International Journal of Pediatric Otorhinolaryngology (2007) 71, 283—290

www.elsevier.com/locate/ijporl

Histological analysis of palatopharyngeal musclefrom children with snoring and obstructive sleepapnea syndrome

Isabela Mattos De Vuono a,*, Edmar Zanoteli b,Acary Souza Bulle de Oliveira b, Reginaldo Raimundo Fujita a,Shirley Shizue Nagata Pignatari a, Gilberto Ulson Pizarro a,Marcia Lurdes de Cassia Pradelle-Hallinan c, Gustavo Antonio Moreira c

aDepartment of Otorhinolaryngology and Head and Neck Surgery, Universidade Federal deSao Paulo (UNIFESP), Sao Paulo, BrazilbDepartment of Neurology, UNIFESP, Sao Paulo, BrazilcDepartment of Sleep Medicine and Biology, UNIFESP, Sao Paulo, Brazil

Received 16 June 2006; received in revised form 19 October 2006; accepted 20 October 2006

KEYWORDSSleep disorders;Palatopharyngealmuscle;Muscle histology;Soft palate;Snoring

Summary Obstructive sleep apnea syndrome (OSAS) is an upper airway obstructionthat occurs during the sleep. One of the suggested mechanisms involved in thisprocess is a neuromuscular abnormality of the palatal muscles. Whether childrenwith OSAS develop into OSAS adults, or children and adult OSAS are two distinctdisorders occurring at different ages are questions to be answered. Here, we madethe histological analysis of palatophryngeal muscle in 34 oral-breathing children ofboth genders, aged 5—12 years old, with hypertrophic tonsils and adenoids. Accord-ing to the polysomnographic study the participants were divided into childrenwithout sleeping disorders (group I) and children with primary snoring (group II)or apnea (group III). The main histological findings were fiber size variability in 70%cases from groups II and III and in 71% from group I; perimysial connective tissueinfiltration in 48% children from groups II and III and in 71% from group I; intracy-toplasmatic mitochondrial proliferation in 63% cases from groups II and III and in 57%cases from group I. Muscle necrosis was only observed in one case, in association withsubglandular inflammation. Others findings observed in all groups included fiberswith internal architecture alteration, such as moth-eaten and lobulated fibers, type2 fiber predominance, and small areas of fiber type grouping. The presence of similarhistological findings in the palatopharyngeal muscle in children with primary snoringor apnea but also in children without sleeping disorders indicate that such changes

* Corresponding author. Tel.: +55 11 55397723.E-mail address: ivuono@uol.com.br (I.M.D. Vuono).

0165-5876/$ — see front matter # 2006 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ijporl.2006.10.019

284 I.M.D. Vuono et al.

could be a normal histological feature of this muscle rather than a neurogenic ormyopathic pathology.# 2006 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Obstructive sleep apnea syndrome (OSAS) is abreathing disorder characterized by upper airwayobstruction with an abnormal sleep pattern [1].Childhood OSAS is relatively common, with a pre-valence of 2% in children aged 2—18 years [19]. Thenatural course of OSAS in children is still to bedetermined, and whether OSAS children developinto OSAS adults, or both are two distinct breathingpathologies is not clearly known [16].

In children, palatine and pharyngeal hypertrophyhave been associated with OSAS [11,5], and most ofthe cases have been successfully treated with ade-notonsillectomy [22]. However, the surgical treat-ment of OSAS has been proved not effective inpatients in whom the only risk factor for OSAS isthe adenotonsillar hypertrophy [22,14,18]. In addi-tion, it has been shown that OSAS is most probablecaused by a combination of structural and neuro-muscular factors [16,15].

The pathophysiological mechanism for the col-lapse of upper airways during inspiration and sleepin patients with heavy snoring and OSAS is not clear.Many studies have suggested that OSAS is probablythe final stage of a progressive disease, starting withsimple snoring [9]. One of the mechanisms sug-gested to be involved in this process is the neuro-genic lesion of the palatal muscles [7,13]. As thefunctional activity of the muscles of the upper air-way is of vital importance to maintain the stabilityand patency of the upper airway during inspiration,a muscle dysfunction might be one of the causes ofOSAS [7,13,20,4,23].

