OTITIS MEDIA IN GREENLAND STUDIES ON ...

OTITIS MEDIA IN GREENLAND STUDIES ON HISTORICAL, EPIDEMIOLOGICAL, MICROBIOLOGICAL, AND IMMUNOLOGICAL ASPECTS by Preben Homøe, MD, Ph.D.

International Journal of Circumpolar Health vol. 60 suppl 2, 2001

Denne afhandling er i forbindelse med de nedenfor anførte offentliggjorte afhandlinger af Det Sundhedsvidenskabelige Fakultet ved Københavns Universitet antaget til offentligt at forsvares for den medicinske doktorgrad. Københavns Universitet, den 29. maj 2001 Hans Hultborn dekan Forsvaret finder sted fredag den 9. november 2001, kl 14.00 præcis i Anneksauditorium A, Studiestræde 6, o.g., København.

This thesis is based on the folIowing publications which are referred to in the text by their Roman numerals: I. * Homøe P. Pneumatization of the temporal bones and otitis media in ancient and modern Greenlanders. Meddr Grønland, Man & Society 1997; 22: 1-42. II. Homøe P, Bretlau P. Cholesteatomas in Greenlandic Inuit: a retrospective study and follow-up of treated cases from 1976-91. Arct Med Res 1994; 53: 86-90. III. Homøe P, Christensen RB, Bretlau P. Hearing in elementary school children in Nuuk and Sisimiut, Greenland. Arct Med Res 1995; 54: 145-50. IV. Homøe P, Christensen RB, Bretlau P. Prevalence of otitis media in a survey of 591 unselected Greenlandic children. Int J Pediatr Otorhinolaryngol 1996; 36: 215-30. V. Homøe P, Prag J, Farholt S, Henrichsen J, Homsleth A, Kilian M, Jensen JS. High rate of nasopharyngeal carriage of potential pathogens among children in Greenland: results of a c1inical survey of middle-ear disease. Clin Infect Dis 1996; 23: 1081-90. VI. Homøe P, Prag J, Olsen CB, Farholt S. Nasopharyngeal bacteria found on blood agar plates from healthy children in Greenland. Int J Circumpolar Health 1998; 57: 32-39. VII. Homøe P, Christensen RB, Bretlau P. Acute otitis media and age at onset among children in Greenland. Acta Otolaryngol (Stockh) 1999; 119: 65-71. VIII. Homøe P, Christensen RB, Bretlau P. Acute otitis media and sociomedical risk factors among unselected children in Greenland. Int J Pediatr Otorhinolaryngol 1999; 49: 3752. IX. Homøe P, Madsen HO, Sandvej K, Koch A, Garred P. Lack of association between mannose-binding lectin, acute otitis media, and early Epstein-Barr virus infection among children in Greenland. Scand J Infect Dis 1999; 31: 363-66. * In a slightly different form publication I has been successfully defended on May 15th, 1996 as a Ph.D. dissertation at the Faculty of Health Sciences, University of Copenhagen with the title "Pneumatization ol the temporal bones and otitis media in ancient and modern Greenlandic Inuit".

To my wife Lotte and daughters Louise and Anne-Sophie

"When I had known Isaiah for six months, he contracted his fourth severe middle ear infection in eight weeks. When he came out of the penicillin daze, he was hard-ofhearing. After that I would sit in front of him when I read so that he could follow the movement of my lips." Peter Høeg: Miss Smilla's feeling for snow Farrar, Straus & Giroux Ine. and Harper Collins Publishers 1993

Abbreviations OM AOM rAOM SOM COM CSOM MEE STD HL URTI CI OR EBV GAS IFA MBL PCR RSV

otitis media acute otitis media recurrent acute otitis media secretory otitis media synonymous with otitis media with effusion (OME) chronic otitis media chronic suppurative otitis media middle ear effusion simple tubal dysfunction hearing loss upper respiratory tract infection confidence interval odds ratio Epstein-Barr virus Group A streptococci immunofluorescense-antibody test mannose-binding lectin polymerase chain reaction respiratory syncytial virus

Acknowledgements The studies on which this thesis is based were mainly carried out during my appointment as a research fellow in the period 1993-1995 at the Department of Otolaryngology, Head & Neck Surgery, Rigshospitalet, University of Copenhagen. I wish to express my great appreciation and thank for the support and excellent working conditions provided to me by professor Poul Bretlau, MD, DMSc., Department of Otolaryngology, Head and Neck Surgery, Rigshospitalet. I am also grateful for the invaluable help and support from chief physician Rene Birger Christensen, MD, Godthåb/Nuuk District, Primary Health Care Clinic. The laboratory examinations in this thesis could not have been performed without the technical help and assistance from my co-authors and the staff at their departments. I owe a special thank to Jørgen Prag, MD, for many fruitful and provocative discussions, and laboratory technician Charlotte Birk Olsen for skilled and reliable assistance. I also wish to thank Niels Lynnerup, MD, Ph.D., Stense Farholt, MD, Ph.D, professor Allan Hornsleth, MD, DMSc., professor Mogens Kilian, Dr. Odont., Jørgen Henrichsen, MD, DMSc., Kristian Sandvej, MD, Ph.D., and Peter Garred, MD, DMSc., who generously have shared their knowledge with me. Ester Eriksen is kindly thanked for Greenlandic translations and Pernille Homøe for language advice. The help from the many persons at the Primary Health Care Clinic and the Clinical Laboratory in Nuuk and the Primary Health Care Clinic in Sisimiut has been of tremendous importance. The willingness of the children and the parents participating in the studies was a "sine qua non" (Qujanarsuaq). Financial support has been granted by the Research Council at Rigshospitalet, the Commission for Scientific Research in Greenland, the Greenlandic Homerule Government, the Danish Hospital Foundation for Medical Research: Region of Copenhagen, The Faroe Islands and Greenland, Director Jacob Madsens & Wife Olga Madsens Foundation, Christian & Ottilia Brorsons Travel Foundation for young scientists, DAKO A/S, Greenland Air and SAS.

Preben Homøe, Køge, November 1999

TABLE OF CONTENTS

1.

INTRODUCTION AND AIMS ............................................................................11

2.

HISTORICAL ASPECTS (INCL. ARTICLE I)................................................... 12

3.

FORMER STUDIES............................................................................................. 14

4.

DEFINITIONS ................................................................................................. ...19

5.

PRESENT STUDIES ............................................................................................21 5.1. Cholesteatoma study and follow-up (article II)........................................................21 5.1.1. Aims .................................................................................................................21 5.1.2. Material & Methods ........................................................................................21 5.1.3. Results..............................................................................................................22 5.1.4. Discussion........................................................................................................22 5.1.5. Conclusion .......................................................................................................23 5.2. Screening of hearing in elementary schoo1 children (article III) ..................... ...23 5.2.1. Aims ............................................................................................................. ...23 5.2.2. Material & Methods .................................................................................... ...24 5.2.3. Results ......................................................................................................... ...24 5.2.4. Discussion.................................................................................................... ...25 5.2.5. Conclusion................................................................................................... ...25 5.3. Clinical epidemiological survey and follow-up (article IV) ............................. ...25 5.3.1. Aims ............................................................................................................. ...25 5.3.2. Subjects & Methods ..................................................................................... ...26 5.3.3. Results.......................................................................................................... ...27 5.3.4. Discussion.................................................................................................... ...28 5.3.5. Conclusion ................................... ............................................................... ...28 5.4. Microbiological survey (articles V & VI)......................................................... ...29 5.4.1. Aims ............................................................................................................. ...29 5.4.2. Subjects & Methods ..................................................................................... ...29 5.4.3. Results.......................................................................................................... ...31 5.4.4. Discussion.................................................................................................... ...31 5.45. Conclusion ........................................................................................... ...33

5.5. Risk factor survey (articles VII & VIII)........................................................... …33 5.5.1. Aims............................................................................................................. …33 5.5.2. Subjects & Methods .................................................................................... …34 5.5.3. Results ......................................................................................................... …35 5.5.4. Discussion ................ .................................................................................. …35 5.5.5. Conclusion .................................................................................................. …37 5.6. Mannose-binding lectin, AOM and Epstein-Barr virus study (article IX) ........…37 5.6.1. Aims.................. ...........................................................................................…37 5.6.2. Material & Methods..................................... :..............................................…38 5.6.3. Results ..........................................................................................................…39 5.6.4. Discussion ....................................................................................................…41 5.6.5. Conclusion ...................................................................................................…41 6.

FUTURE STUDIES ......................................................................................................... 42

7.

CONCLUDING REMARKS ............................................................................................42

8.

ENGLISH SUMMARY ................................................................................................... 43

9.

DANISH SUMMARY ......................................................................................................45

10. REFERENCES ................................................................................................................. 47

International Journal of Circumpolar Health vol. 60; suppl. 2, 2001

9

10

1. INTRODUCTION AND AIMS Otitis media (OM) is a world-wide c1inical manifestation and a frequent disease in childhood. OM consists of different but interrelated disease manifestations which are covered by the WHO diagnostic groups H65-66, H68, H71-72, and H74 (WHO 1992). These groups involve acute OM (AOM), secretory OM (SOM) synonymous with OM with effusion (OME), chronic OM (COM) without or with suppuration (CSOM), OM with sequelae, and cholesteatoma. OM is especially prevalent during the time where language and social behaviour are developed. The epidemiology of the diseases varies and although the c1inical manifestations are quite similar throughout the world, some populations are at increased risk (Bluestone 1998). One of these populations is the Inuit (Eskimos) of the Arctic region. The Inuit peoples are distributed throughout a vast territory extending from Chukotka in Siberia to Alaska, Canada, and Greenland. Almost 167,000 people identify themselves as Inuit, approximately 50,000 of which live in Greenland, and despite differences in lifestyle, socioeconomic opportunities, and health care systems, the Inuit often share the same medical problems (Bjerregaard and Young 1998). This is also the case for upper respiratory tract infections (URTI's) and OM (Kaplan et al. 1973, Baxter and Ling 1974, Pedersen and Zachau-Christiansen 1986, Homøe et al. 1996 (article IV)). The written medical history of the Inuit population is relatively sparse and brief but since the 1950's several studies on URTI's and OM have been performed (Hayman and Kester 1957, The McGrath Project 1962, Littauer and Elm 1964, Brody et al. 1965, Reed et al. 1967, Reed and Dunn 1970, Ling et al. 1969 and 1974, Maynard 1969, Beal et al. 1972 and 1981, Johonnott 1973, Kaplan et al. 1973, Baxter and Ling 1974, Manning et al. 1974, Brodovsky et al. 1976, Lupin 1976, Tower 1979 and 1982, Baxter 1982 and 1983, Pedersen and ZachauChristiansen 1986 and 1988, DiSarno and Barringer 1987, Stewart 1989, Bretlau et al. 1991, Baxter et al. 1992, Middaugh et al. 1993, Duval et al. 1994, Woods et al. 1994, Kramer and McCullough 1998). All these studies, of which the majority are from Alaska and Canada, have shown that OM is prevalent among Inuit children, and especially CSOM has been reported to have high prevalence rates (4-32%). COM without suppuration (dry perforations) and COM sequelae have also been reported with high prevalence rates in Inuit children (15-17%) (Baxter and Ling 1974, Pedersen and Zachau-Christiansen 1986). Furthermore, these diseases have also been found to affect many adults (8-18%) (Baxter and Ling 1974, Pedersen and ZachauChristiansen 1988). Modern epidemiological studies dealing specifically with OM in Greenland are few and before the present investigations only two such studies were published (Littauer and Elm 1974,

International Journal of Circumpo1ar Health vol. 60; supp1. 2, 2001

11

Pedersen and Zachau-Christiansen 1986 and 1988). A few other studies have dealt partly with OM or with specific issues of OM (Berg and Adler-Nissen 1976, Bjerregaard 1983 and 1985, Bretlau et al. 1991, Prellner et al. 1993). The most comprehensive of these studies was carried out by Pedersen and Zachau-Christiansen in the summers of 1983 and 1984 in the towns of Ilulissat (Jakobshavn) and Maniitsoq (Sukkertoppen) and in the settlement Kangamiut near Maniitsoq. The survey included 142 children and 608 adolescents and adults and the main result was a confirmation of the high prevalence of OM in Greenland especially among the children. The survey raised many questions but the dataset did not yield any risk factors or microbiological answers and did not involve immunological aspects. Furthermore, studies on anatomical, bacteriological, and immunological factors in relation to OM in Greenlanders were scarce. Among several research areas recommended by both The Fifth and Sixth International Research Conference on Recent Advances in Otitis Media in 1991 and 1995 was OM research in special populations and high risk populations, especially concerning epidemiology, risk factors, microbiology, and immunology (Klein et al. 1994 and 1998). The Greenlandic society has undergone profound changes during the last three to four decades and still continues to change. Western lifestyle becomes increasingly prevalent and the socioeconomic, sanitary, and housing conditions, as well as the availability of health care, have been enhanced and these changes are likely to have an impact on the disease pattern in Greenland (Bjerregaard and Young 1998). Consequently, at the beginning of the 1990's, a need became apparent for further research on OM in Greenlandic children and especially focusing on the folIowing items: 1. clinical epidemiology of OM 2. possible consequences of the high prevalence of OM 3. examination of the OM associated microbiological pathogens 4. exploration of potential risk factors or risk markers associated with OM, and 5. search for immunological explanatiom: of the high prevalence of OM.

