Craniofacial morphology and otitis media witheffusion in children

Renata Di Francesco a,*, Bruno Paulucci a, Claudio Nery b,Ricardo Ferreira Bento a

aOtolaryngology Department, University of Sao Paulo, School of Medicine, Sao Paula, BrazilbDepartment of Olaryngology, University of Sao Paulo, School of Medicine and Departmentof Orthodontics, Federal University of Goias, School of Dentistry, Sao Paula, Brazil

Received 8 January 2008; received in revised form 31 March 2008; accepted 31 March 2008Available online 2 June 2008

International Journal of Pediatric Otorhinolaryngology (2008) 72, 1151—1158

www.elsevier.com/locate/ijporl

KEYWORDSOtitis media;Eustachian tubedysfunction;Craniofacialabnormalities;Children growth anddevelopment

Summary Otitis media with effusion (OME) affects 28—38% of pre-school children,and it occurs due to the dysfunction of the auditory tube. Anatomical development ofthe auditory tube depends on the craniofacial growth and development. Deviations ofnormal craniofacial morphology and growth using cephalometric studies, may predictthe evolution of otitis.

Our goal in this paper is to determine if there are differences in craniofacialmorphology between children with adenoid enlargement, with and without otitismedia with effusion.

This is a prospective study in which the sample consisted of 67 children (male andfemale) from 5 to 10 years old. All patients presented chronic upper airway obstruc-tion due to tonsil and adenoid enlargement (>80% degree of obstruction). Thirty-three patients presented otitis media with effusion, formore than 3months and 34 didnot. The latter composed the control group.

Standardized lateral head radiographs were obtained for all subjects. Radiographswere taken with patient positioned by a cephalostat and stayed with mandibles incentric occlusion and lips at rest. Radiographs were digitalized and specific landmarkswere identified using a computer program Radiocef 2003, 5th edition. Measurements,angles and lines were taken of the basicranium,maxilla andmandible according to themodified Ricketts analysis. In addition, facial height and facial axis were determined.

Children with otitis media with effusion present differences in the morphology ofthe face, regarding thesemeasures: N—S (anterior cranial base length), N—ANS (upperfacial height), ANS—PNS (size of the hard palate), Po—Or.N—Pog (facial depth), Ba—N.Ptm—Gn (facial axis), Go—Me (mandibular length) and Vaia——Vaip (inferior phar-yngeal airway).# 2008 Elsevier Ireland Ltd. All rights reserved.

* Corresponding author. Tel.: +55 11 3889 0359; fax: +55 11 3889 8307.E-mail address: renatadifran@uol.com.br (R. Di Francesco).

0165-5876/$ — see front matter # 2008 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ijporl.2008.03.027

1152 R. Di Francesco et al.

1. Introduction

Otitis media with effusion (OME) affects 28—38% ofpre-school children, and it occurs due to the dys-function of the auditory tube [1—3] which have amultifactorial origin [4]. The most common ethiol-ogies of tubal dysfunction are functional problems[1] and its intrinsic system of opening and closing[5—8], immunological [9] and genetic disorders [10],gastroesophageal reflux [11] and rhinosinusitis[5,12]. Enlarged adenoid is another associated fac-tor, once it obstructs the pharyngeal ostia of theauditory tube [12,13].

Anatomical development of the auditory tubedepends on the craniofacial growth and develop-ment [14,15]. In an anterior series we describedcraniofacial differences between adults with andwithout otitis media [16]. The influence of cranio-facial morphology in syndromic children is evident,such as Down and Crouzon, etc. [17—19].

Deviations of normal craniofacial morphology andgrowth using cephalometric studies may predict theevolution of otitis.

Otitis media with effusion is a common disease inchildren with enlarged tonsil and adenoids, com-prehending about 30% of them; however, why do justsome of these children develop OME? The purpose ofthis study was to test the hypothesis that there is acorrelation between craniofacial morphology andotitis media with effusion in children with enlargedtonsils and adenoid.

