live births (confidence in-terval [CI] 8.6,17.9). The calculated mutationrate was 6.2 × 10−6 per gene per generation.Asians had the highest prevalence (22.3 permillion live births; CI 7.1,61.3) and Hispan-ics the lowest (7.6 per million, CI 3.3–16.4). Inthe large population-based CBDMP sub-sample, there was an almost equal numberof affected males and females, (sex ratio0.94) but in the clinical CCA subsample,there were more affected females (sex ratio0.79). For all cases, the mean age of motherswas 28.9±6.0 years, and of fathers was34.1±6.2 years. Almost half of fathers wereolder than 35 years when the child was born;for more than 20% of cases, both parentswere older than 35 years. These findingsmay support the view that point mutationsappear to be more commonly associatedwith paternal than with maternal alleles.Representing the largest systematically as-certained population-based study of Apertsyndrome to date, they provide a reliable

basis for genetic counseling and decision-making, and for focused research to definethe cause of this syndrome. Am. J. Med.Genet. 72:394–398, 1997. © 1997 Wiley-Liss, Inc.

KEY WORDS: Apert syndrome; acrocepha-losyndactyly; birth preva-lence; sex ratio; parental age;epidemiology

INTRODUCTION

Apert syndrome (acrocephalosyndactyly [Apert,1906]), comprises craniosynostosis and severe syndac-tyly and results from a mutation of fibroblast growth-factor receptor-2 (FGFR2) allelic with Crouzon syn-drome [Wilkie et al., 1995]. Since the syndrome wasfirst recognized toward the end of the 19th century[Wheaton, 1894], well over 300 cases have been re-ported [Cohen, 1986], most of them caused by new mu-tations.

It appears that relatively few adults with Apert syn-drome have had children; only 11 such instances arerecorded [Cohen and Kreiborg, 1991; Lewanda et al.,1993]. Extensive studies and reviews covering a widerange of different aspects of Apert syndrome have beenpublished [Cohen, 1975; Cohen and Kreiborg, 1991;Kreiborg and Cohen, 1991; Cohen and Kreiborg, 1993a;Cohen and Kreiborg, 1993b], and recent findings[Moloney et al., 1996] showed an exclusively paternalorigin of the new mutation in 57 Apert families. How-ever, because this syndrome is rare, relatively fewstudies assess a population-based database largeenough to permit evaluation of epidemiological charac-teristics. Thus, information is scant regarding differ-ences in birth prevalence and potential differences inthe sex ratio that may influence any search for theparental origin of new mutations [Sapienza, 1996].

To remedy this deficiency, Cohen et al. [1992] re-ported a large birth-prevalence study of 57 cases basedon pooled data from 7 geographic areas. With similarintent, we report a population-based study to deter-mine birth prevalence, mutation rate, sex ratio, par-ents’ age, and ethnicity in 31 cases of Apert syndromedrawn from a database of 2,493,331 live births in Cali-

Presented in part at the 51st Annual Meeting of the AmericanCleft Palate-Craniofacial Association, Toronto, Canada, May1994.

*Correspondence to: Marie M. Tolarova, MD, PhD, DSc, Centerfor Craniofacial Anomalies, Department of Growth and Develop-ment, School of Dentistry, University of California, San Francis-co, 513 Parnassus Avenue, Medical Sciences Building S-747, SanFrancisco, CA 94143-0442.

Received 7 August 1996; Accepted 9 May 1997

American Journal of Medical Genetics 72:394–398 (1997)

© 1997 Wiley-Liss, Inc.

fornia and an additional 22 cases from a clinicalsample.

MATERIAL AND METHODSCase Definition

Apert syndrome was defined on the basis of cranio-synostosis, midface hypoplasia, and symmetrical syn-dactyly of hands and feet minimally involving digits 2,3, and 4 [Cohen, 1975; Cohen et al., 1992]. The diag-nosis was confirmed based on records from the birthhospital, physical examination, radiology films, geneticevaluations, and surgery reports.