2. HISTORICAL ASPECTS (INCL ARTICLE I)

The historical aspects have been dealt with in article I in which it was attempted to estimate the frequency of OM sequelae in the temporal bones of Eskimos from different time periods. The main outcomes of this study were that it was possible to estimate the risk of having had OM based upon the distribution of pneumatized cell areas in historicaI populations and that the frequency of OM sequelae based on this method has increased from before the European colonization of Greenland in 1721 until modern time (see Table 2.1) (Homøe 1997 (article I). Furthermore, it was found that the mean area size has diminished from past to present which may be associated with a higher frequency of URTI and OM. Modern epidemiological methods were not used in the description of URTI and OM in Greenland prior to the 1960's. Earlier reports were mainly case reports or approximative statements of the number of diseased patients in different communities. Thorough descriptions of these reports including his own observations have been given by the district physician Gustav Meldorf in a report from 1907 and by the district physician Alfred Bertelsen in his thesis from 12

Table 2.1

Frequency of OM sequelae in three different samples of Eskimos. The frequencies are based on classification by use of a polychotomous logistic regression model. CI denotes confidence intervals.

Sample Ancient Eskimos (before colonization) Recent Eskimos (l00-200 years old) Modern living Inuit (Eskimos)

n 127 56 34

OM frequency 6 10 8

4.7% 17.9% 23.5%

95% Cl 1.8 - 10.0 8.9 - 30.4 10.8 - 41.2

X2 = 13.19 with 2 df; p< 0.002

1940. According to these, the first written report of possible OM in Greenland most probably dates to 1752. D. Cranz (1770), a German missionary, observed a disease in the Inuit causing severe headache, chest-pain, and violent earache. The first district physician in Greenland was F. Bloch, who was sent out in 1839 to practise in southwest Greenland. Bloch reported after one year that URTl's and also earache were very common (ref in: Meldorf 1907). Ibsen, another district physician, reported from Godthåb (now Nuuk) in 1881 and 1883 about OM as a frequent complication to flue epidemics (ref in: Meldorf 1907). In 1889, district physician Kiær reported from northwest Greenland of frequent Eustachian tube disease and OM often active longer than the acute stadium (ref in: Meldorf 1907). This was confirmed by Lindemann in Julianehåb in 1891 (ref in: Meldorf 1907). Meldorf himself stated in a report from Julianehåb district in 1899 that COM with ear discharge was very common especially among the children. This was probably caused by the frequent URTI's in Greenlanders which are transmitted to the middle ear along the Eustachian tube (Meldorf 1907). Bertelsen (1940) confirmed Meldorf´s observations and c1aimed that OM was almost a constant symptom among children. Bertelsen ascribed these findings to anatomical characteristics of the Eskimos but did not suggest any specific characteristics of importance for this tendency. Since the 1950's otologists from Denmark have regularly visited Greenland and have with remarkable consistency reported high frequencies of OM. The first Danish otologist to give consultations in Greenland was V. Melton who was sent to Greenland in 1954 (Melton 1955). During a three-month period, Melton examined 929 patients, 50% were children, from South Greenland to Uummannaq in the North. Melton found that more than 16% of the examined children suffered from chronic ear discharge and more than 15% suffered from hearing loss (HL). Melton recommended that ENT specialists visited Greenland every year. This was supported by O. Malver who was sent out in 1956 (Malver 1956). Eastern Greenland was first visited by an ENT specialist in 1959. In this more remote and isolated part of Greenland, discharging ears were only present in 2.5% of the examined ears (Arnvig 1959). Later in 1963 and 1965, comparable results were found in the remote Thule (Qaanaaq) district which was also considered an isolated region (Poulsen 1962, Smidt 1982). The U.S. Airforce base was established in 1953 in the Thule district and contact with the outside world was therefore established on a regular basis. The pioneer in audiology in Greenland was the audiologist C. Røjskjær, Head of the State Hearing Center in Odense. Røjskjær performed the first audiological examinations in Greenland in 1962. He examined almost 278 referred patients along the West coast of Greenland (Røjskjær


13

and Andersen 1963). The examinations revealed a need for audiological treatment in Greenland, and this service has since then been undertaken by the State Hearing Center in Odense. However, Bertelsen had already stated that in 1937 there were 13 deaf persons in the adult West Greenlandic population which relatively was approximately twice as many as in Denmark at that time (Bertelsen 1940).

3. FORMER STUDIES

Studies of modern scientific standard conceming OM in Greenland are few, and only three of these studies have been published in peer-reviewed international journals (Littauer and Elm 1964, Berg and Adler-Nissen 1976, Bjerregaard 1983 and 1985, Pedersen and Zachau-Christiansen 1986 and 1988, Bretlau et al. 1991, Prellner et al. 1993). Audiological reports in addition to Littauer and Elm (1964) include Røjskjær (1963 and 1974), Baxter and Røjskjær (1979), and Counter and Klareskov (1990). In the winter of 1962-63, Littauer and Elm (1964) examined the otological health status and hearing ability in 372 of 374 school children in the Uummannaq district. The children living in Uummannaq town (150 children) were examined in December 1962 and January 1963, while the children living in the surrounding settlements (222 children) were examined in March and April 1963. Due to the high participation rate, the study resembled a modern community-based population survey. The children were otoscopically examined and audiometricaIly screened at 20 dB throughout the frequencies 250, 500, 1000, 2000, 4000, and 8000 Hz. Children who failed the first audiometric screening at one or more frequencies were screened once again to improve the validity of the screening procedure and to reduce the number of children with impaired hearing due to URTI or cerumen. Children with abnormal hearing after this control audiometry was subjected to a pure tone air and bone conduction hearing test without the use of a camera silenta. Additionally, all the parents were uniformly interviewed concerning former ear diseases of their children while only the parents of children with HL were asked for relevant hearing disorder parameters. The frequency of CSOM was 3.5% and the frequency of recurrent suppurative OM was 11.3%. Former OM episodes were reported in 35.3% of the children from the town but only in 22.1 % of the children from the settlements (Fisher's exact test: Х2 = 7.26; p = 0.007 (own calculation». Approximately 50% of these children had had their first OM episode befare one year of age. HL at one or both ears was found in 33.3% of the children from the town and in 14.4% of the children from the settlements (Fisher's exact test: Х2 = 17.56; p < 0.0001 (own calculation)). Only 7 children (1.9%) showed evidence of noise induced HL. However, more severe HL > 25 db as average of the frequencies 250 - 2000 Hz was present in 32 children, 18 unilaterally and 14 bilaterally. Of these, 13.3% were from the town and 5.4% from the settlements (Fisher's exact test: Х2 = 6.18; p = 0.014 (own calculation)). OM was claimed to be the cause ofthe more severe HL in 25 ofthe 32 children (78%) while therest was ofthe sensorineural type probably caused by viral infections or complications at birth. Littauer and Elm suggested that the differences between the findings in the town and the findings in the settlements were due to a better nutritional status in the settlements and that URTI's were more prevalent in the town. This was not, however, documented. The study was generally of high standard and the results therefore valuable especially in the light of the difficulties of performing studies in Greenland. 14

Unfortunately, the exact age and gender distribution and the definitions of the different OM disease entities were not described. Additiona1ly, no statistical test results were supplied, but based on the given data, this could be calculated by the author of this thesis. No information was given concerning the exact procedures for the hearing tests and of the rooms in which they were performed. Berg and Adler-Nissen (1976) examined the hypothesis that the incidence of common cold, gastroenteritis, tonsillitis, and OM was influenced by housing standards and by crowding. In Narssaq and Julianehåb ~ Qaqortoq) municipalities in southern Greenland, a total of 238 of 286 Greenland households with children under 16 years were interviewed. The households covered 1,268 persons. The study was carried out in the summers of 1969 and 1970. Disease registration was based partly on medical records from the Primary Health Care Clinics in the districts and partly on the memories of the interviewed persons. A housing quality index was estimated based on several quality variables of the house (heating, toilet conditions, water supply, number of rooms, etc.). Crowding was estimated as a day and night burden index. A statistically significant inverse relation was found between incidence of OM and housing quality index (correlation coefficient: -0.14) but not for the number of persons with OM and housing quality index. Additionally, the number of OM cases was significantly less frequent with increasing day crowding burden index (less crowding) but not for the number of persons with OM. OM was not correlated to night crowding index. Berg and Adler-Nissen concluded that common cold and OM were affected by housing quality and crowding while gastroenteritis and tonsillitis were affected to a smaller degree. The study was extensive but failed to give information on numbers of persons affected and also on the distribution according to age and gender. Likewise, disease definitions and statistical methods were not described. Index values were dubious and it was not clearly described what was meant by a low versus a high index value. Furthermore, the correlation coefficients were low and test values for significance were not given. Tests for influence by potential confounding variables (e.g. age, gender, heredity, number of siblings, etc.) were not performed. Thus, the results of this study should be taken with some caution and need to be reproduced. In 1979-80 Bjerregaard (1983 and 1985) examined the association between housing standards and social group in relation to morbidity of respiratory infections which included lower respiratory tract infections, URTI's and OM. The study was a prospective one-year communitybased study in Upernavik town concerning health care service contacts due to respiratory tract infections in 166 of 310 children aged 0-14 years. AOM accounted for 7% of new infectious disease episodes and 5% of all contacts for infectious diseases while CSOM accounted for 13% and 25%, respectively. AOM could not be related to housing conditions or social grouping but the contact rate for acute tonsillitis and AOM was higher with better housing standards and high social grouping. There was a non-significant trend for higher incidence of new CSOM episodes and a higher contact rate for CSOM with lower housing standards and with low social group. The study had the advantage of including all children in the town of Upernavik but the number of children was probably too small to show a possible effect of the measured variables. Thus, the risk of type 2 error was high. Furthermore, it was not possible to speculate about cause and effect or the temporal associations. However, the study indicated that housing and social group were not very strong predictors for AOM or CSOM. The study gave no information of age related contact rates or contact causes, and it was based on health care contacts, which may explain the higher contact rate of acute tonsillitis and AOM in the group with relatively good housing

International Journal of Circumpolar Health vol. 60; suppl. 2, 200 l

15

standards and high social level. In a prospective study Bjerregaard (1985) also examined the one-year contact rate at the primary health care clinic and the disease pattern of 836 children and adults living in Upernavik town in 1979-80. A total of 41 % of all contacts were due to acute and chronic infections and, in particular, to acute respiratory infections (13.5%) which included OM. CSOM accounted for 1.8% and AOM/SOM for 0.6% of all contacts. Bjerregaard found a seasonal peak for acute and chronic respiratory infections in July. Pedersen and Zachau-Christiansen (1986 and 1988) performed a survey of OM in children, adolescents, and adults in Greenland. The child survey (1986) was carried out in May/June 1984 and included 142 children of 155 possible participants in the age group 3-8 years. In Maniitsoq, 45 children 3 and 4 years old randomly chosen from a Population Register in addition to all children (n = 45) attending first and second school grade were called for an otological examination. Furthermore, all children (n = 52) in the nearby settlement of Kangamiut who were born between 1977 and 1980 were examined (11 were 3 or 4 years old). An interview with the parents concerning former OM episodes and sociomedical variables was performed for the 56 children who were 3 and 4 years old. Medical records from the local Primary Health Care Center and also school records were checked. The children were examined with an otomicroscope (Zeiss OPMI 99) and a tympanometer (Madsen ZS 330). Ethnically, 103 were found to be of Inuit origin, 11 of Danish origin, 22 of mixed ethnical origin, and 6 of unknown origin. The results were, however, given for the combined ethnical groups. Nine children (6%) had CSOM, 18 (13%) had sequelae of COM (not further defined), 37 (26%) had reduced middle ear pressure, and 73 (51 %) had normal middle ear status. The morbidity from middle ear disease was found to be lower in the settlement but statistics were not given (17 of 52 or 33% had OM in Kangamiut compared with 47 of 90 or 52% in Maniitsoq; Fisher's exact test: p = 0.04 (own calculation)). However, the proportion of 3 and 4 year-olds was also lower in Kangamiut and the figures were therefore not readily comparable. No differences were found between boys and girls. Former episodes of AOM were reported in 73% of the 56 interviewed children. Having an unemployed mother from a low social stratum was reported to be the most important etiological determinant for middle ear disease in these 56 children. Type of housing, birth weight, tonsillectomy rate, or penicillin therapy were not related to middle ear disease. Furthermore, it was stated that children without middle ear disease lived in the most densely populated areas, were more frequently milk drinkers (having drunk milk the day before the examination), and had more often been breastfed and for longer periods. No data and no statistics were given for these variables and the material was small (n = 56) in order to calculate risk factor estimates. Furthermore, it was not clear whether middle ear disease meant former AOM episodes or middle ear disease found at the otological examination. Thus, the risk factor part of this study was inadequate. However, the strength of the survey was the high participation rate and the description of the prevalence of infectious middle ear disease diagnosed by otomicroscopy. The survey concerning adolescents and adults (1988) included 608 persons aged 11-20 years (n = 303) and 41-50 years (n = 305), respectively. The test subjects were randomly chosen from a Population Register. There was an even distribution of males and females. An otological examination was performed including otomicroscopy. The survey was carried out in Ilulissat in the summer 1983 ( n = 293) and in the summer 1984 in Maniitsoq including the settlement Kangamiut (n = 315). In Ilulissat CSOM was found in 1 % of the 41-50-year-olds and in 9% ofthe 11-20-year-olds. The corresponding figures in Maniitsoq/Kangamiut were 3% and 7%, 16