2. Subjects and methods

This is a prospective study in which the sampleconsisted of 67 children (male and female) from 5to 10 years old, under treatment in the Otolaryn-gology Department of the University of Sao Paulo,School of Medicine. All patients presented chronicupper airway obstruction due to tonsil and adenoidenlargement. Children were randomically chosen inthe waiting list for tonsillectomy and adenoidect-omy (T&A surgery) with or without indication fortympanotomy for ventilarion tubes. In general, the

Table 1 Cephalometric points according to Radiocef progr

N: nasion Me: mentonOr: orbitale Pog: pogonionS: sella turcica Gn: gnathionPo: porion M: mentaleBa: basion ANS: anterior nasal spineCo: condylion PNS: posterior nasal spineAr: articulare Go: gonion

indications for tympanotomy tubes in this depart-ment followed the American Academy of Otolaryn-gology guidelines for ventilation tube placement.The recommendation for tube placement is a validindicator of Eustachian tube dysfunction. All thepatients presented at least 80% of nasopharyngealobstruction due to adenoid enlargement, deter-mined by the lateral head X-ray evaluation. Allthe children presented tonsils staged grades IIIand IV, according to the Brodsky classification [20].

Thirty-three patients presented otitis media witheffusion, for more than 3months and 34 did not. Thelatter composed the control group.

Otitis media with effusion was diagnosed by thephysical exam (otoscopy) and confirmed by theaudiogram and tympanometry. All of them pre-sented an air—bone gap of at least 20 dB, absentestapedian reflex and a flat tympanometry curve(B). These children were followed for at least 3months and there was no resolution of the otitis.These children with OME have the indication forventilation tube inserction. The waiting list forT&Awith or without ventilation tubes in our hospitalis about 8—10 months long.

Exclusion criteria were as follows: personal orfamily history of cleft palate or other craniofacialsyndromes, chronic medical conditions previousoral, pharyngeal, craniofacial or nasal surgery,and being currently or previously under orthodontictreatment.

Caregivers answered a questionnaire about thepast history of the otitis and about time of breast-feeding. All subjects underwent otolaryngologicanamnesis and physical examination.

Standardized lateral head radiographs wereobtained for all subjects. Radiographs were takenwith patient positioned by a cephalostat and stayedwith mandibles in centric occlusion and lips at rest.All radiographs were digitalized and specific land-marks were identified (Table 1), using a computerprogram Radiocef 2003, 5th edition. Measurements,angles and lines (Tables 2 and 3) were taken of thebasicranium, maxilla and mandible according to themodified Ricketts analysis. In addition, facial heightand facial axis were determined (Fig. 1).

am (Fig. 1)

Ptm: pterigomaxillaryVasa: upper anterior pharyngeal wallVasp: upper posterior pharyngeal wallVaia: lower anterior. pharyngeal wallVaip: lower posterior pharyngeal wallA: maxilar concavityB: mandibular concavity

Craniofacial morphology and otitis media with effusion in children 1153

Table 2 Linear measures used the cephalogram

N—S Anterior cranial base lengthS—Ba Middle cranial base lengthGo—Me Length of the body of mandibleN—Me Total anterior facial heightN—ANS Superior facial heightANS—PNS Size of the maxillaVasa—Aasa Superior pharyngeal airwayVaia—Vaip Inferior pharyngeal airway

Table 3 Angular measures used in the cephalogram

Po—Or.Ba—N Cranial deS—N.Go—Me RelationshANS—PNS.Po—Or RelationshPo—Or.Go—Me MandibulaBa—N.Ptm—Gn Facial axisXi—ANS.Si—Pm Inferior faDc.Xi—Xi—Pm MandibulaPo—Or.N—Pog Facial depGo—Me.N—Pog RelationshN—S.Ba RelationshPo—Or.N—A Maxillar d

Fig. 1 Cephametric points.

Regarding the abbreviations used for the mea-surements discussed in this article, a dash [—] repre-sents a line and a period [.] represents an angle.