Population Samples

Data were collected as a sample from each of 2 cen-ters. The 1st subsample was drawn from the CaliforniaBirth Defects Monitoring Program (CBDMP), a re-gional population-based registry of congenital anoma-lies currently based on 300,000 annual births [Croen etal., 1991]. CBDMP staff visit all hospitals and outpa-tient genetic centers to abstract data on all childrenwith congenital anomalies diagnosed up to the age of 1year. For each specific anomaly diagnosed, CBDMPstaff record the type of physical and/or medical proce-dure and/or confirmatory test and the name of the spe-cific subspecialist who made the diagnosis. In this way,the accuracy of each diagnosis can be calculated. Dataaccrued over the 11 years from 1983 through 1993 wereused, comprising 2,509,881 births, of which 2,493,331were live births. The counties included in the registryeach year varied; at present the registry covers ap-proximately half of California counties.

The second subsample consisted of children treatedand followed in the Center for Craniofacial Anomalies(CCA) at the University of California, San Francisco(UCSF). The CCA provides highly specialized teamcare and follow-up evaluations for a wide spectrum ofcraniofacial anomalies and, for children who have cra-niosynostosis syndromes, works closely with the De-partment of Neurological Surgery at UCSF.

Seven children were registered in both subsamples.Two children in the UCSF subsample who were bornduring years when the CBDMP registry existed, butwere not registered by the CBDMP, were born in coun-ties where the registry was not active that year.

Methods

For calculations of the birth prevalence, mutationrate, prevalence by mother’s age group, and ethnicity,only the population-based CBDMP subsample was

used. Birth prevalence of Apert syndrome per millionlive births was calculated with 95% upper and lowerconfidence limits based on the Poisson distribution.The mutation rate was calculated as described by Co-hen et al. [1992] and was estimated by using the equa-tion m 4 1⁄2p, in which ‘m’ is the mutation rate per geneper generation, and ‘p’ is the frequency of sporadiccases at birth. For this estimate, it was assumed thateach sporadic case represented a de novo mutation. Toevaluate age of the parents, we assessed each sub-sample separately and then the whole sample. For sta-tistical comparisons, where appropriate, we used thet-test. When not otherwise specified, statistical testswith a P value of less than 0.05 were considered sig-nificant; P values greater than 0.05 but less than 0.10were considered to be of borderline significance.

RESULTSPopulation Samples

Data for this study were collected as a sample fromeach of 2 centers accruing case records of children whohad a diagnosis of Apert syndrome. The first sub-sample, drawn from the California Birth Defects Moni-toring Program (CBDMP) registry, consisted of 31cases derived from the California population of2,493,331 live births during the years 1983–1993. The2nd consisted of 29 cases, children born between 1976and 1995 who were treated and followed-up in the Cen-ter for Craniofacial Anomalies (CCA) of the Universityof California, San Francisco (UCSF). All were liveborn,and all but 2 were born in California. This subsamplereflected the composition of the population of patientsseen at UCSF. Seven children were registered in bothsubsamples. Thus, the total sample comprised 53 chil-dren.

Birth Prevalence

The birth prevalence of Apert syndrome, calculatedfrom the CBDMP registry data for 1983–1993, was 12.4per million live births [confidence interval (CI) 8.3–17.4] (Table I). This figure represents 1 case of Apertsyndrome in every 80,430 live births. A separate evalu-ation of data for the years 1988–1993 in the CBDMPregistry showed a birth prevalence of 12.9 per millionlive births (CI 8.4,19.7).

Mutation Rate

The mutation rate of Apert syndrome derived fromthe California birth population of 2,493,331 births dur-

TABLE I. Estimates of Apert Syndrome Birth Prevalence and Mutation Rate

Author andyear

Geographicarea

Years ofstudy

No. ofcases with

Apertsyndrome

No. oflive births

Birthprevalence

per 1,000,000live births

Confidenceinterval 95%

Mutation rateestimate

(per gene, pergeneration)

Cohen et al., 1992 Pooled from 7populationsa

1958–1989 57 3,686,592 15.46 11.8, 20.2 7.73 × 10−6

Czeizel et al., 1993 Hungary 1980–1989 13 1,309,583 9.93 5.5, 17.5 4.96 × 10−5

Tolarova et al., 1996 California 1983–1993 31 2,493,331 12.43 8.6, 17.9 6.22 × 10−6

aNebraska, Washington, Denmark, Italy, Spain, Atlanta, Northern California.

Apert Syndrome 395

ing 1983–1993 was 6.2 × 10−6 per gene per generation(Table I).