respectively. More females than males in Maniitsoq/Kangamiut had CSOM in the young age group (10% versus 5%). Sequelae of COM including dry perforations were found in 7% of the 41-50-year-olds and in 5% of the 11-20-year-olds in Ilulissat. The corresponding figures in Maniitsoq/Kangamiut were 18% and 23%, respectively. No sex differences were observed in the two age groups. CSOM was reported to be twice as frequent in grown-ups from the settlement of Kangamiut compared with the town of Maniitsoq while there was no difference in the findings of sequelae of COM. It was estimated that 9% of the persons with CSOM or sequelae of COM would benefit from otosurgical intervention. Cholesteatoma was found in three persons and suspected in one while traumatic incus luxation was found in three persons. It was concluded from the study that there was a need for an intensive otosurgical effort in Greenland mostly involving those with chronic draining ears. With a high participation rate, the study was large, and benefitted especially from the use of an otomicrosope by an experienced otologist and otosurgeon. However, the study gave no explanation as to why there was a higher prevalence of COM sequelae in Maniitsoq and Kangamiut than in Ilulissat, and also no statistics were supplied. Exact numbers were not given but instead frequencies. Ethnically approximately 10% were Danes with a higher percentage of especially Danish males in the older age group. The results were not given in detail for the different ethnical groups. However, the ethnical differences. between the groups may represent a potential bias and may explain the higher prevalence of CSOM in females and in the younger participants. Conflicting results were found when comparing the surveys of children, adolescents, and adults. Thus, the prevalence of CSOM and COM sequelae was lower in children 3-8 years old in the settlement of Kangamiut than in the town of Maniitsoq while the opposite was found for CSOM in 11-20 and 41-50-year-olds. Bretlau et al. (1991) reported the results from two separate studies in one article. A descriptive study concerning OM in the Thule district in northwest Greenland was carried out in the summers of 1988 and 1990. Included were 203 persons out of a total population numbering 775 Polar Eskimos. Of these, 78 were children 1-13 years old while 125 were adults asking for otological assistance. No Danes participated in the study. In addition to otological examination and tympanometry, the study included a questionnaire concerning former episodes of AOM, CSOM, HL, and exposure to noise. CSOM was found in 2.5% and COM sequelae was found in 8% of the children. In adults, the corresponding frequencies were 2% and 2.5%, respectively. It was concluded that CSOM and COM sequelae were less common in the Polar Eskimos, probably due to a lower frequency of URTI's in that part of Greenland. However, no data were supplied on the frequency of URTI's in this special population. Additionally, no information was given of the selection procedure, exact numbers, or age and gender distributions. Furthermore, the results of the questionnaires were not given. The other study by Bretlau et al. (1991) dealt with the results of 221 otosurgical treatments in adult Greenlanders. Only 102 (46%) of the operated patients showed up at the follow-up examinations by the travelling specialists in several districts in Greenland during the summer of 1988. Fifty-six were females, mean age 25 years (range 5 - 62), and the follow-up period was from 1 to 5 years (mean: 29 months). The operations included 74 myringoplasties and 28 ossiculoplasties including 20 mastoidectomies for 97 central perforations, four cholesteatomas, and one ossicular chain disruption. All the operations were performed at the Department of Otolaryngology, Head & Neck Surgery, Rigshospitalet in Copenhagen. Pre- and postoperative otological examinations and mean pure tone air conduction hearing thresholds at the frequencies 500, 1000, and 2000 Hz were registered. In total, 70% of the operated ears were dry but the

International Journal of Circumpolar Health vol. 60;suppl.2, 2001

17

perforations were only healed in 39% of the patients. It was concluded that the success rate of otosurgical procedures was below that in Danes but similar to that in Canadian and Alaskan Inuit although the follow-up periods differed between these studies (Beal et al. 1972 and 1981, Brodovsky et al. 1976, Lupin 1976, Baxter 1977, Tower 1979 and 1982, McCullough 1982 and 1990). The hearing thresholds were improved from 10-40 dB in 58%. Details of the hearing thresholds before and af ter surgical intervention including preoperative air-bone gap and hearing thresholds of the non-operated ears were not given. Furthermore, the rate and gain of social hearing including the variations in thresholds at specific frequencies pre- and postoperatively were not mentioned. The main problem in this study was the low follow-up rate and the lacking possibilities of accurate audiometrical testing. In 47 adult Polar Eskimos with COM (n = 17) and without COM (n = 30) Prellner et al. (1993) examined the distributions of IgG specific pneumococcal antibodies against pneumococcal types 3, 6A and 19F and the genetic marker G2m(n) which when negative is associated with poor antibody response against polysaccharide antigens (Sarvas et al. 1989). G2m(n) was negative in 95% of the Eskimos with and without COM compared with only 25% in Scandinavians. There were no significant differences in IgG specific pneumococcal antibody levels between Polar Eskimos with or without COM, but all the median levels were lower than those in a sample of healthy Swedes. A description of the selection procedure and a timeframing for when the samples were obtained were not given. The history of otological diseases of the participants was not obtained and in particular OM episodes in childhood were not evaluated. The sample size was small and accounted only for the Polar Eskimos who settled down in the northwestern part of Greenland in the 19th century while the population inhabiting the rest of Greenland presumably migrated into Greenland in the 12th century. The results may therefore not be transferable to the population elsewhere in Greenland. Røjskjær (1963 and 1974) performed the first audiological examinations in 1962 along the West Coast of Greenland. All 278 examined persons were referred by the local district physicians. The primary purpose of this examination was to describe the need for audiological treatment and service in Greenland. Røjskjær found that a minimum of 2-3% of the Greenlandic population would benefit from audiological treatment and that a very large proportion of these people had hearing impairment caused by OM. The examination was not designed as a population study and the obtained frequencies of audiological problems are, therefore, debatable. In 1974 Røjskjær reported his immense experience with audiological problems in Greenland including the total number of audiological treatments and services, but this report was afflicted with the same problems as the first report. However, based on the regular audiological control of more than 1000 patients and the 600-700 patients who so far had been treated with hearing aids, Røjskjær demonstrated that audiological treatment was well established in Greenland. Baxter and Røjskjær (1979) studied the sources of noise in 34 Inuit hunters from the Uummannaq district in Greenland and in 59 Inuit hunters from the Baffin Zone in the Eastern Canadian Arctic region. Of these 93 hunters, six were excluded due to conductive HL caused by COM. Pure tone audiometry was performed in as silent an environment as possible. The otological examinations revealed that 19 had minimal tympanic membrane scarring while the rest had normal tympanic membranes. Eighteen of the hunters had been treated for tuberculosis with streptomycin and 23 reported former episodes of discharging ears. Forty-one complained of intermittent or constant tinnitus. All examined hunters from Greenland had moderate to severe high frequency sensorineural HL while all Canadian Inuit, except for one hunter aged 23, had

18

high frequency sensorineural HL. The causes were predominantly c1aimed to be high intensity noise from riffles, shotguns, motor boats, and especially in Canada also snowmobiles. Of other limited noise exposure sources were power-houses, ship-engines, aircrafts, helicopters, mining, and construction. Although the study was retrospective and the selection procedure unknown, it was reasonable to conc1ude that Inuit hunters are highly exposed to noise and at high risk for developing sensorineural HL. Hereditary and congenital HL were not studied. Counter and Klareskov (1990) performed a survey of the incidence of HL in a sample of 118 Polar Eskimos living in the Thule district in northwest Greenland. The participants were from 6 to 80 years old and 75 were males. Pure tone audiometry air conduction thresholds were obtained at frequencies from 250 Hz - 8000 Hz in as silent circumstances as possible. Testretest procedures were carried out in a subgroup and showed good reliability. In some participants bone conduction thresholds were obtained especially in cases with suspected conductive HL. Otoscopic evaluation revealed only nine peop1e with signs of OM. The results showed considerably (no statistical tests) poorer hearing in adult males than in adult females. A total of 77% of the males had hearing thresholds greater than 25 dB above 2000 Hz in one or both ears compared with 35% of the females. Furthermore, 72% of the males compared with 21 % of the females had hearing thresholds ≥ 50 dB above 2000 Hz in one or both ears. In males ≤ 40 years old, 46% had high frequency HL. High frequency HL was already apparent in the group of 11-20-year-olds. Rapid age related deteriorations in the high frequency hearing thresholds were found especial1y among 11-40 years old males. Thereafter a slower deterioration occurred probably due to presbyacusis. Almost similar results have been reported in the Keewatin Health Assessment Study (Moffatt et al. 1994). Thresholds exceeding 25 dB at 250 Hz and/or 500 Hz, indicating conductive HL, were found in 8% of males and 16% of females. Hereditary or congenital HL were reported not to be evident in this sample. Counter and Klareskov proposed that the young hunters should be provided with hearing protectors of high quality. Furthermore, programmes for hearing rehabilitation involving the use of hearing aids, and aural and speech reading training should be set into action. The selection. procedure in this survey was unknown but the survey inc1uded approximately 15% of the total population living in the Thule district. The different age groups were represented with small numbers of participants (range: 3-20 for males and 3-13 for females). However, it seemed without any doubt that high frequency HL in males and even in young males was a prevalent problem among the Polar Eskimos.

4. DEFINITIONS The different OM entities are difficult to separate c1inically as well as temporally, and overlapping therefore occur. Some authors prefer to study the total OM disease entity, but due to e.g. differences in age specific prevalence rates of the specific OM entities this strategy may be problematical in several ways. Other authors prefer to study the specific disease entities such as AOM or SOM, but due to diagnostic inaccuracy in transitional cases and different disease manifestations in an individual's two middle ears this is also problematical. The present studies deal with OM specific disease entities according to the definitions mentioned below and according to a selfdetermined scale of severity. AOM: Along with contemporary c1inical symptoms as earache, fever, or irritability of less


19

than 48 hours duration, the AOM diagnosis relied on findings of a red/grey/yellowish or bulging tympanic membrane with absence of specific normal landmarks or presence of purulent otorrhea in the middle ear or the external ear canal as seen in an otoscope (inc1uding pneumatic otoscopy) in a child without a chronic perforation of the tympanic membrane prior to the AOM episode. Historical AOM: Episodes with earache, otorrhea, or treatment for AOM as reported by the parents or guardians. This was cross-checked in medical records by written evidence from district physicians of otoscopic findings of otorrhea or a red or bulging tympanic membrane or simply AOM as diagnosis which resulted in treatment with weak analgetics and/or antibiotics. This procedure was considered to resemble the c1inical situation of AOM most accurately. Historical rAOM: ≥ 5 AOM episodes since birth. SOM, middle ear effusion (MEE), and tubal dysfunction: SOM was defined as an abnormal tympanometric profile with Cl (-100 to -199 daPa), C2 (-200 to -300 daPa), or B-curve according to Jerger (1970). MEE was defined as a tympanometric B-curve with a relative gradient ≤ 0.1 or Cl or C2 curve with absence of ipsilateral stapedial reflex according to Fiellau-Nikolaisen (1983). This c1assification was used due to the high sensitivity (96%) and specificity (98%) reported by Fiellau-Nikolajsen (1983). The relative gradient was calculated as the ratio between the distance from a 100 daPa wide horizontal line intersecting the two legs of the tympanogram to the top of it divided by the maximal height of the tympanogram. Practically, this meant a compliance maximum below 0.25 ml which was used as criteria by Pedersen and ZachauChristiansen (1986). Children with Cl or C2 curves and presence of ipsilateral stapedial reflexes were grouped as simple tubal dysfunction. COM and CSOM: COM was defined as a permanent tympanic membrane perforation and CSOM was defined as COM with findings of serous or purulent otorrhea in the middle ear or external ear canal as seen by otoscopy. The chronicity was confirmed in the medical records available at the District Health Care Clinics. OM Sequelae: Sequelae in the tympanic membrane was defined as scarring, tympanosc1erosis, or atrophy as seen by otoscopy. Cholesteatoma: Presence of stratified squamous epithelium in the middle ear cavity as found under operation by use of an otomicrosope. This was only recorded in study II. Community-based: Population-based in selected small and isolated communities. Relative risk (RR): Rate of exposed AOM children divided by rate of unexposed AOM children. Odds ratio (OR): Ratio between odds of exposure in AOM children divided by odds of exposure in children without AOM. Each child was appointed to only one c1inical category according to the worst possible finding in the two ears, i.e., a child with COM in one ear and sequelae in the other was c1assified as having unilateral COM. The scale from worst to best was chosen according to the following sequence: COM inc1uding CSOM, AOM, MEE, OM sequelae, and simple tubal dysfunction (STD).

20

5. PRESENT STUDIES 5.1. Cholesteatoma study and follow-up (article II) 5.1.1. Aims Cholesteatomas in Greenlanders were already mentioned by Malver in 1956. While Ratnesar (1976) and others claimed that cholesteatomas did not exist in the Inuit, Pedersen and ZachauChristiansen (1986) found three people with otomicroscopically diagnosed cholesteatomas. This provoked a debate concerning the incidence and treatment results of cholesteatomas among Greenlanders. Therefore, the aim of the present study was to describe these factors in isolated Greenlandic regions where ENT service is infrequent. 5.1.2. Material & Methods Cholesteatoma is a rare disease and hence the study was retrospective looking at patients treated at the Department of Otolaryngology, Head & Neck Surgery, Rigshospitalet in Copenhagen which has served as referral hospital for Greenlanders with complicated otolaryngological diseases since the 1950's. The data sources were the patient record archive files at the department. Patients have been registered according to diagnosis and operative treatment if any such was performed. Until 1979, registration books were used, but since 1980 registrations have been computerized. In the 1976 files, an 11 year-old Greenlandic boy was registered with an operatively and histologically confirmed diagnosis of cholesteatoma in the middle ear. It was stated in the medical record that this was the first case of cholesteatoma in a Greenlander to be treated at the department. The study was therefore decided to include the period 1976 to 1991. From 1976 to 1979, all medical records of patients with middle ear operations for cholesteatomas were examined. Those with any kind of Greenlandic relations were checked for Greenlandic birthplace and included if so. From 1980 all medicaI records from patients born in Greenland and registered with the diagnosis cholesteatoma and/or the operation "resectio processus mastoidei" were examined and included if the operation had revealed cholesteatoma. Ethnic status as Greenlander relied upon birthplace in Greenland and the ability to speak Greenlandic. The latter was checked in the referral papers from Greenland and by telephone contact with those who lived in Denmark at the follow-up time. Variables in this part of the study contained: gender, age at time of operation, clinical symptoms before operation including audiological status, type and extension of the cholesteatomas, pathological findings, complications, and type of operation. A follow-up study was carried out in 1992 by the ENT specialists who visited Greenland. Otomicroscopy was not used. The variables encountered in the follow-up study were: state of the ear (dry or wet), intact or perforated tympanic membrane, signs of residual or recurrent cholesteatoma, and complications. Postoperative audiological status was not evaluated because accurate thresholds could not be obtained in Greenland.