2.1. Statistical analysis

Potential confounders of the relationship betweencraniofacial morphology and otitis media witheffusion were those suggested by the literature(breast-feeding, tonsil and adenoid enlargement.

flexionip between anterior cranial base and mandibular planeip between palatine and Frankfurt planesr plane

cial heightr archthip between facial height and mandibularip between anterior and medium cranial basesepth

1154 R. Di Francesco et al.

Table 4 Clinical conditions associated to OME

Breast-feeding 0.452Tonsil enlargement 0.964Adenoid enlargement 0.479Otitis media with effusion 0.633

Mann—Whitney test.

Table 5 Sex distribution and otitis media

Sex Control Otitis Total

FemaleNumber 13 17 30% by sex 43.3 56.7 100% with otitis 38.2 51.5 44.8% total 19.4 25.4 44.8

MaleNumber 21 16 37% by sex 56.8 43.2 100.0% with otitis 61.8 48.5 55.2% total 31.3 23.9 55.2

TotalNumber 34 33 67% by sex 50.7 49.3 100.0% with otitis 100.0 100.0 100.0% total 50.7 49.3 100.0

That correlation was analyzed by the Mann—Whit-ney test.

Descriptive statistics (mean � standard devia-tion) were calculated from the observed valuesfor each measurement. Mean differences wereassessed using Student’s t-test for independentgroups. A p value of 0.05 or less was consideredstatistically significant. Student’s t-test was alsoused for comparing mean values for the otitis groupwith normal values found in the literature.

Correlations between variables were assessedusing Pearson’s chi-square test.

The study protocol was approved by the Ethics inResearch Committee of the University of Sao Paulo,School of Medicine. And informed consent wasobtained from the legal caretaker of each patient.

Table 6 Age and otitis media

Age Otitis Total

Yes No

5—8 (years)Number 16 23 39% control 41.0 59.0 100.0% otitis 47.1 69.7 58.2% total 23.9 34.3 58.2

8—10 (years)Number 18 10 28% control 64.3 35.7 100.0% otitis 52.9 30.3 41.8

3. Results

The final sample comprises of age range from 5 to 10years: 55.2% boys and 44.8% girls. The Mann—Whit-ney test showed that there were no significantdifferences between the two samples according totonsil and adenoid enlargement and breast-feeding,as it can be seen in Table 1. Table 4 shows thedistribution according to sex.

Sex distribution is found in Table 5.Distribution by age is described in Table 6: There

is a tendency for the occurrence of otitis witheffusion in children between 5 and 8 years old.

Table 7 shows the comparison of all angular andlinear measurements

That table shows that there are 7 statisticallysignificant measurements: shorter N—S (the lengthbetween the Nasium and the Sela), shorter Go—Me(length from the Gonion to the Menton), shorter N—ANS (superior facial height), ANS—PNS (shorterlength of the maxilla), shorter Vaia—Vaip (inferiorpharyngeal airway), bigger Ba—N.Ptm—Gn (facialaxis) and bigger Po—Or.N—Po (facial depth) (Fig. 2).

% total 26.9 14.9 41.8

TotalNumber 34 33 67% control 50.7 49.3 100.0% otitis 100.0 100.0 100.0% total 50.7 49.3 100.0

Fischer’s exact test, p = 0.05.

4. Discussion

The main cause of otitis media is auditory tubedisfunction [1—3]. Other factors may contributesuch as: inflammatory processes of auditory tube,

adenoid enlargement, etc. [3,6,16]. However, cor-relation with adenoid enlargement was not found inthis series, once all the children presented it. Arecent paper [21] shows that adenotonsillectomy donot change the course of otitis with effusion.