Sex Ratio

In the population-based CBDMP subsample, almostthe same proportion of males and females was affectedwith Apert syndrome (sex ratio 0.94). However, in theCCA subsample 12 males and 17 females were affected,for a sex ratio of 0.7. Evaluation of the whole sampleshowed that the higher proportion of females in theCCA subsample affected the sex ratio, which was 1:1.3or 0.77 (23 males and 30 females).

Age of Parents

Mother’s age. In the CBDMP subsample of chil-dren with Apert syndrome the mean age of motherswas 28.26±5.94 years, whereas in the CCA subsampleit was 29.72±6.06. As there was no significant age dif-ference between the subsamples (t 4 1.68, P < 0.19),we combined them to determine mean age for mothersin the whole sample, which was 28.87±5.98 years. Formothers of male propositi the mean age was 30±5.29years, and for mothers of female index cases it was28±6.4 years.

Prevalence by age groups for the mothers was ana-lyzed on the population-based CBDMP subsample. Thehighest prevalence was 17.1 per million births (CI9.5,30.1) among mothers 25–29 years old. The secondhighest prevalence was 16.5 per million (CI 8.0,32.5)among mothers 30–34 years old. The lowest prevalencewas 4.6 per million births (CI 1.2,14.6) in the subgroupof mothers of 20–24 years old (Table II).

Father’s age. In the CBDMP subsample the meanage of fathers was 32.72±6.1 years, whereas in the CCAsubsample it was 36.15±6.03. The difference betweenmeans for these 2 subsamples was significant (t 41.68, P # 0.03) and larger for the fathers than for themothers. For both subsamples combined, the mean ageof fathers was 34.12±6.24 years. For fathers of maleindex cases, the mean age was 35.36±5.23 years and forfathers of female propositae, 33.1±6.9 years. Almosthalf (46.9%) of the fathers of children with Apert syn-drome in the whole sample were older than 35 years.

Combined ages. The whole sample was also ana-lyzed in regard to age combinations for mothers andfathers. For over 20% of all cases (11 couples), bothparents were older than 35 years (Table III). The sec-ond most common combination (15%) was a mother age25–29 and a father 30–34 years old.

Ethnicity

The registry of birth defects among California’s mul-tiracial population provides an opportunity to evaluatebirth prevalence by race. In this study, birth preva-lence of Apert syndrome among racial groups was cal-culated from the CBDMP subsample only. Because Ap-ert syndrome is a rare condition, even a registry aslarge as the CBDMP yields only a small number ofcases when analyzed by race. With that consideration,the findings of prevalence by ethnicity showed thehighest birth prevalence among Asians, with 22.3 permillion live births (CI 7.1,61.3), and the lowest amongHispanics, with 7.6 per million (CI 3.3,16.4). AmongWhites, who constituted the majority of cases in bothsubsamples (CBDMP 56.7%, CCA 68.9%), the birthprevalence of Apert syndrome was 16.6 per million livebirths (CI 10.2,26.9). There was no case of Apert syn-drome among Blacks in our sample.

DISCUSSION

Table I permits comparison of the birth prevalence ofApert syndrome as determined in this study with theestimated pooled prevalence from 7 sources reported byCohen et al. [1992] and with the birth prevalence in aHungarian study [Czeizel et al., 1993]. Birth preva-lence values in the pooled populations reported by Co-hen’s group [1992] vary from as high as 23.1 per millionlive births in a Nebraska population to as low as 11.2per million in California. However, all values they re-ported are higher than those reported in earlier studiesand those of 9.9 per million found by Czeizel et al.[1993] in Hungary.

In the original work by Cohen’s group [1992], one ofthe pooled populations came from the same source asthat used in our study: CBDMP data from 1983–1987contributed 8 cases to their study, leading to the esti-mated birth prevalence of 11.2 per million live births.The CBDMP subsample was the 3rd largest among thepooled data and contributed a population of 713,028live births (estimate prevalence 11.2 per million; CI5.2,23.1). A separate evaluation of data for the years1988 through 1993 in the CBDMP registry showed abirth prevalence of 12.9 per million live births (CI