21

5.1.3. Results During the 16-year study period, 35 Greenlanders were treated for cholesteatoma at the department. One patient was treated conservatively without operation. Three had bilateral disease. The male/female ratio was 21/14 and median age at operation was 19 years, range 5-67 years. Children younger than 15 years of age accounted for 37% of the patients. The total incidence of hospital treated cholesteatoma patients was 5 per 100,000 Greenlanders. Age specific incidence rates were 6.6 per 100,000 Greenlandic children aged 0-14 years and 4.4 per 100,000 adults 15 years or older. The symptoms were in all cases ear discharge and HL. Four (11%) had complications with either fistula of the horizontal semicircular canal or history of facial palsy. The cholesteatomas were generally extensive with affection of the stapes in 32% and extension to the antrum or mastoid in 71%. Attic cholesteatomas were found in 32%, sinus cholesteatomas in 24%, tensa cholesteatomas in 42% while one showed a spontaneous open cavity. Follow-up could be performed in 30 of the 35 patients (80%). Median observation time was 6 years, range 1-13 years. Perforation of the tympanic membrane was found in 47% and all of these reported intermittent episodes of ear discharge. Five patients had signs of residual or recurrent cholesteatoma, three of which were confirmed later at operation. Additional complications were not found. During the study period, 16 patients (46%) had had residual or recurrent cholesteatoma. Median time between the first and second operation was 23 months, range 4-49 months. Median age of the re-operated patients was 14 years, range 7-44 years when using age at first operation. Mastoidectomy was performed in 28 of the 37 primary operations. Canal walldown procedure was used in 11 of these with two recurrences (18%) and canal wallup in 17 with nine recurrences (53%) (p = 0.15).

5.1.4. Discussion This study represents the first systematic attempt to evaluate the epidemiology of cholesteatomas in Inuit populations. Beal and co-workers (1972) and Tschopp (1977) reported results from the same study programme of the native population in Alaska, also including Indians, and found that 3.4% (144 of 4,193) of the children had cholesteatomas confirmed by operation. These cases were not, however, further described. Ratnesar (1976) reported COM in the populations living in Labrador and northern Newfoundland. The study comprised 150 selected Eskimos, Indians, and Caucasians referred due to history of deafness or discharging ears. No case of cholesteatoma was found in the Eskimos. Ratnesar claimed that the reason for this was that the Eskimos in the study (n = 30) had larger Eustachian tubes as assessed by ureteric catheterization during operation. The present study revealed incidence rates in children and adults that were consistent with comparable studies from U.S.A., Denmark, and Finland (Harker 1977, Pedersen 1983, Karma et al. 1989). The incidence rates in children and in adults were respectively higher and lower compared to the findings by Tos (1988) who reported incidence rates of 2.9 per 100,000 in children and 12.6 per 100,000 adults in Copenhagen county. These calculations were not, however, adjusted for age specific population size. Consequently, the incidences reported by Tos are actually higher for both children and adults. The high proportion of extensive cholesteatomas in Greenlanders seemed to be the most striking difference compared with other studies (Tos 1988, Rosenfeld et al. 1992). This may be one of the reasons for the poor otosurgical success rate in Greenlanders. 22

Furthermore, the far from optimal conditions for postoperative control in Greenland may partly explain the high frequency of residuals or recurrences. Surprisingly, the complication rate was only half of the complication rate reported in Malaysian patients with cholesteatomas (Lee 1991). This study may be criticized in terms of selection of the referred patients, unknown ethnical admixture, small sample size, possible registration faults of patients treated for cholesteatoma at the department, the possibility that Greenlanders may have been treated in ENT-departments elsewhere in Denmark, and the lack of use of an otomicroscope in the follow-up study in Greenland. However, cholesteatoma is much too rare a disease to be examined for in a classical survey and comparable studies have also used hospital samples (Harker 1977, Pedersen 1983, Tos 1988, Karma et al. 1989, Lee 1991, Rosenfeld et al. 1992). Furthermore, only experienced otologists were consultants in Greenland which enhanced the diagnostic awareness and accuracy despite the lack of an otomicroscope. Substantial changes in the present incidence figures would require that a large number of patients with cholesteatomas have been ignored during the 16 year study period. Thus, one ignored patient per year, would only change the incidence to 7 per 100,000. 5.1.5. Conclusion The incidence of cholesteatoma in Greenlanders was approximately the same as that in the Western world for children but probably less in adults. The age distribution and clinical characteristics were also analogous to findings in the Western world and childhood cholesteatomas were most aggressive. However, the results of the operations were below expectations, and residual or recurrent cholesteatomas were frequent. The best results concerning residuals or recurrences were obtained by use of extensive otosurgical procedure. Regular follow-up of patients with cholesteatomas for at least five years, if not lifelong, is adviced.

5.2. Screening of hearing in elementary school children (article III) 5.2.1. Aims One of the most important complications of OM is HL. The study by Litttauer and Elm (1964), which has thoroughly been described in chapter 3, showed that HL was frequent (20%) among school chiIdren in the Uummannaq district. This has also been demonstrated in studies from Canada and Alaska (Brody et al. 1965, Reed et al. 1967, Kaplan et al. 1973, Ling et al. 1974, DiSarno and Barringer 1987). Kaplan and co-workers (1973) showed that mean verbal skills were significantly poorer in chiIdren with many and early episodes of otorrhea and with a HL defined as > 25 dB or a conductive HL with a minimum of 15 dB air-bone gap compared to chiIdren with no episodes of OM and normal hearing. Furthermore, there was a trend for lower performance score and achievement tests in reading, mathematics, and language development. DiSarno and Barringer (1987) confirmed this in an almost similar study involving 45 ninthgrade students. They found lower mean grade levels in reading, mathematics, spelling, and general information in the children with early history of OM, tympanoplasty, and conductive HL than in a comparable control group without history of OM and with normal hearing. Thus, it


23

was of great interest to examine the hearing ability of school children in the two towns of Nuuk and Sisimiut where the epidemiological studies (articles IV, VII, and VIII) also were carried out. An additional purpose was to evaluate the need for audiometric screening of school children in Greenland. 5.2.2. Material & Methods It was possible to evaluate retrospectively 167 audiograms and health care registration charts obtained in school children one year prior to the epidemiological studies. The schools and school classes (grades 1, 4, 8 and 9) had been selected in accordance to the Chief Medical Officer's recommendations in Greenland (1976). The children were between 5 and 14 years old, 89 were from Nuuk and 78 were from Sisimiut. Ethnic status was not known, but in Nuuk two classes were Danish speaking. Pure tone air conduction audiometry was obtained in as silent rooms as possible at the schools. An audiometer (Madsen TBN 85, ISO 389: 1991 (E)) was used for the screenings which started at 20 dB at all frequencies from 250-8000 Hz with the usual frequency doubling intervals. HL was defmed as failure to hear at pure tone thresholds = 20 dB. Prior to the examination the parents had filled out a questionnaire on past and present health . status of their child including history of OM.

5.2.3. Results In total, 43% of the children had hearing thresholds exceeding 20 dB at one or more frequencies in one or both ears. At the mid-frequencies (500-2000 Hz) 19% had HL > 20 dB, 10% at two or all of the mid-frequencies, and 4% were bilaterally affected. HL was significantly more frequent among the children in Sisimiut both in total (Fisher's exact test: p < 0.002 with 95% CI = 11-40%) and in comparable grade s (X2-test: 4.57, p = 0.03 for grade l and X2-test: 5.39, p = 0.02 for grade 8 to 9). Furthermore, 30% reported OM episodes in Sisimiut while 10% did so in Nuuk and none of these were from the Danish speaking classes. Report of OM was associated with impaired hearing (24/31 compared to 48/135, p < 0.0001 (Fisher's exact test)). Surprisingly, HL was almost equally frequent in the different grades. 5.2.4. Discussion The results obtained in this study were in accordance with those obtained in similar studies in Canada and Alaska of school children (see Table 5.2.4.1) (Brody et al. 1965, Reed et al. 1967, Kaplan et al. 1973, Ling et al. 1974, Woods et al. 1994). The studies were carried out almost similarly without the optimal facilities for audiometric testing such as a camera silenta, bone conduction, etc.. This implies that some of the children with especially a mild low frequency HL may have normal hearing but also that comparison with studies from the Western world are problematical. However, the prevalence of HL in school children living in the Arctic are

24

Table 5.2.4.1 Prevalence of pure tone air conduction hearing loss (HL) > 20 dB in mid-frequencies (500-2000 Hz) in different studies of Inuit school children. Author

n

Age in years

Frequency of HL

Brody et al. 1965

327

school children

39%

Reed et al. 1967 Kaplan et al. 1973 Ling et al. 1974 Homøe et al. 1995 (III)

378 361 194 167

3-5 7 - 10 4 - 14 5 - 14

31% 16% 13% 19 %

reported to be consistently high (see Table 5.2.4.1). All studies c1aim that OM is associated with the high prevalence of HL in Inuit children and this is also indicated in the present survey. The same association was found in a Danish survey of children at school start (Madsen et al. 1991). In Greenland, noise traumas due to rifle shooting and noise from motor-boats may also be involved (Baxter and Røjskjær 1979). The prevalence of hereditary or congenital HL is unknown. 5.2.5. Conclusion HL was prevalent in Greenlandic schoo1 children and the study uniformly indicated that hearing screening programmes are of great importance in Greenland in order to identify children with HL and to ensure their control and rehabilitation. Furthermore, it was indicated that OM has an impact on the high prevalence of HL in Greenlandic school children.

5.3. Clinical epidemiological survey and follow-up (article IV) 5.3.1. Aims The epidemiology of OM among the Inuit in the Arctic region has mainly been studied in surveys, the designs being either cross-sectional or attempting to inc1ude whole populations in selected towns and settlements (Hayman and Kester 1957, The McGrath Project 1962, Littauer and Elm 1964, Brody et al. 1965, Ling et al. 1969 and 1974, Maynard 1969, Beal et al. 1972 and 1981, Johonnott 1973, Baxter and Ling 1974, Manning et al. 1974, Tower 1979, Baxter 1982 and 1983, Pedersen and Zachau-Christiansen 1986 and 1988, Bretlau et al. 1991, Baxter et al. 1992, Middaugh et al. 1993, Woods et al. 1994, Kramer and McCullough 1998). Some of these surveys also contain follow-up studies. Stewart (1989) performed a historical cohort study while the only prospective cohort study was initiated in 1960 by the Arctic Health Research Center and was conducted in 27 Eskimo villages in southwestern Alaska. This study inc1uded a total of 643 live born children (Reed et al. 1967, Reed and Dunn 1970, Kaplan et al. 1973). Due to the vast geographical distances, the obvious logistic obstac1es, the lifestyle of the people, the climatic conditions, and the heavy economic costs, all the studies have to some extent experienced methodological problems especially concerning selection of the study International Journal of Circumpolar Health vol. 60; suppl. 2, 2001

25

populations, ethnicity, seasonal variation, diagnostic accuracy, and diagnostic tools. However, the studies have consistently reported high prevalences of the different OM disease entities. Baxter and coworkers reported in 1992 a dec1ine in the prevalence of COM/CSOM among elementary school children in the Eastern Canadian Arctic. This observation was based on longitudinal surveys since 1965 in three settlements (see Table 5.3.1.1). Furthermore, it was reported that 93% of ears tested had adequate hearing. Thus, Baxter and co-workers claimed to be able to see "the light at the end of the tunnel associated with the high prevalence of COM in children in the Eastern Canadian Arctic.". It was speculated that this change may have been due to adaptation to contact, improvement of medical and social services, housing facilities, and diet. The present study aimed, therefore, to examine whether the same change had occurred in Greenland and, if this was not the case, to provide epidemiological data inc1uding follow-up data which could be supportive in order to prevent or diminish the problem of OM in Greenlandic children.