Duration of breast-feeding is believed to be oneof the protecting factors for otitis media, and thiscorrelation was not also found in this paper. We alsoknow that breast-feeding plays an important role incraniofacial development and growth and also in thedevelopment of oro-facial muscles [22,23]. But wedid not found association between otitis and breast-

Craniofacial morphology and otitis media with effusion in children 1155

Table 7 Comparison of all angular and linear measurements

Otitis Control Significance

Media Standarddeviation

Media Standarddeviation

N—S, anterior cranial base length 64.3909 3.2566 66.7947 3.0922 0.003S—Ba, medial cranial base length 43.7340 4.3228 44.0462 3.9851 0.759Go—Me, length of the mandible body 61.1730 5.4855 64.0209 5.2330 0.033N—Me. total facial height 110.083 8.114 111.106 18.571 0.772N—ANS, superior facial height 46.387 4.506 48.960 4.915 0.029ANS—PNS, length of the maxilla 45.3355 5.6322 49.0024 3.6077 0.003Vasa—Vasp 5.8845 2.9423 6.8882 2.9540 0.168Vaia—Vaip, inferior pharyngeal airway 12.3600 4.4866 14.3474 3.2599 0.043Po—Or.Ba—N, cranial deflexion 26.7058 5.2296 27.0732 4.1678 0.751S—N.Go—Me 41.0473 4.6008 39.6547 5.5733 0.270ANS—PNS.Po—Or 176.3597 3.5632 175.7685 2.8734 0.457Po—Or.Go—Me 33.7261 6.6014 31.3624 6.0448 0.131Ba—N.Ptm—Gn, facial axis 96.2524 4.6383 93.8094 4.1887 0.027Xi—ANS.Xi—Pm 131.8521 3.9911 132.2456 3.9106 0.685Dc—Xi.Xi—Pm 154.8155 7.0201 156.9347 6.3735 0.200Po—Or.N—Pog, facial depth 96.2867 5.5287 93.7900 4.1012 0.039Go—Me.N—Pog, facial cone 116.1664 4.1619 113.1276 18.1491 0.352N—S.Ba, angle between the anterior

and posterior cranial bases48.366 3.827 48.945 5.412 0.614

Po—Or.N—A 88.9994 5.8668 90.0503 3.7787 0.389

Student’s t-test p < 0.05.

Fig. 2 The measurements that present statistical differences between the groups.

1156 R. Di Francesco et al.

feeding. Our sample is composed bymouth breathingchildren due to the adenoid and tonsil enlargementand probably their hypotonic oro-facial musclesshould be important in the genesis of this kind ofotitis [24]; however, no correlation was found.

There was a predominance of male children, andit corroborates other papers that show mouthbreathing is predominant in boys [14].

Incidence of otitis media is larger in children [25]which is explained by the different anatomy ofauditory tube in children. We found a prevalenceof otitis in children less than 8 years of age.

The main differences in the auditory tubebetween children and adults lie not only in thelength (it is shorter in children) [26] and diameter,but also in its angle with the cranial base which issmaller in this phase of life [25,27]. With craniofa-cial growth and development, the auditory tubeacquires the characteristics of the adult [14,28]exerting a mandatory influence in its function[6,7,26]. Consequently, there is a reduction of pre-valence of otitis in adolescence [6].

Thinking about that, we studied craniofacial mor-phology in children with and without otitis media.We chose for this study children with tonsil andadenoid enlargement because it is one of the mostcommon associations to otitis media with effusion,even though only a small part of them develops otitismedia. And this was our main question. Why onlysome of them? Based on the fact that children withcraniofacial anomalies are more prone to otitismedia, and that some craniofacial differences arepresent in adults with chronic otitis media [16]; wehypnotized that besides the pharyngeal obstruction,craniofacial morphology would exert some influencein the genesis of otitis media.

Computerized cephalometry is very precise in thestudy of craniofacial morphology and the cephalo-metric measures were based on a previous series[16].

We do found some differences between thegroups.

Fig. 3 Auditory tube development followi

Cranial base is the key to the development andgrowth of the face [28], and its morphology mayinfluence other related structures [29]. This is thecounter part theory. It says that the cranial basedetermines the sizes of maxilla and mandible [29].The auditory tube extends from the cranial base(bony part) to the pharynx. The inclination of thispart depends on the projection of the maxilla, as itwill be showed below.