TABLE II. Prevalence of Apert Syndrome in 5-Year MaternalAge Groups

Maternalage group

No. ofmothers

Rate per1,000,000live births

95% confidenceinterval

<20 3 10.5 2.7, 33.520–24 3 4.6 1.2, 14.625–29 13 17.1 9.5, 30.130–34 9 16.5 8.0, 32.535+ 3 11.4 2.9, 36.2

TABLE III. Parental Age Combinations

Maternalage group

Paternalage group No. %

<20 20–24 2 3.8Unknown 1 1.9

20–24 20–24 1 1.925–29 1 1.930–34 3 5.735+ 1 1.9

Unknown 2 3.825–29 25–29 5 9.4

30–34 8 15.135+ 6 11.3

Unknown 1 1.930–34 30–34 5 9.4

35+ 6 11.335+ 35+ 11 20.7Total 53 100.0

396 Tolarova et al.

8.4,19.7). A comparison of these estimates suggeststhat the birth prevalence of Apert syndrome in Califor-nia may have increased; however, CIs for each estimateoverlapped.

Although Apert syndrome seems to be a conditioneasily recognized at birth and during infancy, somecases of Apert syndrome are still misdiagnosed atbirth. We think that our method of ascertainment frommultiple sources and follow-up reports from the CCAhas enabled us to avoid misdiagnoses and ascertaincompletely the cases of Apert syndrome that occurredin the specific birth population we targeted. An almostequal number of males and females affected with Apertsyndrome was found when only the population-basedCBDMP subsample was evaluated; and more affectedfemales were found when the total sample was evalu-ated (sex ratio 0.77). These findings corroborate thatthere is either an equal proportion of cases of Apertsyndrome in both sexes, or a slightly greater number ofaffected females. The sample reported by Blank [1960]consisted of 17 males and 20 females (sex ratio of 0.85),and the cases reported previously in the literature,which he summarized, comprised 46 females and 42males (sex ratio 0.91). In the largest series, 136 casesevaluated by Cohen and Kreiborg [1993a,b], the sexratio was 0.94.

Because most previously published studies are basedon samples in which the index case was ascertained atthe medical, or usually the surgical, facility, the expla-nation for the higher prevalence of females may be thatthe parents of girls affected with Apert syndrome aremore likely to bring their child for treatment; theprevalence of females in our CCA subsample could alsobe explained in this way.

In previous reports, an older age of parents has beenassociated with sporadic Apert syndrome [Blank, 1960;Erickson and Cohen, 1974]. In our study, the mean ageof mothers was 28.87 years and of fathers, 34.12 years,for a difference of 5.25 years; and the difference was thesame whether the index case was male (5.36 years) orfemale (5.11 years) (Table IV). For almost half of ourcases (46.9%), the father was age 35 years or older.These findings may support the view that point muta-tions appear to be more commonly associated with pa-ternal alleles than with maternal alleles [Moloney etal., 1996; Carlson et al., 1994; Becker et al., 1996].

Although most of the published studies comprise aCaucasian population, Apert syndrome was also ob-served in other ethnic groups. California’s multiracialpopulation is an excellent source for studies of racialdifferences in the birth prevalence of birth defects. Inour sample, most children were Caucasian (63.5%), butour analysis of birth prevalence by race showed thehighest value in Asians (22.3 per million live births).

This finding is surprising because Asians in Californiaare known to have a low birth prevalence of birth de-fects in our registry. A study of an Asian population inSingapore during 1947–1961 [Pillay, 1964] showed amuch lower value (8.0 per million). Also, a low preva-lence found in the Hispanic population (7.6 per million)was in contrast to a usually higher overall birth preva-lence of birth defects among Hispanic births in the CB-DMP registry. However, it is difficult to compare ourfindings with those from other, and particularly ear-lier, sources because those studies lack identical meth-ods of ascertainment.

We conclude that these findings, derived from a sys-tematically ascertained population and representingthe largest population-based study of Apert syndrometo date, can be considered a reliable basis for geneticcounseling and decision-making, and provide a basisfor focused research to define the cause of this syn-drome.

ACKNOWLEDGMENTS

We thank Susan Eastwood ELS(D) for editorial ad-vice and Jenny Lockwood for her excellent secretarialwork.

REFERENCES

Apert E (1906): De l’acrocephalosyndactylie. Bull Soc Med Paris 23:1310–1330.