5.3.2. Subjects & Methods Due to practical and economical circumstances the study was planned as a randomly sampled cross-sectional survey of children with partly follow-up and drop-out examination. It was attempted to inc1ude different study sites representing populations from the capital of Nuuk (Godthåb; 64°10’N, 51°46’W) with a more or less European lifestyle, Sisimiut (Holsteinsborg; 66°56'N, 53°40W), a large town with a more traditional lifestyle, and Qeqertarsuaq (Godhavn; 69°15'N, 53°38'W), a large settlement or small town with a typical traditional lifestyle. These sites were chosen according to population size and practical and logistic reasons. However, the study was cancelled in Qeqertarsuaq due to local staff related problems. In order to be able to compare with the results obtained by Pedersen and Zachau-Christiansen (1986), it was decided to inc1ude 3, 4, 5, and 8-year-old children. Furthermore, it was expected that by choosing these age groups a positive effect would be a higher participation rate and better diagnostic accuracy than if choosing younger age groups. The sample size was estimated prior to the survey in order to obtain robust prevalence data. SOM was expected to be the most common OM disease entity and was used as baseline for the calculations of number of participants. Thus, an expected disease prevalence of 20% with 95% CI between 16%-24%, which was considered acceptable, would demand a minimum of 248 children. With an estimated overall participation rate of approximately 75%, 330 children had to be inc1uded at each study site. The study periods were chosen to be late autumn (October - November) and late winter (March). The children in each age group were chosen from the Danish Population Register according to day of birth in the months for the first 110 children in Nuuk and for the first 75 children in Sisimiut. All children were born in Greenland. A total of 740 children were inc1uded, 440 in Nuuk, and 300 in Sisimiut, representing 18% of the children in the chosen age groups in Greenland. Contact was made by letter approximately two weeks before the examination, which took place at the Primary District Health Care Clinic in each town. One reminder was sent out. The overall participation rate was higher in Sisimiut (89%) than in Nuuk (74%). In Nuuk 325 children participated, 168 girls and 157 boys, while 266 children participated in Sisimiut, 128 girls and 138 boys. The age and sex distribution of the participants and the drop-outs did not infer any significant selection 26

Table 5.3.1.1 Prevalences of chronic otitis media (COM) and chronic suppurative otitis media (CSOM) obtained in three longitudinal surveys of elementary school children in three settlements in the Eastern Canadian Arctic region (Baxter et al. 1992). Cape Dorset Year COM / CSOM in %

Year

Iqaluit COM / CSOM in %

Year

Pangnirtung COM / CSOM in %

1968

30.0

1965

50.0

1972

5.6

1979 1990

28.2 15.4

1975 1990

35.7 4.0

1979 1990

11.4 3.1

bias. All children underwent otoscopy, pneumatic otoscopy in case of problems by simple otoscopy, and tympanometry with a Madsen ZS 330 tympanometer except if COM was present. The tympanometer was the same type as that used by Pedersen and Zachau-Christiansen (1986). The definitions of the different OM disease entities have been described in detail in chapter 4. Oneyear follow-up analysis was carried out in Sisimiut and inc1uded the 144 children who had some kind of middle ear pathology at the primary survey. af these children 82 (57%) attended. Thirtytwo children had changed addresses and could therefore not be reached. The procedure followed the above described methods. 5.3.3. Results The point prevalence rates of the different OM disease entities found in the survey are shown in Table 5.3.3.1. COM inc1uding CSOM combined for the two towns was equally frequent in the age groups, except for the 4-year-olds, and with an equal sex distribution. In Nuuk there was, however, more girls (n = 16) than boys (n = 6) with COM. The usual age specific prevalence differences were found for the different abnormal tympanometric profiles with the youngest children having the highest prevalences. Furthermore, the survey c1early showed that myringotomy, ventilation tubes, adenoidectomy, and tonsillectomy were relatively rare treatment procedures in Greenlandic children compared to the high frequencies of these procedures in Western countries inc1uding Denmark (Madsen et al. 1991). The folIow-up study in Sisimiut showed that 29.3% (24/82) of the children had recovered complete1y from middle ear pathological conditions. In three of the COM children, the tympanic membrane perforations had healed, in one child even bilaterally. However, these children still suffered from unilateral COM or MEE. Improvement was found in 58.2% (32/55) of the children with MEE, STD, or sequelae. Deterioration was only found in 6.2% (5/82) of the children, three of whom had deve1oped COM from MEE or COM from sequelae.

International Journal of Circumpolar Health vol. 60; suppI. 2, 2001

27

Table 5.3.3.1 Point prevalences, shown as frequencies, of acute otitis media (AOM), chronic otitis media (COM), chronic suppurative otitis media (CSOM), middle ear effusion (MEE), simple tubal dysfunction (STD) and sequelae among 325 children from Nuuk and 266 children from Sisimiut. Fisher's exact test was used to test for difference. Clinical group

Nuuk

Sisimiut

p-value

AOM

1.5%

0.4%

ns

COM CSOM MEE STD

5.8% 0.9% 23.0% 13.4%

7.9% 3.8% 28.2% 8.4%

ns 0.039 ns ns

Sequelae

10.8%

10.9%

ns

Total

51.7%

54.1%

ns

5.3.4. Discussion The study confirmed the high prevalence rate s of especially SOM and COM including CSOM compared with prevalence studies from Denmark and Sweden (Lous and Fiellau-Nikolajsen 1981, lngvarsson et al. 1982, Fiellau-Nikolajsen 1983, Tos 1983, Birch et al. 1984). The present survey did not use otomicroscopy, and took place in two relatively short and different time periods at two different localities. Thus, the prevalence figures may be fluctuating according to season and between the districts in Greenland. However, the prevalence figures resembled the figures obtained by use of otomicroscopy in Maniitsoq in May/June 1984 (Pedersen and Zachau-Christiansen 1986). Furthermore, the findings corresponded with two late reports from the Keewatin district in Canada (Woods et al. 1994, Kramer and McCullough 1998).

5.3.5. Conclusion This epidemiological survey involving a large sample size confirmed the findings by Pedersen and Zachau-Christiansen (1986). It could therefore be concluded that almost no changes had occurred during the 10 year period between the studies despite the improvement in socioeconomic welfare in Greenland since the 1960 - 70's. The survey underlined the need for increased awareness, control, and probably more active treatment procedures of children with SOM and COM in the two surveyed towns. Furthermore, tympanometry seems to be an important diagnostic tool.

28

5.4. Microbiological survey (articles V & VI) 5.4.1. Aims No systematic reports existed concerning the microbiological flora in relation to episodes of AOM in Inuit children prior to the present study. Furthermore, the nasopharyngeal microbiological flora in healthy children had not been investigated. Reinhard et al. (1970) studied the bacteriology in an Alaskan Inuit population which predominantly included COM patients, and therefore the bacteria found were multiple species of Gram~negative rods. Reed and Dunn (1970) mentioned unpublished bacteriological cultures which indicated that the most prevalent bacteria from AOM cases were Diplococcus, Haemophilus, and Streptococcus. Sampling of the nasopharynx has been claimed to be a valid alternative to middle ear specimens when attempting to determine the potentially pathogenic bacteria in AOM (Box et al. 1961, Kamme et al. 1971, Branefors-Helimder et al. 1975, Schwartz et al. 1979, Long et al. 1993, Faden et al. 1991). In the Western world, many studies have been performed concerning the nasopharyngeal carriage rates of potentially pathogenic bacteria in children with and without AOM or SOM (Box et al. 1961, Kamme et al. 1971, Branefors-Helander et al. 1975, Schwartz et al. 1979, Henderson et al. 1982, Ingvarsson et al. 1982a, Long et al. 1983, Freijd et al. 1984, Prellner et al. 1984, Sørensen et al. 1988, Faden et al. 1991, Aniansson et al. 1992). In most of these studies the nasopharyngeal colonization with potentially pathogenic bacteria have been shown to be associated with AOM, rAOM, and SOM while the quantitative aspects of colonization have shown more contradictory results maybe due to estimation difficulties. The most frequent bacteria to be cultured from middle ear aspirations in children with AOM are Streptococcus pneumoniae (20-50%), Haemophilus influenzae (6-30%), Moraxella catarrhalis (0-10%), and group A streptococci (GAS) (0-5%) while Staphylococcus aureus and Chlamydia trachomatis have been found irregularly (Kamme et al. 1971, Branefors-Helander et al. 1975, Schwartz et al. 1979, Chang et al. 1982, Henderson et al. .1982, Karma et al. 1983, Long et al. 1983, Giebink 1989, Arola et al. 1990). Viruses have been found in coexistence with potentially pathogenic bacteria in 10 - 30% of middle ear fluids and in the nasopharynx of 30 55% of children with AOM (Henderson et al. 1982, Arola et al. 1990, Ruuskanen et al. 1991). Especially influenza A and B virus and respiratory syncytial virus (RSV) seem to be independent pathogens inAOM (Henderson et al. 1982, Klein et al. 1982, Arola et al. 1990, Ruuskanen et al. 1991, Buchman et al. 1995). The objectives of the present studies were therefore to determine the bacterial and viral microflora in the nasopharynx of healthy Greenlandic children and in children with and without AOM including the pathogens in ear discharge.

5.4.2. Subjects & Methods Fifty-four children with AOM and 201 children without AOM (controis) were included in this survey which was carried out along with the epidemiological survey (Homøe et al. 1996, article IV). The AOM children were consecutively enrolled as they contacted the Distriet Primary Health Care Clinic and the inclusion criterion was: AOM in 0-1 O-year-old children with less than 48 hours duration of clinical symptoms. Ear discharge was present in 24 of the 54 children with AOM. Thirty were girls and 24 boys. Forty were from Nuuk (median age = 23 months,


29

range = 4-112 months) and 14 were from Sisimiut (median age = 18 months, range 4-72 months). The 201 control children (89 girls and 112 boys) were part of the subjects described in article IV except for children ≤ one-year-old, who were picked from the routine vaccination programme. The controls were examined by otoscopy and those aged three or more were also examined for SOM by tympanometry. Former episodes of AOM were registered in all children by use of medicai records and parental information as decribed in articles VII and VIII. Furthermore, use of antibiotics two weeks prior to enrolment and foreign traveis during the last six months were registered. On the basis of these registrations 70 children were classified as very healthy being with no signs and symptoms of acute or chronic disease, with no evidence of AOM or SOM, with no history of more than one episode of AOM, with no use of antibiotics for the last two weeks, and with no traveis outside Greenland during the last six months. Two nasopharyngeal swabs were obtained via the oral route. One 5% horse blood-agar and one chocolate agar culture plate was instantly inoculated with material from one of the swabs. The plates were incubated aerobically ovemight at 35°C. The same laboratory technician, who was blinded from the clinical findings, read the culture plates on the two folIowing days. Standard techniques were used for identification of S. pneumoniae, GAS, H. influenzae, M. catarrhalis, and S. aureus. Semiquantitative estimation was done by growth in number of three streaks. Pure cultures were produced and frozen in ox-broth at -80°C in Nuuk and -20°C in Sisimiut. Non-pathogens were only registered. The other swab wasimmediately placed in chlamydia transportmedium for later detection of Chlamydia trachomatis, Chlamydia pneumoniae, and Mycoplasma pneumoniae and stored. Ear discharge was aspirated into a cryo tube after cleaning of the outer external ear canal. Two swabs were dipped into the sample and treated as described for the nasopharyngeal swabs. The surplus was immersed in virus transport-medium for viral detection. Aspiration of nasopharyngeal secretions for virus detection was done via the nasal route by introducing a tube in which was injected 2 ml sterile virus transport-medium. Slides were produced for immunofluorescense-antibody (IFA) examinations and frozen. Nasopharyngeal aspirations for IFA examinations were obtained from 122 of the 127 control children in Nuuk, from 39 of the 40 AOM children, and from ear discharge of 14 out of 16 children. In Sisimiut, nasopharyngeal aspirations for IFA examinations were obtained from all 14 AOM children, from 14 age matched controls, and from ear discharge of 6 out of 8 children. Virus cultures and virus PCR examinations were only examined for in a case-control study of children from Nuuk where the specimens were stored at the mandatory -80°C. Included were 17 boys and 22 girls with AOM and 39 (22 boys and 17 girls) age matched controls. All frozen bacterial and viral specimens were transported on dry ice to Copenhagen. Further bacteriological and viral analyses were performed at specialized laboratories in Denmark and all analyses were performed blinded from clinical findings. The applied techniques have been described in detail in article V. The analyses included capsular typing and oxacillin susceptibility of S. pneumoniae (Statens Serum Institute), grouping and T-typing of GAS (Statens Serum Institute), biotyping, serotyping and B-Iactamase production of H. influenzae (Department of Medical Microbiology and Immunology, University of Aarhus), M. catarrhalis identification and B-lactamase production (Department of Clinical Microbiology, Rigshospitalet), S. aureus phage-typing and methicillin susceptibility (Statens Serum Institute), detection of Chlamydiae by culture and PCR (Statens Serum Institute), and detection of Mycoplasma pneumoniae by PCR (Statens Serum Institute). Virus detection included direct IFA techniques for 30

adenovirus, RSV, influenza type A- and B-virus, and parainfluenza type 1,2, and 3 -virus. All were examined by commercially available diagnostic kits: IMAGENTM Adenovirus, IMAGENTM RSV, IMAGENTM Influenza Virus A and B, and IMAGENTM Parainfluenza Virus Group (Types l, 2, and 3) (DAKO A/S, Glostrup, DK). The company reported between 93% and 100% sensitivity and specificity of the applied kits. Primary monkey (cynomolgus) kidney, human diploid cell line (MRC5), and Hl-Hela cells were established for the detection of rhinoand enterovirus by culture. Polymerase chain reaction (PCR) was furthermore us'ed for rhinoand enterovirus detection (Department of Virology, University of Copenhagen). 5.4.3. Results Potentially pathogenic bacteria were carried in the nasopharynx by 98% with AOM, 91 % without AOM, and by 94% of healthy children. Furthermore, more than one species of potentially pathogenic bacteria (mixed flora) was carried by 78% with AOM and 57% without AOM. H. influenzae was carried by 92% with AOM, 77% without AOM, and by 80% of healthy children. There were statistically significant differences in age specific carriage rate s for H. influenzae non-b, M. catarrhalis, and mixed flora which were found in higher frequencies in the youngest control children. S. aureus was found statistically more often in the youngest healthy children. S. pneumoniae was, however, the only potentially pathogenic bacteria to be carried significantly more often in the nasopharynx of AOM children (48%) compared with age matched controls (26%) (X2 = 10.78 with 1 df; P < 0.03). H. influenzae non-b was found frequently in all age groups (57-82%) of control and healthy children. H. influenzae serotype b (Hib) was carried by 11% of AOM children, 9% of controls, and 11% of healthy children. H. influenzae and S. pneumoniae were found in 22 of 24 ear discharges. Chlamydiae were detected in the nasopharynx of nine children, three of whom had AOM and seven of whom reported rhinitis. No healthy children carried chlamydiae. All ear discharge specimens, except one, contained bacteria from the nasopharyngeal flora. The bacteria cultured from ear discharge were predominantly S. pneumoniae and H. influenzae while M. catarrhalis was cultured in four ear discharge specimens and only in combination with other potentially pathogenic bacteria. Mixed flora was cultured in 11 of 24 (46%) ear discharge specimens. The IFA examinitions (n = 209) only revealed influenza type B virus in one ear discharge specimen and in one nasopharyngeal specimen from two children in Sisimiut. Picornaviruses (enteroviruses or rhinoviruses) were detected in 23 (59%) nasopharyngeal specimens from 39 children with AOM and in 13 (33%) nasopharyngeal specimens from 39 children without AOM (p = 0.04), and in 4 (29%) of 14 ear discharge specimens. Findings of virus were associated with reports of rhinitis, but reports of rhinitis were not associated with findings of virus. 5.4.4. Discussion Other indigenous populations also suffer from high prevalence of OM (Sunderman and Dyer 1984, Todd and Bowman 1985, Giles and Asher 1991) and studies of the nasopharyngeal colonization with potentially pathogenic bacteria in two of these populations (children in Papua