A previous study has shown that the degree ofneurogenic process on palatopharyngeal muscle ofadultpatientswasassociatedwith thedegreeof sleepbreathing disorder [9]. The authors compared adultswith habitual snoring andupper airway obstruction tonormal individuals and found that the degree ofhistological abnormalities correlated significantlywith the degree of obstructive breathing. They sug-gested that a progressive local neurogenic lesioncaused by the trauma of snoring might be a potentialcontributory factor to the upper airway collapsibility[9]. The determination of the presence of neurogenicmuscular alteration in children with OSAS could indi-cate a genetic risk factor for the condition.

The aim of this work was to detect the presenceof possible myopathic or neurogenic changes

on palatopharyngeal muscle in children withOSAS.

2. Patients and methods

2.1. Patients

Thirty-four consecutive children (16 males and 18females, mean age 7 years old, range 5—12) sched-uled for tonsillectomy participated in the presentstudy. The patients presented with mouth breathingfor at least 6 months, palatine tonsils hypertrophygrades III and IV [3], and adenoid hypertrophy occu-pying at least 70% of the nasopharynx, assessed byclinical examination and nasofibroscopy. Accordingto their polysomnographic study and based on theAmerican Thoracic Society Criteria [1], the childrenwere divided into three groups: group I (normalnocturnal polysomnography), group II (habitual snor-ing) and group III (OSAS) (Table 1). None of thesubjects had clinical signs or symptomsof neuropathyor neuromuscular disorder. This study received theapproval of the institutional ethics committee.

2.2. Muscle biopsy

The muscle biopsies were obtained from the cranialpart of the palatopharyngeal muscle. Every surgerywas performed under general anesthesia, withoutthe use of local anesthetic. Each specimen wasquickly frozen in liquid nitrogen and stored in a deepfreezer at �75 8C. Cryosections (5—8 mm, �25 8C)were stained with hematoxylin—eosin (H&E), mod-ified Gomori trichome (GO), NADH-tetrazoliumreductase (NADH-tr), succinate dehydrogenase(SDH), and myofibrillar adenosine triphosphatase(ATPase) after alkaline (pH 9.6) and acid (pH 4.6and 4.3) preincubations [6].

The blind evaluation of histopathological changesof palatopharyngeal muscle was performed by twospecialists (EZ, ASBO) in agreement, and the mainhistological abnormalities included for analysiswere:fiber type variability, endomysium and perimysiumconnective tissue proliferation, nuclei positioning,muscle necrosis, endomisyum and perimysiuminflammation, atrophic rounded fibers, hypertro-phied fibers, subsarcolemal and/or intracitoplas-matic mithocondrial proliferation, mouth-eatenappearance, fiber splitting, nuclear bags and

Histological analysis of palatopharyngeal muscle 285

Table 1 Sleep respiratory findings in habitual snoring and obstructive apnea syndrome patients and in non-snoringcontrol children

Case/group Age (year) Sex/race Diagnosis AHI (event/h) Nadir SaO2 (%)

1/III 5 F/B OSAS 2.3 862/III 12 F/B OSAS 110.6 393/III 11 F/B OSAS 4.1 844/III 5 F/B OSAS 9.1 825/III 6 F/B OSAS 3.4 906/III 6 M/W OSAS 24.6 557/III 5 F/B OSAS 5.1 898/III 6 M/W OSAS 10.7 779/III 6 M/B OSAS 5.4 7610/III 7 M/B OSAS 3.2 8711/III 5 F/W OSAS 16.2 4612/III 6 F/W OSAS 3.5 8913/III 10 M/W OSAS 6.9 8914/III 6 M/W OSAS 23.4 6015/II 11 M/W HS 0.9 9316/II 9 M/W HS 0 9317/II 7 F/W HS 0 9618/II 9 F/B HS 0 9619/II 7 M/W HS 0 9520/II 7 M/B HS 0 9221/II 7 M/B HS 0.4 9422/II 9 F/B HS 0 9523/II 7 F/B HS 0 9224/II 6 F/W HS 0 9225/II 8 F/B HS 0 9526/II 7 M/B HS 0 9727/II 7 F/B HS 0.4 9328/I 6 F/W Normal 0 9029/I 6 M/B Normal 0.3 9030/I 6 M/W Normal 0 9431/I 9 M/W Normal 0 9332/I 11 F/B Normal 0.2 9433/I 5 F/W Normal 0.6 9334/I 8 M/B Normal 0 96