Becker J, Schwaab R, Moller-Taube A, Schwaab U, Schmidt W, Brack-mann HH, Grimm T, Olek K, Oldenburg J (1996): Characterization ofthe factor VIII defect in 147 patients with sporadic hemophilia a: Fam-ily studies indicate a mutation type-dependent sex ratio of mutationfrequencies. Am J Hum Genet 58:657–670.

Blank CE (1960): Apert’s syndrome (a type of acrocephalosyndactyly): Ob-servations on a British series of thirty-nine cases. Ann Hum GenetLond 24:151–164.

Carlson KM, Bracamontes J, Jackson CE, Clark R, Lacroix A (1994): Par-ent-of-origin effect in multiple endocrine neoplasia type 2b. Am J HumGenet 55:1076–1082.

Cohen Jr MM (1975): An etiologic and nosologic overview of craniosynos-tosis syndromes. In Bergsma D (ed): ‘‘Malformation Syndromes.’’ NewYork: Elsevier Inc., for the National Foundation– March of Dimes, pp137–189.

Cohen Jr MM (1986): ‘‘Craniosynostosis: Diagnosis, Evaluation and Man-agement.’’ New York: Raven Press, pp. 413–590.

Cohen Jr MM, Kreiborg S (1991): Genetic and family study of the Apertsyndrome. J Craniofac Genet Dev Biol 11:7–17.

Cohen Jr MM, Kreiborg S, Lammer EJ, Cordero JF, Mastroiacovo P, Erick-son JD, Roeper P, Martinez-Frias ML (1992): Birth prevalence study ofthe Apert syndrome. Am J Med Genet 42:655–659.

Cohen Jr MM, Kreiborg S (1993a): Growth pattern in the Apert syndrome.Am J Med Genet 47:617–623.

Cohen Jr MM, Kreiborg S (1993b): Visceral anomalies in the Apert syn-drome. Am J Med Genet 45:758–760.

Croen LA, Shaw GM, Jensvold NG, Harris JA (1991): Birth defects moni-toring in California: A resource for epidemiological research. PaediatrPerinat Epidemiol 5:423–427.

Czeizel AE, Elek C, Susanszky E (1993): Birth prevalence study of theApert syndrome [Letter to the editor]. Am J Med Genet 45:392.

Erickson JD, Cohen Jr MM (1974): A study of parental age effects of theoccurrence of fresh mutations for the Apert syndrome. Ann Hum GenetLond 38:89–96.

TABLE IV. Mean Age of Parents by Sex of Proband

Age of father Age of mother Difference betweenmean age of

of father and motherSex ofProband Mean SD Mean SD

Male 35.36 5.23 30.0 5.29 5.36Female 33.11 6.89 28.0 6.41 5.11All cases 34.12 6.24 28.87 5.98 5.25

Apert Syndrome 397

Kreiborg S, Cohen Jr MM (1991): The infant Apert skull. Neurosurg C NAm 2(3):551–554.

Lewanda AF, Cohen Jr MM, Hood J, Morsey S, Walters M, Kennedy Jr JL,Jabs EW (1993): Cytogenetic survey of Apert syndrome. Reevaluationof a translocation (2;9)(p11.2;q34.2) in a patient suggests the break-points are not related to the disorder. Am J Dis Child 147:1306–1308.

Moloney DM, Slaney SF, Oldridge M, Wall SA, Sahlin P, Stenman G,Wilkie AOM (1996): Exclusive paternal origin of new mutations inApert syndrome. Nature Genet 13:48–53.

Pillay VK (1964): Acrocephalosyndactyly in Singapore. A study of five Chi-nese males. J Bone Joint Surg 46B:94–101.

Sapienza C (1996): A paternal wash in Apert syndrome. Nature Genet13:9–10.

Wheaton SW (1894): Two specimens of congenital cranial deformity ininfants associated with fusion of fingers and toes. Trans Pathol Soc,Lond 45:238–241.

Wilkie AOM, Slaney SF, Oldridge M, Poole MD, Ashworth GJ, Hockley AD,Hayward RD, David DJ, Pulleyn LJ, Rutland P, Malcolm S, WinterRM, Reardon W (1995): Apert syndrome results from localized muta-tions of FGFR2 and is allelic with Crouzon syndrome. Nature Genet9:165–172.

398 Tolarova et al.