International Journal of Circumpolar Health vol. 60; suppl. 2, 200l

31

New Guinea and Aboriginal infants) disclosed early and heavy colonization (Lehmann 1992, Leach et al 1994). The findings in Greenlandic children are in accordance with these reports and suggest an association between massive and early load of potentially pathogenic bacteria in the nasopharynx and a high rate of OM. In comparable studies from e.g. Sweden, and Buffalo, U.S.A. the overall nasopharyngeal carriage rate of potentially pathogenic bacteria was 63% and 60% in control children compared with 91 % in Greenlandic children (1ngvarsson et al. 1982a, Faden et al. 1991). S. pneumoniae, H influenzae non-b, and mixed flora were all carried in higher frequencies in control children in Nuuk and Sisimiut than in control children in Malmø and Buffalo. Technically, it was a marked difference that direct inoculation on culture plates was performed in the present study whereas transport media have been used in most other studies thereby causing delayed inoculation. However, delayed inoculation was also used in the studies of children in Papua New Guinea and infant Australian Aborigines (Lehmann 1992, Leach et al. 1994). The bacterial findings in the nasopharynx have been shown to be influenced by age, season, geography, sampling technique, and selectivity of isolation media (Box et al. 1961, Schwartz et al. 1979, Henderson et al. 1982, Long et al. 1983, Freijd et al. 1984, Kuklinska and Kilian 1984, Prellner et al. 1984, Sørensen et al. 1988, Arola et al. 1990, Faden et al. 1991, Ruuskanen et al. 1991, Aniansson et al. 1992). Thus, differences between studies in these matters may to some extent explain the differences in results. However, the present findings were markedly distinct from the results obtained in Sweden and U.S.A. but in accordance with the findings in Australian Aborigines and, hence, most likely represent a true difference in nasopharyngeal carriage rates of potentially pathogenic bacteria. In ear discharge, all potentially pathogenic bacteria were found more frequently in this study compared with others (Karma et al. 1983, Giebink 1989). However, the studies were not directly comparable due to several factors of which the most important ones probably were that i) the present samples represented acute suppurations from spontaneously ruptured tympanic membranes and not suppurations obtained from the middle ear after myringotomy, ii) the present study used direct inoculation on culture plates, and iii) the present sample size was small (n = 24).

Species specific examinations revealed the same pneumococcal types in Greenlandic children with AOM as in other populations, which indicates that a conjugate pneumococcal vaccine immunogenic in children below the age of two years should contain the same pneumococcal types as elsewhere (Lund and Henrichsen 1978, Long et al. 1983, Prellner et al. 1984, Sørensen et al. 1988). Hib was found very frequently in AOM, control and healthy children compared with 2%-4% Hib carriage rates in healthy persons as reported in the review by Turk (1984) and 1.3% in Chinese and Vietnamese children in Hong Kong (Sung et al. 1995). Hib vaccination was not initiated at the time of the present study in Greenland. In accordance with other studies examination for H. influenzae biotypes showed that biotype II was significantly more often cultured from the nasopharynx of AOM children and accounted for 42% of H. influenzae biotypes found in ear discharges (Long et al. 1983, Freijd et al. 1984). This indicated differences in the virulence of the various biotypes of H. influenzae. Chlamydiae and M. pneumoniae were not of major importance in AOM in Greenland. In other studies these microorganisms have only sporadically been associated with AOM while two recent studies found Chlamydiae pneumoniae rather frequently by PCR in middle ear fluids from AOM and SOM children (Kamme et al. 1971, Chang et al. 1982, Storgaard et al. 1997, Falck et al. 1998). 32

All capsular pneumococcal species were susceptible to oxacillin and surprisingly only 5% of the isolated H. influenzae strains produced B-lactamase while an expected 75% of M. catarrhalis strains produced B-lactamase. Thus, although the prescription of Β-lactams was frequent in Greenland, bacterial resistance was not a major problem (Niclasen et al. 1995, Homøe et al. 1996 (article V)). Rhinitis is an important risk factor for AOM (Maynard 1969, Henderson et al. 1982, Fleming 1987, Harsten et al. 1989, Arola et al. 1990, Alho et al. 1990, Ruuskanen et al. 1991, Aniansson et al. 1994, Daly et al. 1999). In this study significantly more AOM children (87%) reported rhinitis than controls (25%) (p < 0.001). Entero- and rhinoviruses were detected in high rates either by PCR or culture in nasopharyngea1 aspirations from children in Nuuk compared with other studies (Henderson et al. 1982, Arola et al. 1990, Ruuskanen et al. 1991). However, PCR was not applied in these studies. Influenza type B virus was only found in two AOM children while adeno-, RS-, parainfluenza- type 1,2,3, and influenza- type A virus were not detected at all in contrast to other reports (Sarkkinen et al. 1981, Henderson et al. 1982, Arola et al. 1990, Ruuskanen et al. 1991). The negative results of the latter viruses may be exp1ained by a re1atively short sampling period (1 and 10 month) and the unique epidemiological conditions in Greenland with small and relatively isolated populations. Considering this, the results suggest that entero- and rhinoviruses occur endemically and are of importance in the pathogenesis of . AOM in Greenland.

5.4.5. Conclusion Bacteria associated with AOM in Greenlandic children were of the same species as in other parts of the world but the bacterial load of potentially pathogenic bacteria in the nasopharynx of children with AOM, without AOM, and in healthy children was early and massive. Entero- and rhinoviruses were detected frequently and furthermore rhinitis was reported prevalently. These combinations may be crucial for the high prevalence of OM among Greenlandic children.

5.5. Risk factor survey (articles VII & VIII) 5.5.1. Aims Risk factors for OM may be host related, i.e., genetic or related to the environment. These factors probably function in a complicated interplay. In 2,750 pairs of Norwegian adult monozygotic and heterozygotic twins Kværner et al. (1997a) estimated that the variation in liability for recurrent ear infections consisted of 74% heritability in females and 45% in males. URTI's have been claimed to be the most powerful risk factor associated with OM (Maynard 1969, Fleming et al. 1987, Harsten et al. 1989, Arola et al. 1990, Alho et al. 1990, Aniansson et al. 1994, Daly et al. 1999). Eustachian tube dysfunction as a cause of immaturity or anatomical factors has also been shown to be of importance (Doyle 1977, Shapiro 1987, Maw et al. 1991, Stenstrom et al. 1991). It has been indicated that early and massive nasopharyngeal colonization with potentially pathogenic bacteria may p1ay a role in the pathogenesis of OM especially in high risk populations (Leach et al. 1994, Homøe et al. 1996, article V). Risk factors in studies from other parts of the world include race, male gender, cleft palate, young age at first AOM,


33

episode, siblings in the household with URTI, parental history of OM, attendance in large size daycare centers, exposure to passive smoking, lack of breast feeding, feeding position, atopy, prematurity, use of comforter, poor housing standard, and social poverty (i.e., poor living conditions) (Christensen 1956, Howie et al. 1975, Strangert 1977, Ingvarsson et al. 1982b, Saarinen 1982, Vinther et al. 1982, Ståhiberg et al. 1986, Fleming et al. 1987, Kero and Piekkala 1987, Shapiro 1988, Harsten et al. 1989, Teele et al. 1989, Alho et al. 1990 and 1993, Froom and Culpepper 1991, Duncan et al. 1993, Bastos 1994, Aniansson et al. 1994, Ey et al. 1995, Marx et al. 1995, Sorri et al. 1996, Uhari et al. 1996, Kværner et al. 1997a and b, Stenstrom and Ingvarsson 1997a and 1997b, Daly et al. 1999). However, results are conflicting for several of these risk factors, and most have only slightly been found to increase the risk for AOM. At the time the present study was p1anned little was known about risk factors for OM in Greenlanders as well as in Inuit populations in Alaska and Canada. However, a number of studies in Alaskan Inuit populations have reported that early age at first AOM episode is frequent and probably related to chronic problems later in life (Reed et al. 1967, Maynard 1969, Reed and Dunn 1970, Johonnott 1973, Kaplan et al. 1973, Stewart 1989). Reed and Dunn (1970) also found an association between episodes of OM and hearing deficiency but no relation between incidence of OM and season, weather conditions, allergy, size of family, indicators of crowding, we1fare indicators, sanitary conditions, or indicators of parental interest in the study. Johonnott (1973) reported significantly higher prevalence of COM among rural than urban Eskimo children, but no difference between frequency of otorrhea before the age of two and no significant difference in family history of AOM or COM. Schaefer (1971) and Timmermans and Gerson (1980) reported an association between COM and insuffIcient breast feeding or ear1y bottle feeding in Canadian Inuit. Both studies suffered from methodological problems, being retrospective and involving children up to 15 years of age and adults. Thus, other factors may have influenced or confounded the findings. From Greenland, Littauer and Elm (1964) reported higher OM morbidity in urban compared to rural children and Berg and Adler-Nissen (1976) reported an association between crowded housing conditions and episodes of OM. Bjerregaard (1983 and 1985) did not find any significant associations between AOM or CSOM episodes and housing conditions or social grouping even though there was a trend for CSOM with low social group. Pedersen and ZachauChristiansen (1986) did not demonstrate any significant risk factors but claimed that having an unemployed mother from a low social stratum was a determinant of middle ear disease. The four latter studies have been thoroughly described in chapter 3. Thus, the objective of this survey was to determine the possible sociomedica1 risk factors for AOM, rAOM, and COM in unselected Greenlandic children.

5.5.2. Subjects & Methods The subjects and inclusion procedure in these studies were the same as in article IV (see section 5.3.2). The parents or guardians answered a questionnaire concerning sociomedical variables which included number of former episodes of AOM, age at first AOM episode, myringotomy, adenoidectomy, tonsillectomy, traumatic head injuries, cleft palate, recent disease episodes, chronic diseases in general, and allergy. Most of these variables were cross-checked in the medical records which also formed the basis for the drop-out analyses. The social variables included native land of parents as a marker for ethnicity, history of OM in the family, i.e., parents and siblings if any, type or size or insulation standard of housing, crowding, type of daycare, passive exposure to cigarette smoking at home, breast feeding, and type of diet. Social classification was not obtained because no validated method for stratification of Greenlanders was available. The statistical analyses included simple frequency tests, calculation of RR and OR, and multiple logistic regression using backward stepwise exclusion. 34

5.5.3. Results Former episode of AOM was reported by 2/3 of the children, rAOM by 20%, and COM by 9%. Of all children in the survey, 40% had AOM during the first year of life. Median age of the first episode was 10 months, range 1-84 months and with no sex difference. The highest risk of AOM was between 7 and 12 months of age. Children with rAOM had their first AOM episode in a significantly younger age (median: 7 months, range: 2-48 months) than children with only one episode (median: 15 months, range: 1-84 months) (Kruskal-Wallis test: Χ2 = 45.4 with 1 df; p < 0.0001) and 83% of children with rAOM had the first AOM episode before 12 months of age compared with 53% of children with < 5 episodes (p < 0.0001). The relative risk of rAOM was eight times greater if first episode of AOM was before six months of age compared with first episode when older than 24 months. Thirtyfive percent of children with rAOM also had COM compared with only 4% of children with < 5 AOM episodes. The following variables were found significantly more often in children with AOM by univariate frequency testing: Parental (OR = 1.83), sibling (OR = 1.62), and parental plus sibling (OR = 2.56) history of OM, crowding (OR = 5.55), long period of exclusive breast feeding (> 4 months) (OR = 2.47), and recent acute disease (p = 0.034). The following variables were found significantly more often in children with rAOM or COM by univariate frequency testing: Parental history of OM (OR = 1.60; OR = 2.11, respectively) and no recall of breast feeding (p = 0.005; p = 0.003, respectively). COM was found significantly more often in children with two Greenlandic parents (OR = 3.07). In a multivariate test using multiple logistic regression only parental history of OM (OR = 1.82) and long period of exclusive breast feeding (OR = 1.14) were denoted as significant predictors of AOM. 5.5.4. Discussion Risk factor estimation is not optimal in a survey due to the temporal segregation between obtained risk factors (independent variables) and the outcome measure (dependent variable) and due to the problem of recall bias for several variables. However, especially variables determined early in life and hereditary variables such as parental OM and early age of AOM are in all probability reliable in the temporal aspect while recalling is more problematical. In this survey the AOM diagnosis was retrospectively registered and relied upon parental recalling and district physicians diagnoses written in the medical records. According to the definitions applied (see chapter 4), it was aimed to resemble the clinical situation as closely as possible. Furthermore, it