Y, year; F, female; M,male;W,white; B, black; PLS, polysomnography; HS, habitual snoring; OSAS, obstructive sleep apnea syndrome;Nadir SaO2, minimal arterial SaO2; AHI, apnea-hypopnea index.

ragged-red fibers, fiber type 1 or 2 predominance(>60%), type grouping (at least 12 fibers of the sametype), grouped or fascicular atrophy, and angulatedfibers [6]. Themuscle fiberswere classified into types1, 2A and 2B according to their ATPase-staining char-acteristics [6]. Four to six randomly chosen areas ineach section of every muscle sample stained withH&E and ATPase at pH 9.6, 4.6 and 4.3 were scannedwith a video camera. The number of types 1, 2A and2B fibers were counted and expressed as percentageof the total number of fibers.

Measurement of the lesser diameter of each fiberwas performed on sections stained with H&E in atleast 150 fibers of each specimen. Classification offiber types and measurement of the size were per-formed with a 20�-enlargement objective. Thevariability of fiber diameter was expressed as thecoefficient of variation (CV = standard deviation �

1000 per mean fiber diameter) [6]. Mean valuesand standard deviations were calculated for indivi-dual variables according to standard procedures. Thecomparison of the medium of the muscle fiber sizebetween the control, habitual snoring and obstruc-tive sleep apnea syndrome groups wasmadewith theKruskal—Wallis test. The level of significance to testthe null hypothesis was set at 0.05.

3. Results

3.1. General morphology

The main histopathological findings are presented inTable 2. The muscle biopsy of palatopharyngealmuscle contained a high amount of connective,fat and gland tissues. A high amount of perimysial

286 I.M.D. Vuono et al.

Table 2 Histopathological findings of the PF muscle in habitual snoring and obstructive apnea syndrome patients andin non-snoring control children

Case Fiber sizevariability

Endomysial/perimysialfibrosis

MP Type 1/type2/type 2Bfibers (%)

Type grouping Abnormalinternalarchitecture

1 + 0/+++ + 9/91/0.2 A P2 + 0/0 0 38/62/0.6 P (type 1) A3 0 0/0 0 25/75/0.0 A A4 0 0/0 + 20/80/0.3 P (type 1) P5 + 0/0 0 33/67/0.6 A A6 + 0/0 + 8/92/0.8 A P7 + 0/0 0 13/87/0.7 A A8 0 0/+ + 7/93/0.8 A P9 + 0/++ 0 21/79/0.2 A A

10 + 0/+ + 7/93/0.0 A A11 + 0/+ + 14/86/0.0 A P12 + 0/+ + 8/92/0.0 A P13 + 0/+ + 12/88/0.3 P (type 1) P14 0 0/0 0 10/90/0.0 A A15 0 0/0 0 7/93/0.1 P (type 1) A16 + 0/0 + 15/85/0.2 P (type 1) P17 0 0/0 + 37/63/0.6 P (type 1) P18 + 0/++ 0 26/74/0.7 A A19 0 0/0 + 13/87/0.0 A P20 0 0/0 + 14/86/0.0 A P21 + 0/0 + 13/87/0.0 A P22 + 0/0 + 18/82/0.0 A P23 + 0/++ 0 23/77/0.0 P (type 1) A24 ++ +/++ + 24/76/0.8 A P25 ++ ++/++ 0 30/70/0.0 P (type 1) A26 + 0/++ + 8/92/0.0 A A27 + +/++ + 13/87/0.2 A P28 + 0/+++ 0 48/52/0.0 P (type 1) A29 0 0/+ 0 23/77/0.3 A A30 + 0/0 + 26/74/0.0 A P31 0 0/0 + 19/81/0.0 A P32 + 0/++ + 18/82/0.0 A P33 + 0/++ + 30/70/0.0 A P34 + 0/++ 0 22/78/1.1 A P