In this series we found a smaller N—S (lengthof the anterior cranial base), in children withotitis media what corroborates other authors[14,16,19,30]. Mann et al. [31] found that the ante-rior skull base of children with otitis media is shorterthan that of other children. Our results are also inagreement with those authors that found a correla-tion between otitis media and brachycephaly[32,33].

The bony portion of the Eustachian tube islocated in the petroscamous fissure of the temporalbone, which is near the occipital and sphenoid bones[21]. Together these structures form one of the mostimportant basicranium synchondroses involved inthe growth and development of the skull.

In the maxilla, differences were found in N—ANS(maxillar height or superior facial height) and ANS—PNS (antero-posterior length of the maxilla) whichrepresents the size of maxilla. During craniofacialgrowth, the maxilla is projected antero-inferiorly[28,30] and according to Mann et al. [31], thedevelopment of this group of bones leads to thegrowth of the cartilaginous part of the auditorytube, that will acquire the adult characteristics.It influences the inserction of paratubal (tensorand levator veli palatine) muscles [6,14,29](Fig. 3). A shallower maxilla correlates with lessdownward and forward maxillary movement duringcraniofacial growth and development. The shortersuperior height is shorter in otitis media indicates anarrested development of the maxilla. A shortervertical dimension of the face may imply in a poorerEustachian tube dysfunction. Maw and Bawden [22]

ng maxilla projection (antero-inferior).

Craniofacial morphology and otitis media with effusion in children 1157

suggested that children with serous otitis mediahave experienced delayed craniofacial growth.

Shibahara and Sando [30] describe that the inser-tion angle of these muscles may alter the vectors oftraction in the opening of the tube. That explainsthe insertion of the tensor veli palatine muscle thatis more superior and more distant of the lumenadults than in children [34].

Paratubal (tensor and levator veli palatine) mus-cles extend from the cranial base to the hard palate[35] and the cartilaginous part of the auditory tube,along the pharynx wall [36]. So, the smaller projec-tion of the maxilla may explain the poor auditoryfunction in children [6,25,29]. A smaller or delayedgrowth in the vertical axis of the face may interferein tubal function and in a higher incidence of otitis.

Enlarged facial depth (Po—Or.N—Pog) shows thereis a disproportion betweenmaxilla andmandible. Anenlarged facial depth is present when the maxilla isunderdeveloped.

The facial axis (Ba—N.Ptm—Gn) represents thedirection of craniofacial growth. A bigger facial axisangle is associated to a more horizontal growth ofthe face. Many authors [32,33,37] suggest this con-dition leads to otitis media and other abnormalitiesof the tympanic membrane.

Vaia and Vaip represent the upper airway space.It was smaller in children with otitis media, and itmay be explained by the retroposition of the mand-ible, as it will be seen. Smaller proportions ofpharynx are related also to different inserctionand vectors of paratubal (tensor and levator velipalatine) muscles interfering in tubal function.

At last, the length of mandible (Go—Me) smallerin the studied group. The mandible must adjust tothe maxilla, and in these cases there is an under-development of the maxilla and probably thisexplains the smaller dimension according to thecounterpart theory [28]. A smaller mandibleexplains a smaller retrolingual space and conse-quently a smaller Vaia and Vaip.

5. Conclusion

Children with otitis media present differences inthe morphology of the face, regarding these mea-sures: N—S (anterior cranial base length) N—ANS(upper facial height), ANS—PNS (size of the hardpalate), Po—Or.N—Pog (facial depth), Ba—N.Ptm—Gn (facial axis), Go—Me (mandibular length) andVaia—Vaip (pharyngeal airway). Tubal dysfunctionin childhood can be attributed to the relativeposition of the Eustachian tube and paratubalmuscles [3,31]. According to Mann et al. [31]and Kemaloglu et al. [15], the development of

the Eustachian tube is influenced by the way inwhich the face develops. Our data strongly sup-port the idea that the Eustachian tube position isassociated to craniofacial growth and develop-ment. Deviations of this process may affect notonly the tendency toward the otitis media but alsoits prognosis [16].