35

was of primary concern to increase the reliability of the AOM diagnosis and also of the variables by use of medical records which were available for 95% of the included children and for 74% of the drop-outs. Some of the variables were obtained as current status and thus merely possible to ascribe as potentially associated variables. It was, however, attempted to perform a longitudinal, prospective registration of AOM cases in several districts in Greenland but this failed due to inconstant reportations. Some of the variables were unevenly distributed such as e.g. breast feeding, which was reported for 91.5% of the children, exposure to cigarette smoking, which occurred in 79.2% of the homes, and use of daycare, which was attended by 87% of those who did not attend school. The robustness of the test procedures therefore tended to decrease because of the relatively small-sized comparable groups. Problems arose using multiple logistic regression implementing all variables as the dataset was incomplete for some of the variables such as daycare, which ruled out 178 children who attended school. Incomplete data were treated as missing values by the statistical software program. The multiple logistic regression test was therefore also applied on the three variables which were predictive for AOM on a p-level < 0.1 in the first procedure. This was considered acceptable because the correlation coefficients (r) between the different main groups of variables all were small (r < 0.14). Multivariate analysis was not performed for rAOM because this subgroup did not vary much from the group of AOM children in univariate frequency testing or for COM children, because the number of COM children was considered too small for multivariate analysis implementing more than ten main groups of variables. Thus, although this survey could not outline cause and effect in an ideal fashion, the results strongly indicated that early age at first AOM episode was an important risk factor for rAOM and that early first AOM episode occurred frequently among children in Nuuk and Sisimiut compared with fmdings in other studies from the Western world (Howie et al. 1975, Vinther et al. 1979, Ingvarsson et al. 1982, Harsten et al. 1989, Kværner et al. 1997b) but in accordance with reports from Alaska (Reed et al. 1967, Kaplan et al. 1973). Furthermore, familial and ethnical predisposition for AOM, rAOM, and/or COM seems to exist and may be ascribed to both genetical (intrinsic) and environmental (extrinsic) influence. Very heavily crowded households and long-term exclusive breast feeding or no recall of breast feeding were also significantly associated with either AOM, rAOM, or COM. However, these three variables may be confounders for socioeconomic factors and thereby serve as surrogate variables for low social status. In addition, long-term breast feeding without substitution may be an indicator of poor nutritional status, which has been indicated to be associated with OM in some developing countries (Bastos 1994). In a recent survey of 307 Inuit children from the Keewatin district in Canada, no associations were found between the presence of OM and sex, age, housing type, crowding, or socioeconomic status (Woods et al. 1994). However, significantly different rates of OM were found between communities (8-45%). At the same time as the present study, the Greenland Health Interview Survey was undertaken. This survey included questions to mothers concerning OM in their children below 13 years of age. The survey included 661 children of whom some were Danes. Seven percent answered that their child had had ear pain during the preceding 14 days, 33% of the children had had ear pain in some period of their life, 22% had had ear discharge in some period of their life, and 14% had had both (Curtis et al. 1997). Ear discharge was reported significantly more often in boys (25%) and in children cared for outside the home (27%). 36

However, this was not valid for ear pain and the two combined. There was a trend that ear pain and ear discharge were less common among children living in settlements than among children living in towns and this became statistically significant when combined (44% and 34%, respectively; p = 0.04). Crowding was not an associated factor. However, due to the rather simple questions denoting OM without any clinical validation the results should be taken with caution. In a Danish health status survey involving health care examination charts and parentally filled out health care questionnaires of 5,307 children 6-7-years-old, 44% reported ear problems prior to school start (Madsen et al. 1991). The frequency of earproblems was lowest for children cared for at home and highest for children living with only one parent. No associations were found between ear problems and parental occupation or housing conditions. OM as such cannot be ascribed to any single factor but probably to a complicated interaction between several intrinsic and extrinsic variables and occasional circumstances. However, the present survey seems to indicate that children are at increased risk of AOM, rAOM, and COM when the first AOM episode occurs before one year of age, when both parents are born in Greenland, when parents also have had a history of OM, when living in very crowded households, and when having a long period of exclusive breast feeding, or when recalling of breast feeding is impossible. 5.5.5. Conclusion Compared with other studies early onset of AOM (before one year of age) occurred frequently in Greenlandic children and a high proportion of these children developed rAOM. Many of the risk factors usually associated with AOM could not be confirmed as risk factors in this survey. Parental history of OM and long period of exclusive breast feeding were the strongest factors associated with AOM in Greenlandic children, and ethnicity was associated with COM. The study confirmed that AOM also in Greenland is a multifactorial disease determined by a number of genetic and environmental factors.

5.6. Mannose-binding lectin, AOM and Epstein-Barr virus study (article IX) 5.6.1. Aims Investigations of immunological factors associated with the high prevalence of OM in Greenland have been few, and in fact only two such studies exist. One has been published in proceedings (Prellner et al. 1993) and the other exists as an unpublished manuscript (Sørensen et al. (unpublished)). The study in Greenlanders by Prellner and co-workers has been referred in chapter 3. Sørensen and co-workers found no significant correlation between IgD-cell density and otitis proneness in nasopharyngeal lymphoid tissue from 23 Greenlandic children and there was an even distribution in IgD-cell density in Greenlandic and Danish children. Studies in e.g. Swedish children have shown that immunological deficiencies or characteristics are of importance in OM (Freijd et al. 1985, Prellner et al. 1989, Kalm et al. 1992, Lindberg et al. 1994). Thus, lower levels of serum IgG2 subclass have been reported in Swedish rAOM children than in age matched controls (Freijd et al. 1985). Also, significantly lower levels of IgG specific pneumococcal antibodies against pneumcicoccal types 6A and 19F have been found in cord blood sampies and blood from the first year of life in Swedish children later InternationaI JournaI of CircumpoIar Health vol. 60; suppI. 2, 2001 37

developing rAOM than in healthy children (Prellner et al. 1989). HLA-A2 antigen has been reported to be present significantly more often in Swedish children with rAOM than in controls (Kalm et al. 1992) and recently significantly lower levels of the inflammatory cytokines IL-1∃, IL-6 and TNF-∀ were found in nasopharyngeal secretions from children with rAOM than in healthy children (Lindberg et al. 1994). The purpose of the present study was to examine whether MBL genotypes, episodes of AOM, and EBV infection were associated in community-based unselected children from Nuuk. This idea was trickered by studies concerning MBL genotypes and MBL serum levels in Greenlanders and studies of EBV infections in Greenlanders (Garred et al. 1992, Madsen et al. 1994 and 1995, Melbye et al. 1984, Albeck et al. 1985). MBL acts as an opsonizer and has the capability of initiating the classical complement pathway when binding to mannose residues on yeast, bacteria or virus (Super et al. 1989, Anders et al. 1994, Turner 1996). MBL variant alleles (B, C, and D) lead dominantly to low MBL serum concentrations in heterozygotes and absence in homozygotes. The variant gene frequency is 0.13 in East Greenlanders compared with 0.20 in Caucasians (Madsen et al. 1994). Due to a high frequency of the promotor gene variant haplotype HY = 0.83 in Eskimos, the median serum MBL levels in persons homozygous for normal MBL alleles are 2.5 to 5 times higher in East Greenlanders than in Europeans and Africans (Garred et al. 1992, Madsen et al. 1995). Absent or low MBL serum concentrations have been associated with recurrent infections in early childhood in hospitalized children and in immunodeficient patients (Richardson et al. 1983, Super et al. 1989, Garred et al. 1995, Summerfield et al. 1997). It has also been suggested that high MBL levels are associated with certain infections caused by intracellular parasites such as Mycobacterium leprae, Mycobacterium tuberculosis, Legionella pneumophilia, and certain viruses such as herpes simplex type 2 (Fischer et al. 1994, Garred et al. 1994,). Thus, some intracellular microorganisms may use MBL to by-pass the first line immune defence system and promote infection. An association may therefore also exist between MBL genotypes and EBV infection. EBV infections have been reported to be early and frequent in Greenlandic children (Melbye et al. 1984, Albeck et al. 1985). EBV infection is known to stimulate the production of the anti-inflammatory cytokine IL-10 which inhibits the function of macrophages, T-killer lymphocytes, and natural killer cells. Therefore, an outcome of EBV infection could be a facilitation of nasopharyngeal colonization with potentially pathogenic bacteria, which has been shown to be early and massive in Greenlandic children (Homøe et al. 1996 and 1998, articles V and VI). 5.6.2. Material & Methods The children in this study were part of the children from Nuuk who also participated in studies V and VI. Of the 127 unselected community-based children available for the study, 45 children were excluded due to lack of surplus material for the detection of MBL alleles and EBV. Thus, 82 children participated in this study. Of these 36 were girls and 46 boys with a median age of 5 years, range 0-8 years. Fifty-nine children had Greenlandic born parents, 20 children had one Greenlandic born parent and three children had Danish born parents. The nasopharyngeal sampling and the bacterial culture and detection methods are described in article V. DNA for MBL 38

genotype examination was extracted from the nasopharyngeal samples and MBL genotyping of structural alleles was performed according to standard procedures at the Department of Clinical Immunology at Rigshospitalet, Copenhagen (Madsen et al. 1994). Preparation of DNA from the nasopharyngeal samples and analyses for EBV, including control with positive and negative EBV cell lines, were performed at the Laboratory of Immunopathology, Institute of Pathology at Aarhus University Hospital (Sandvej et al. 1994). EBV was detected by nested PCR for two EBV specific genes, the EBNA-2 locus and the EBNA-3c locus. According to Kunimoto et al. (1992) the sensitivity of the first round EBNA-2 PCR was one EBV positive cell of 104 EBV negative cells. The procedures gave results concordant with single step procedure when tested on EBV positive cell lines. The PCR product was verified after gel-electrophoresis stained with ethidium bromide. History of former episodes of AOM and rAOM were obtained for the 3-8year-old children as described in articles VII and VIII. 5.6.3. Results MBL genotypes were detected in 89% (73/82) of the obtained samples. The distribution of the alleles followed the Hardy-Weinberg expectations with a frequency of normal wildtype (A) allele of 0.88 and a frequency of variant alleles (B, D, and C) of 0.12 (0.08 for B and 0.04 for D while C was not found). Homozygosity for AA, implying normal level of MBL, was present in 68%, heterozygosity AB and AD, implying low level of MBL, was present in 20% and homozygosity BB or DD, implying absence of MBL, was present in 1%. EBV DNA was detected in 50% (41/82) of the nasopharyngeal samples. Carriage of EBV increased by age: 10% (1/8) in 01-year-olds to 60% (5/9) in 3-year-olds but thereafter the carriage rate was constant. Children with Greenlandic parents carried EBV significantly more often (34/59) than children with parents of mixed ethnicity (7/20) and children with Danish parents (0/3) (X2 = 6.19 with 2 df; p = 0.046). Homozygosity for A and heterozygosity/homozygosity for B and D were not differently associated with EBV carriage (32/56 and 8/17, respectively, p = 0.65). As shown in Table 5.6.3.1, MBL genotypes were not associated with nasopharyngeal carriage of any of the potentially bacterial pathogens. Table 5.6.3.2 shows that GAS and H. influenzae were found significantly more often in children who were EBV positive. However, when adjusting for mass significance according to the Bonferroni method the significance disappeared. Furthermore, there was no significant difference in the findings of potential pathogens, no pathogens or mixed pathogens between children who had MBL genotype AA and were EBV positive or negative (see Table 5.6.3.3). MBL genotypes, EBV findings, and episodes of AOM and rAOM were not associated in the 74 children in whom a history of AOM episodes were obtained. Furthermore, examination for age at first episode of AOM did not reveal any significant associations.

Internationa1 Journa1 of Circumpo1ar Hea1th vol. 60; supp1. 2, 2001

39

Table 5.6.3.1 Mannose-binding lectin (MBL) genotypes and potentially pathogenic bacteria in the nasopharynx of 82 children from Nuuk, Greenland. X2-test was used to test for difference between the groups. MBL genotypes could not be detected in nine specimens. MBL genotypes AB/DD/AD AA n = 17 n = 56

Potential pathogens

6 8 5 42 11 2 23

Staphylococcus aureus Streptococcus pneumoniae Group A streptococci (GAS) Haemophilus influenzae Moraxella catarrhalis Chlamydiae Mixed pathogens*

0 3 0 10 7 0 6

p

ns ns ns ns ns ns ns

* denotes > one potential pathogen

Table 5.6.3.2 Association between Epstein-Barr virus (EBV) and potentially pathogenic bacteria in the nasopharynx of 82 children from Nuuk, Greenland. X2-test was used to test for difference. EBV Potential pathogens

Positive, n=41

Negative, n=41

p

2

5 5 0 24 17 l 17

ns ns 0.02 0.02 ns ns ns

Staphylococcus aureus Streptococcus pneumoniae . Group A streptococci (GAS) Haemophilus influenzae Moraxella catarrhalis Chlamydiae Mixed pathogens* * denotes > one potential pathogen

40

6 5 34 11 1 16

Table 5.6.3.3 Homozygosity for normal mannose-binding lectin (MBL) genotypes with and without carriage of Epstein-Barr virus (EBV) and nasopharyngeal colonization with potentially pathogenic bacteria in children from Nuuk, Greenland. X2-test was used to test for difference. Nasopharyngeal colonization

MBL AA + EBV pos. MBLAA + EBV neg. n=32 n=24

p

+ potential pathogen - potential pathogen

31 1

22 2

ns ns

Mixed pathogens*

13

10

ns

* denotes > one potential pathogen

5.6.4. Discussion The results of the present study in community-based, unselected Greenlandic children supports the fmdings by Garred and co-workers (1993) in Danish children in whom neither MBL genotype nor MBL serum levels, MBL levels in middle ear effusions, or MBL levels in nasopharyngeal secretions were associated with otitis proneness, rAOM (less than otitis proneness) or SOM. The studies which have shown association between low MBL serum level and recurrent infections have been performed in children hospitalized for a variety of infections including serious invasive infections, and the controls have consisted of children staying in hospital for other reasons (Richardson et al. 1983, Super et al. 1989, Summerfield et al. 1997). Thus, low MBL serum levels may still be associated with infections in children with other immunological disorders as suggested by Garred et al. (1995) and infections in children with low MBL serum levels may be more virulent and cause more frequent hospital admittance. Except for the variant allele D, the MBL allele frequencies in the present study were in accordance with the findings in East Greenlanders (Garred et al. 1992). The presence of the D allele may be explained by Caucasian admixture. 5.6.5. Conclusion This study revealed no associations between MBL genotypes, EBV infections, and episodes of AOM or rAOM in community-based children in Greenland and did not support the hypotheses that MBL levels and EBV infections account for the high prevalence of OM and early first episode of AOM in Greenlandic children.