+, mild; ++, moderate; +++, intense; 0, normal; A, absence; P, present; MP, mitochondrial proliferation.

connective tissue was noted in 13 children of groupsII and III combined (48%) and in 5 of group I (71%)(Fig. 1A). Endomysial connective tissue proliferationwas observed only in three cases of group II.

There was amarked variability in themuscle fibersize, even inside of a same fascicle, with the occur-rence of rounded atrophic fibers and hypertrophiedfibers. Such variability was observed in 19 cases ofgroups II and III combined (70%) and in 5 cases ofgroup I (71%). The variability coefficients found forgroups I (control), II (primary snoring) and III (OSAS)were 27.4, 22.8 and 16.7, respectively. The statisticanalysis comparing the medium of the muscle fibersize between the three groups studied was notsignificant (p = 0.148). It is important to point outthat many fascicles with normal and similar sizewere observed side by side with fascicles whose

fiber size varied greatly (Fig. 1B). The majority ofthe fibers presented a more rounded rather thanpolygonal contour.

Intracytoplasmatic irregular staining with H&Eand GO was common, and in some fibers showinga granular aspect inside the cytoplasm. In GO stain-ing, the granules appeared in red color, indicating amitochondrial proliferation, which was observed in17 cases from groups II and III combined (63%) and in4 cases from group I (57%) (Fig. 1C). Occasionally,this proliferation was observed in the subsarcolem-mic position of the fibers.

No increase of centralized located nuclei (<5%),split fibers or angulated fibers were observed. Mus-cle necrosis was observed in only one child (case 23)(Fig. 1D). Those necrotic fibers were in the peri-fascicular area, in close relation to the submucous

Histological analysis of palatopharyngeal muscle 287

Fig. 1 Coronal sections of palatopharyngeal muscle. (A) A high amount of perimysial connective tissue is seenseparating the muscle fascicles (case 22) (H&E). (B) A fiber size variability and endomysial connective tissue infiltration(case 25) (GO). (C) A granular aspect inside the cytoplasm in several muscle fibers (case 17) (GO). (D) Presence of musclefiber necrosis (case 23) (GO).

Fig. 2 Coronal sections of palatopharyngeal muscle. (A) An irregularly distributed staining for oxidative activityreaction in most of the fibers can be seen, andmany fibers present a lobular or moth-eaten appearance (case 7) (SDH). (B)Small-sized rounded fibers with a marked high level of subsarcolemal oxidative activity in contrast with a weak activity inthe central area of the fibers, which is characterized by a ringed aspect (case 31) (NADH-tr). (C) Presence of clusters offibers of the same fiber type composition (type grouping) (case 28) (ATPase pH 4.6). (D) Predominance of the type 2 fibers(case 17) (ATPase pH 4.6).

288 I.M.D. Vuono et al.

and peri-glandular inflammation area, suggesting alocal process of contiguous miosites.

3.2. Oxidative activity

A large number of fibers presented an irregularlydistributed staining for oxidative activity reaction(NADH-tr and SDH). Fibers with a lobular or moth-eaten appearance were observed in 15 cases fromgroups II and III (56%) and in 5 cases from group I(71%) (Fig. 2A). In general, this fiber aspect wasassociated with those intracytoplasmatic granulesobserved in the H&E and GO staining. In addition, anumber of small-sized rounded fibers presentedwith a marked high level of subsarcolemal oxidativeactivity in contrast with a weak central activity thatwas characterized by a ringed aspect (Fig. 2B).There were no target-like fibers.