References

[1] C.D. Bluestone, Q.C. Berry, Concepts on the pathogenesis ofmiddle ear effusions, Ann. Otol. Rhinol. Laryngol. 85 (1976)182—187.

[2] N.W. Todd, Otitis media and Eustachian tube calibre, ActaOtolaryngol. (Stockh.) 404 (1983) 13—15.

[3] C.D. Bluestone, Current concepts in eustachian tube func-tion related to otitis media, Auris Nasus Larynx 12 (1985)1—12.

[4] M. Straetemans, N. van Heerbeek, E. Tonnaer, K.J. Ingels,G.T. Rijkers, G.A. Zielhuis, A comprehensive model for theaetiology of otitis media with effusion, Med. Hypotheses 57(6) (2001) 784—791.

[5] S.M. Marone, M.C. Lorenzi, R.F. Bento, medio. Doencas doouvido, R.F. Bento, A. Miniti, S.M. Marone, Tratado deOtologia, EDUSP, Sao Paulo, 1998, 173—258.

[6] C.D. Bluestone, J.O. Klein, Otitis media, atelectasis andeustachian tube dysfunction, in: C.D. Blustone, S.E. Stool,M. Kenna (Eds.), 3rd ed., Pediatr. Otolaryngol., 1, Saunders,Philadelphia, 1996, pp. 388—5817.

[7] C.D. Bluestone, Eustachian tube function and managementof otitis media, in: Handout for Instructional Course Pre-sented at Annual Meeting of the American Academy ofOtolaryngology, Washington, DC, September, 2000.

[8] M.A. Kenna, Otitis media with effusion, Bailey B: Head &Neck Surgery–—Otolaryngology, 2, Lippincott Company, Phi-ladelphia, 1993, pp. 57—69.

[9] M. Wierzbicka, W. Szyfter, M. Lawniczak, B. Kaminski, Con-tribution to etiology of otitis media with effusion, Otolar-yngol. Pol. 55 (4) (2001) 405—407.

[10] M.L. Casselbrant, E.M. Mandel, The genetics of otitis media,Curr. Allergy Asthma Rep. 1 (4) (2001) 353—357.

[11] A. Tasker, P.W. Dettmar, M. Panetti, J.A. Koufman, J.P.Birchall, J.P. Pearson, Reflux of gastric juice and glue earin children, Lancet 359 (9305) (2002) 493.

[12] P.B. Dinis, A. Gomes, Immunologic mechanisms in secretoryotitis media–—recent concepts. Part II, Allerg. Immunol.(Paris) 23 (3) (1991) 81—87.

[13] H. Kubba, J.P. Pearson, J.P. Birchall, The aetiology of otitismedia with effusion: a review, Clin. Otolaryngol. 25 (3)(2000) 181—194.

[14] Y.K. Kemaloglu, N. Goksu, S. Ozbilen, N. Akyildiz, Otitismedia with effusion and craniofacial analysis. II: ‘‘Mastoid-middle ear-Eustachian tube system’’ in children with secre-tory otitis media, Int. J. Pediatr. Otorhinolaryngol. 32 (1995)69—76.

[15] Y.K. Kemaloglu, N. Goksu, A. Koybasioglu, E. Inal, S. Ozbilen,Prognostic value of craniofacial growth and development inchildren with secretory otitis media, in: M. Tos, J. Thonsen,V. Balle (Eds.), Otitis Media Today, Kyler Publications, Neth-erlands, 1999, pp. 81—91.

[16] R.C. Di Francesco, P.L. Sampaio, R.F. Bento, Correlationbetween otitis media as craniofacial morphology in adults,ENT J 86 (12) (2007) 738—743.

1158 R. Di Francesco et al.

[17] K.A. Daly, Epidemiology of otitis media, Otolaryngol. Clin.North America 24 (1991) 775—786.