3.3. Fiber type distribution

Clusters of fibers of the same fiber type (type group-ing) were occasionally observed in the specimensfrom eight children in groups II and III (30%) and inone specimen from child in group I (14%) (Fig. 2C).

There was a predominance of the type 2 fibers inall cases from groups II and III, with prevalence rateof 62 and 93%, respectively (Fig. 2D). In only onecase, the proportion between types 1 and 2 fiberswas considered normal. Type 2B fibers were presentin a very small proportion (less than 1%) in allgroups.

4. Discussion

OSAS is thought to be directly associated with theinability of the pharyngeal muscle to respond to thenegative pressure inside the upper airway generatedduring inspiration, which would result in the col-lapse of the upper airway structures while thepatient is sleeping. Some authors hypothesize thatthe chronic stress on the pharyngeal tissues due tothe abnormal vibration of the pharyngeal structuresduring sleep respiration could result in muscularchanges secondarily to neural lesion [9,7,12,8,10].These changes would account for an impaired abilityof the muscles supporting the upper airway tocounter-act the negative pressure and keep theairway patency, thus contributing to the occurrenceof sleep obstructive apnea. Otherwise, many studieshave demonstrated histological abnormalities softpalate muscles in adult patients with OSAS. How-ever, studies linking abnormal muscle structure andOSAS have not been carried out in children. Whetherthe OSAS in adult individuals is a condition already

present in the childhood, or is a primary muscularchange caused by a genetic factor, is a question thatshould be determined.

In our study, the histological feature most fre-quently observed in palatopharyngeal muscle wasthe increased variability in the muscle fiber size.This fiber size variability occurred in children fromgroups II and III (primary snoring and apnea), as wellas in children from group I (control group). Edstromet al. [7] studied the biopsy specimens of palato-pharyngeal muscle from OSAS adult individuals andfound an increased variability in fiber size, andatrophic and hypertrophic fibers alternated withnormal-sized fibers. In a similar study, Friberget al. [9] observed a higher variability in palatophar-yngeal muscle fiber size from apneic individuals incomparison with normal controls. In addition, Lind-man and Stal [13] have also described abnormalvariability in fiber size comparing to muscle biopsyobtained from healthy subjects autopsy. However, aprevious study by these same authors on the softpalatemuscles from postmortem healthy individualssuggested that normal sleep breathers could alsoshow increased variability in the size of the fibers ofmuscles associated with the upper airway [21].

Similar to the studies from Friberg et al. [9], ourstudy found that the palatopharyngeal muscle ofsnoring and apneic children showed roundedatrophic fibers alternating with fibers with normalsize and shape. Additionally, the main histochemicalfindings of our study were the predominance of type2 fibers and the irregularity in the internal archi-tecture of the muscle fibers, in which moth-eatenaspect was especially relevant. The histologicalsections of some cases also showed small areas fibertype grouping. Our study, however, showed histolo-gical characteristics in the control group similar tothose found in the snorers and apneics patients. Thefact that snorers, apneics and controls showed simi-lar histological features may be evidence that thesefeatures are normal aspects of the palatopharyngealmuscle. In addition, the type 2 fiber predominance issomeway related to the anatomic role which thismuscle plays. The action of the palatopharyngealmuscle gives rise to rapid and non-supported move-ments, such as the depression of the soft palate andthe traction of the palatopharyngeal arches causingthe pharynx to shorten [17].

Authors, such as Edstrom et al. [7] and Friberget al. [9] compared the anatomy of the palatophar-yngeal muscle of OSAS patients with that of the limbmuscles (deltoid, biceps and quadriceps), whichhave a homogeneous cytoplasmatic aspect. All sam-ples included in our study presented muscle fiberswith irregular cytoplasmatic aspect (granularaspect) and non-homogenous intracytoplasmatic

Histological analysis of palatopharyngeal muscle 289

oxidative activity, which produced a lobulated ormoth-eaten aspect of the fibers. In limb muscles,these characteristics are considered abnormal andindicate neurogenic and/or myopathic primarychanges [6,2]. The abnormal internal architectureof the muscle fibers, also reported by Edstrom et al.[7] and Friberg et al. [9] may be associated with anincreased of oxidative activity due to increasednumber of mitochondria. The increased level ofoxidative activity indicates increased enduranceconferred to the palatopharyngeal muscle, probablyto compensate the predominance of type 2 musclefibers, which characteristically have anaerobicmetabolism and are, thus, more susceptible to fati-gue.

Stal and Lindman [21] described that palatophar-yngeal muscle fibers presented a more rounded thanpolygonal contour, and contained large amounts ofconnective, fat and gland tissues. We also observed,in some cases, an increased amount of connectivetissue in the perimysial region. Even at endomysiallevel, there was an increased amount of connectivetissue. In these cases, the muscle fibers acquired aless polygonal and a more rounded appearance.

In same individuals, in the present study, it waspossible to observe fascicles with normal homoge-neous fiber size beside fascicles with great varia-bility in fiber size or beside fascicles homogeneouslyatrophic. Atrophic fascicles are typical of neuro-genic changes due to injury to spinal motor neurons.However, the characterization of such neurogenicchanges must include the absence of the mosaicpattern provided by the different muscle fibersgiving rise to fiber type grouping and the presenceof angulated or target-like fibers. These were notsignificantly found in any of the participants in ourstudy, although small areas with type grouping wereobserved, which have also been described even innormal limb muscles [6].

Friberg et al. [9] reported an increased percen-tage of muscle fibers in OSAS patients with theirnuclei positioned centrally. This can be occasionallyassociated with a pathological condition. We havealso observed muscle cells with central nuclei, butthis finding occurred in both patients (snorers andapneics) and controls and at a level below 5% of thetotal cell count, which could be considered a normalanatomic variation [2].

Muscle necrosis was observed only in one case.However, it seemed to be closely related to thesubmucous inflammation process the child pre-sented with. This is a noteworthy fact to emphasizethe importance of treating adequately any inflam-matory process occurring in the oropharyngealregion. An advanced stage of myositis, besides pain,can cause even muscle paralysis.

In conclusion, children with sleep breathing dis-order (snoring or obstructive apnea) and childrenwith normal polysomnography were found to bearsimilar histological findings in the palatopharyngealmuscle: high amount of perimisium and endomisiumconnective tissue, fiber size variability, abnormalintracytoplasmatic oxidative activity, and type 2fiber predominance. In contrast to what have beenreported, we have demonstrated that the histolo-gical findings above are normal morphological char-acteristics of the palatopharyngeal muscle.Therefore, at least in children, this normal anatomicvariation is probable not related to the pathophy-siological mechanisms of OSAS.

Acknowledgements

We would like to thank CNPQ (National Counsel ofTechnological and Scientific Development) thefinancial support, Marlene Silveira Rocha, RobertoDias B. Pereira and Selene Maria M. Costa for tech-nical assistance.

References

[1] American Thoracic Society, Standards and indications forcardiopulmonary sleep studies in children, Am. J. Respir.Crit. Care Med. 153 (2) (1996) 866—878.

[2] B.Q. Banker, A.G. Engel, Basic reactions of muscle, in: A.G.Engel, C.F. Armstrong (Eds.), Myology, second ed., McGraw-Hill Inc., New York, 1994, pp. 832—888.

[3] L. Brodsky, Modern assessment of tonsils and adenoid,Pediatr. Clin. North Am. 36 (6) (1989) 1551—1569.

[4] R.T. Brouillette, B.T. Thach, A neuromuscular mechanismmaintaining extrathoracic airway patency, J. Appl. Physiol.46 (4) (1979) 772—779.

[5] R.T. Brouillette, S.K. Fernbach, C.E. Hunt, Obstructive sleepapnea in infants andchildren,J.Pediatr. 100 (1) (1982)31—40.