[18] Y.K. Kemaloglu, T. Kobayashi, T. Nakajima, Analysis of cra-niofacial skeleton in cleft children with media with effusion,Int. J. Pediatr. Otorhinolaryngol. 47 (1999) 57—69.

[19] P.M. Brown, G.T.R. Lewis, A.J. Parker, A.R. Maw, The skullbase and nasopharinx in Down’s syndrome in relation tohearing impairment, Clin. Otolaryngol. 14 (1989) 241—246.

[20] L. Brodsky, Tonsillitis, tonsillectomy and adenoidectomy, in:B.J. Bailey (Ed.), Head and Neck Surgery–—Otolaryngology,JB Lippincott, Philadelphia, 1993, pp. 838—847.

[21] K.P. Oomen, M.M. Rovers, E.H. van den Akker, B.K. vanStaaij, A.W. Hoes, A.G. Schilder, Effect of adenotonsillect-omy on middle ear status in children, Laryngoscope 115(April (4)) (2005) 731—734.

[22] A.R. Maw, R. Bawden, Factors affecting resolution of otitismedia with effusion in children, Clin. Otolaryngol. 19 (1994)125—130.

[23] J. Magnan, Secretory otitis media, Rev. Prat. 48 (8) (1998)860—865.

[24] Junqueira, Paula Andreya de Souza, Reabilitacao miofuncio-nal da tuba auditiva em criancas com otite media secretora,Tese de mestrado, Faculdade de Medicina da Universidadede Sao Paulo, 2003.

[25] C.D. Bluestone, W.J. Doyle, Anatomy and physiology ofeustachian tube and middle ear related to otitis media, J.Allergy Clin. Immunol. 81 (1988) 997—1003.

[26] D. Sadler-Kimes, M.I. Siegel, J.S. Todhunter, Age-relatedmorphologic differences in the components of the eusta-chian tube/middle ear system, Ann. Otol. Rhinol. Laryngol.98 (1989) 854—858.

[27] I. Sando, C. Suzuki, Three-dimensional reconstruction andmeasurement study of human eustachian tube structures: ahypothesis of eustachian tube function, Ann. Otol. Rhinol.Laryngol. 107 (7) (1998) 547—554.

[28] C.M. Myer, Growth of the pediatric skull base: assessmentusing magnetic resonance imaging, Laryngoscope 105 (9 Pt 2Suppl. 75) (1995) 1—10.

[29] D.H. Enlow, M.G. Hans, Essentials of Facial Growth, Saun-ders, Philadelphia, 1996, 303p.

[30] Y. Shibahara, I. Sando, Histopathologic study of eustachiantube in cleft palate patients, Ann. Otol. Rhinol. Laryngol. 97(4 Pt 1) (1988) 403—408.

[31] W. Mann, I. Jonas, G. Munker, Growth influence on tubalfunction, Acta Otolaryngol. 87 (1979) 451—455.

[32] G. Worley, T.E. Frothingham, R.S. Stuner, Head shape andmiddle ear effusion in children, AJDC 141 (1987).

[33] J.P. Stolovitzky, N.W. Todd, Head shape and abnormalappearance of tympanic membranes, Otolaryngol. HeadNeck Surg. 102 (4) (1990) 322—325 (Links).

[34] J.D. Swarts, S.R. Rood, The morphometry and three-dimensional structure of the adult eustachian tube: impli-cations for function, Cleft Palate J. 27 (4) (1990) 374—381.

[35] N. Albin, The anatomy of Estachian tube, Acta Otolaryngol.(Stockh.) 414 (Suppl.) (1984) 34—37.

[36] S.R. Rood, W.J. Doyle, The nasopharyngeal orifice of theauditory tube: implications for tubal dynamics anatomy,Cleft Palate J. 19 (2) (1982) 119—128.

[37] N.W. Todd, Cranial anatomy and otitis media: a cadaverstudy, Am. J. Otol. 19 (5) (1998) 558—564 (Comment in: AmJ Otol. 1999; 20 (4): 556-8).

Available online at www.sciencedirect.com