[6] V. Dubowitz, Muscle Biopsy: A practical Approach, seconded., Bailliere Tindall, London, 1985.

[7] L. Edstrom, H. Larsson, L. Larsson, Neurogenic effects on thepalatopharyngeal muscle in patients with obstructive sleepapnea: a muscle biopsy study, J. Neurol. Neurosurg. Psy-chiatry 55 (1992) 916—920.

[8] D. Friberg, B. Gazelius, T. Hokfelt, B. Norlander, Abnormalafferent nerve edings in the soft palatal mucosa of sleepapneics, Regul. Pept. 71 (1997) 29—36.

[9] D. Friberg, T. Ansved, K. Borg, B. Carlsson-Nordlander, H.Larsson, E. Svanborg, Histological indications of a progres-sive snorers disease in a upper airway muscle, Am. J. Respir.Crit. Care Med. 157 (1998) 586—593.

[10] D. Friberg, B. Gazelius, L.E. Lindblad, B. Nordlander, Habi-tual snorers and sleep apnoics have abnormal vascularreactions of the soft palatal mucosa on afferent nervestimulation, Laryngoscope 108 (1998) 431—436.

[11] C. Guilleminaut, R. Korobkin, R. Winkle, A review of 50children with obstructive sleep apnea syndrome, Lung 159(1981) 275—287.

290 I.M.D. Vuono et al.

[12] H. Larsson, B. Carlsson-Nordlander, L.E. Lindblad, O. Nor-beck, E. Svanborg, Temperature thresholds in the orophar-ynx of patients with obstructive sleep apnea syndrome, Am.Rev. Respir. Dis. 146 (1992) 1246—1249.

[13] R. Lindman, O.S. Stal, Abnormal palatopharyngeal musclemorphology in sleep-disordered breathing, J. Neurol. Sci.195 (2002) 11—23.

[14] C. Marcus, J.L. Carrol, O. Bamford, P. Pyzik, G.M. Loughlin,Supplemental oxygen during sleep in children with sleep-disorder breathing, Am. J. Respir. Crit. Care Med. 152 (1995)1297—1301.

[15] C.L. Marcus, Sleep-disordered breathing in children, Curr.Opin. Pediatr. 12 (2000) 208—212.

[16] C.L. Marcus, Sleep-disordered breathing in children, Am. J.Crit. Care Med. 164 (2001) 16—30.

[17] K.L. Moore, A Faringe, in: K.L. Moore (Ed.), Anatomiaorientada para a clınica, third ed., Editora Guanabara Koo-gan S.A., Rio de Janeiro, 1994, pp. 747—759.

[18] S. Morton, C. Rosen, E. Larkin, P. Tishler, J. Aylor, S. Redline,Predictors of sleep-disordered breathing in children with a

history of tonsillectomy and/or adenoidectomy, Sleep 24 (7)(2001) 823—829.

[19] S. Redline, P.V. Tishler, M. Schluchter, J. Aylor, K. Clark, G.Graham, Risk factors for sleep-disordered breathing in chil-dren: associations with obesity, race, and respiratoryproblems, Am. J. Respir. Crit. Care Med. 159 (1999)1527—1532.

[20] J.E. Remmers, W. de Groot, E.K. Sauerland, A.M. Anch,Pathogenesis of upper airway occlusion during sleep, J.Appl. Physiol. 44 (1978) 931—938.

[21] O.S. Stal, R. Lindman, Characterisation of human soft palatemuscles with respect to fibre types, myosins and capillarysupply, J. Anat. 197 (2000) 275—290.

[22] J.S. Suen, J.E. Arnold, L.J. Brooks, Adenotonsillectomyfor treatment of obstructive sleep apnea in children,Arch. Otolaryngol. Head Neck Surg. 121 (5) (1995) 525—530.

[23] B.T. Woodson, J.C. Garancis, R.J. Toohill, Histopathologicchanges in snoring and obstructive sleep apnea syndrome,Laryngoscope 101 (1991) 1318—1322.