International Journal of Circumpo1ar Health vol. 60; supp1. 2, 2001

41

6. FUTURE STUDIES The present studies have outlined a rosette of new research projects concerning the high prevalence of OM in Greenlandic children. There is a need for more knowledge on descriptive epidemiology of OM in remote districts and settlements, and studies on OM complications such as mastoiditis do not exist. Furthermore, the epidemiology of OM in Greenlandic children living in Denmark would be interesting to investigate as it may contribute with further knowledge of hereditary and environmental factors. Such a study is being considered at present by the local ethics committee. However, most important are prospective studies of birth cohorts focusing on regular clinical otological examinations and registrations of disease episodes, especially URTI's and OM. Risk factor variables should also be registered regularly due to common and frequent changes in some of these factors. Information concerning the importance of the bacteriological and the virological nasopharyngeal colonization and the colonization dynamics either alone or in interplay with risk factors for URTI and OM could also be obtained from such a study design. In fact such a study is already started in Sisimiut and is at present in the process of data examination (Koch et al. 1996). Long-term follow-up studies are needed, especially of children with rAOM and COM, focusing on side effects such as developmental, cognitive, and educational aspects of OM in a small, isolated population such as the Greenland one which is constantly exposed to the world's increasing educational needs. Still our knowledge is sparse or lacking in respect to the role of possible anatomical differences in Greenlanders, especially concerning the Eustachian tube and neighbouring spaces in the newborns and infants. Furthermore, a deeper insight is wanted into the immunological system and function including immunoregulatory mediators in the highly selected indigenous population in Greenland. Intervention studies aiming at reducing the high prevalence of URTI's and OM are recommended such as e.g. examination of the effects of longer maternal leave or the effects of increased public information on these diseases. Studies concerning hearing and HL including hereditary and congenital HL are called for. Another very interesting question deserving scientific study is the impact of outside influences on health and OM morbidity in Greenlanders, including the immunological homeostasis.

7. CONCLUDING REMARKS Studies in Greenland are difficult to perform due to logistic obstacles and heavy financial burdens, but also very fruitful and rewarding. In general, it is my impression that Greenlanders are very concerned about their health and especially the health of their children. Furthermore, the medical staff is very concerned about the many and severe health problems such as the high infant mortality rate, the high prevalence of suicides, mental disorders, etc.. OM may seem a small problem but cause, along with URTI's, frequent health care contacts, high economic costs for the society, and also frequent familial and individual problems. Thus, there is a need for increased concern and medical effort toward this prevalent problem. It is my hope, therefore, that the presented results will prove to be of clinical advantage to the Greenlanders and conducive to further investigations in this area.

42

8. ENGLISH SUMMARY This thesis describes the different aspects of otitis media (OM) in the population of Greenland viewed in a historical and modern c1inical perspective. Chapter 1 outlines the addressed problems and aims while chapters 2 and 3 deal with historical studies and an evaluation of the present knowledge based on the literature. Physical anthropological studies, using skeletal samples of adult Eskimo crania from before and after the colonization of Greenland in 1721 and information about modern living Eskimos (Inuit), have shown that OM sequelae of the temporal bones were significantly less common in pre-colonization Eskimos and that the mean area size of the pneumatized cell system in the temporal bone was significantly larger in pre-colonization Eskimos. These findings indicated an increase in OM after the colonization most likely caused by the social, cultural, habitary, and dietary changes due to increased contact with the outside world. Historical reports after the colonization confirm a high prevalence of OM especially in children. Modern epidemiological studies from the 1960's to 1980's in the Arctic region of Alaska, Canada, and Greenland along with reports from visiting consultant otologists in Greenland almost uniformly mention prevalent OM problems in children as well as in adults. The aim was therefore to further describe the epidemiological pattern of the different OM disease entities (acute OM (AOM), chronic OM (COM), COM with suppuration (CSOM), secretory OM (SOM), and cholesteatoma) and investigate the potentially associated risk factors in especially Greenlandic children because these diseases are primarily established and problematical in childhood. Chapter 4 describes the definitions used in the thesis and chapter 5 describes the studies included. Section 5.1 describes a study of cholesteatoma in Greenlanders. The study revealed an almost similar incidence of hospital treated children with cholesteatoma (6.6 per 100,000) as seen in comparable studies from other parts of the world. Furthermore, childhood cholesteatomas were the most aggressive. The frequency of residuals or recurrences after otosurgical treatment was high with a trend for better results when using the extensive canal wall-down procedure. It could be concluded that these patients urgently need close follow-up for at least five years postoperatively, if not lifelong. Section 5.2 describes a hearing screening survey of 167 school children using school registration charts. A high prevalence of hearing loss (HL) was found. A total of 43% of the children had hearing thresholds exceeding 20 dB at one or more frequencies between 250-8000 Hz in one or both ears, and 19% had the same type of HL in the frequencies 500-2000 Hz. HL was significantly associated with episodes of OM. These findings were in accordance with reports from Alaska and Canada. It is therefore concluded that a hearing screening programme of school children is important and that OM seems to have an impact on hearing in school children in Greenland. In section 5.3 an epidemiological survey is described concerning the prevalence of the different OM disease entities. The survey was carried out in Nuuk and Sisimiut and involved 740 children aged 3, 4, 5, and 8 years. A total of 591 children participated and selection bias was not found when controlling for age, sex, and episodes of AOM. The survey revealed that 52% of children in Nuuk and 54% in Sisimiut had some kind of pathological affection of their middle ear. COM and CSOM were found in 9%, but more prevalent among children in Sisimiut (12%) than in Nuuk (7%). Middle ear effusion (MEE) diagnosed by tympanometry was found in 23% in Nuuk and 28% in Sisimiut while simple tubal dysfunction (STD) was found in 13% and 8%,


43

respectively. MEE and STD were associated with young age. Sequelae of OM was apparent in 11% in both towns. When comparing the results with a 10-year-older, almost similar survey of 142 children, it was evident that the OM situation had not changed in the period between the studies. The survey underlines the need for increased focus on the different OM entities in Greenlandic children. Section 5.4 deals with microbiological aspects. The nasopharyngeal microflora and ear discharge microflora of potential pathogens were evaluated in 54 children with AOM and in 201 control children without AOM. Very high carriage rates expressed qualitatively and semiquantitatively of potentially pathogenic bacteria were found in the nasopharynx of children with AOM (98%) but also in that of the control children (91%) and even in children denoted as being very healthy (94%). However, the same bacterial species were cultured from the nasopharynx and ear discharge as in comparable studies world-wide. Only S. pneumoniae was carried significantly more often in the nasopharynx of AOM children compared with age matched control children. Chlamydiae, M. pneumoniae, adenovirus, respiratory syncytial virus, parainfluenza- type 1,2, and 3 virus, and influenza- type A and B virus were not major pathogens. In contrast, enteroand rhinoviruses were detected significantly more frequent in nasopharyngeal specimens from AOM children (59%) compared with age matched controls (33%) and also in 29% of the examined ear discharge specimens. It is therefore concluded that the potentially pathogenic bacterial load is early and massive. This alone or in interplay with entero- and rhinovirus infection and occasionally with other viruses may play an important role in the high prevalence of OM among children in Greenland. Section 5.5 deals with an examination of potential risk factors for AOM, recurrent AOM (rAOM), and COM in the same 591 children as studied in section 5.3. Early age at first AOM episode was associated with rAOM episodes (≥ 5 episodes since birth). Thus, the relative risk of developing rAOM was eight times higher if the first episode of AOM occurred before 7 months of age than after 24 months of age. Furthermore, compared with studies elsewhere in the world, a high proportion (40%) of the children in this survey had their first AOM episode during their first year of life and 41% of these children developed rAOM. It was also found that children had an increased risk of AOM, rAOM, or COM when both parents were born in Greenland, when parents also have had OM, when living in very crowded households, and when having experienced a long period of exclusive breast feeding, or when recalling of breast feeding was not possible. Gender, type, and size of housing, insulation standard of housing, daycare, exposure to passive cigarette smoking, and dietary habits were not associated with AOM, rAOM, or COM in the surveyed children. It is concluded that early onset of AOM occurs frequently in Greenlandic children and that a high proportion of these children develop rAOM. The study confirms that AOM is a highly multifactorial disease determined by a number of genetic and environmental factors. Finally, section 5.6 is a hypothesis generating study attempting to explain the high prevalence of early episodes of AOM in community-based children in Nuuk. The hypothesis is based on a possible association between findings of mannose-binding lectin genotypes, early EpsteinBarr virus infections and episodes of AOM, rAOM, or nasopharyngeal colonization with potentially pathogenic bacteria. However, the study does not support any of this hypothesis. In chapter 6, future studies are suggested and chapter 7 presents concluding remarks.

44

9. DANISH SUMMARY Afhandlingen behandler mellemørebetændelse (OM) blandt befolkningen i Grønland i et historisk og moderne klinisk perspektiv. I kapitel 1 beskrives problemstillingen og formål. I kapitlerne 2 og 3 beskrives de historiske undersøgelser i forbindelse med en litteratur gennemgang. Fysisk antropologiske undersøgelser af dels skeletmaterialer bestående af voksne eskimokranier fra henholdsvis før og efter koloniseringen af Grønland i 1721 og dels levende voksne grønlændere viste, at tegn på gennemgået OM bedømt ved forandringer i tindingeben var signifikant mindre hyppigt forekommende i materialet fra tiden før koloniseringen. Ligeledes fandtes middelarealet af det pneumatiserede cellesystem i tindingebenene at være signifikant større i kranierne før koloniseringen. Disse fund tyder på en øget hyppighed af OM efter koloniseringen og er således sammenfaldende med de sociale, kulturelle, boligmæssige og kostmæssige levestilsforandringer som koloniseringen af Grønland medførte. I den tidlige litteratur efter koloniseringen beskrives høj forekomst af OM blandt specielt de grønlandske børn og moderne epidemiologiske undersøgelser udført i 1960'erne til 1980'erne af inuit befolkningerne i henholdsvis Alaska, Canada og Grønland har alle bekræftet denne høje prævalens af OM problemer blandt såvel børn som voksne. Denne afhandlings formål har således været at beskrive epidemiologien af de forskellige OM sygdomme (akut mellemørebetændelse (AOM), kronisk mellemørebetændelse (COM), kronisk flydeøre (CSOM), mellemørekatar (SOM), og kolesteatom) samt mulige årsagssammenhænge hertil blandt specielt børn i Grønland, idet disse sygdomme overvejende grundlægges og er problematiske i barndommen. I kapitel 4 beskrives de benyttede definitioner og kapitel 5 omhandler de enkelte studier som indgår i afhandlingen. Afsnit 5.1 omhandler undersøgelsen af kolesteatom blandt 34 grønlændere i perioden 1976-1991. Undersøgelsen, som er baseret på et hospitalsbehandlet materiale, afslørede en stort set identisk incidens af kolesteatom (6.6 per 100,000) i Grønland som rapporteret i sammenlignelige undersøgelser fra andre steder i verden. Kolesteatom blandt børn var ligeledes mest aggressive. Residual eller recidiv kolesteatom efter otokirurgi forekom hyppigt med en tendens til bedre resultater ved ekstensiv kirurgi. Patienter med kolesteatom bør kontrolleres nøje i mindst 5 år postoperativt og formentlig livslangt. Afsnit 5.2 beskriver en skoleaudiometrisk screeningsundersøgelse af 167 skolebørn i Nuuk og Sisimiut. Hørenedsættelse blandt børnene forekom hyppigt og sammenlagt havde 43% af børnene høretærskler dårligere end 20 dB på et eller begge ører ved en eller flere frekvenser mellem 250-8000 Hz. Samme type hørenedsættelse fandtes hos 19% af børnene i de vigtige frekvenser mellem 500-2000 Hz. Hørenedsættelse var signifikant hyppigere hos børn med tidligere OM problemer. Disse resultater er analoge med resultater af undersøgelser blandt inuit i Alaska og Canada. Undersøgelsen konkluderer, at audiometrisk skolescreening er vigtig og, at OM blandt børn har betydning for den høje prævalens af hørenedsættelse i Grønland. I afsnit 5.3 beskrives i en epidemiologisk tværsnitsundersøgelse prævalensforholdene af de forskellige infektiøse mellemøresygdomme blandt børn i Nuuk og Sisimiut. Der indgik 740 børn i aldrene 3, 4, 5, og 8 år. Ialt deltog 591 børn, og der kunne ikke konstateres selektionsbias blandt deltagere og udeblevne ved kontrol for alder, køn og antal OM episoder. Undersøgelsen viste at 52% af børnene i Nuuk og 54% af børnene i Sisimiut havde en patologisk mellemøretilstand. COM og CSOM fandtes hos 9%, men hyppigere blandt børnene i Sisimiut (12%) end i Nuuk (7%). Tympanometrisk vurderet væske i mellemøret (MEE) forekom hos 23% i Nuuk og 28% i Sisimiut, hvorimod simpel tubadysfunktion forekom blandt henholdsvis 13% og 8%. SOM var hyppigst hos de yngste børn. Trommehindesequelae forekom blandt 11% i begge byer. Undersøgelsens resultater var i overenstemmelse med resultater opnået i en 10 år Internationa1 Journa1 of Circumpo1ar Hea1th vol. 60; suppl. 2, 2001 45

ældre sammenlignelig undersøgelse af 142 børn i Grønland. Undersøgelsen understreger behovet for øget fokusering på OM problemerne blandt børn i Grønland. Afsnit 5.4 beskriver næsesvælgfloraen blandt AOM børn og kontrolbørn samt floraen i øreflåd fra AOM børn. Ialt indgik 54 konsekutive børn med AOM og 201 kontrolbørn uden AOM. Der blev fundet høje bærerfrekvenser af mulige patogene bakterier såvel kvalitativt som semikvantitativt i næsevælgpodninger blandt AOM børn (98%), men også blandt kontrolbørn (91%) og endda blandt børn med speciel lille sygelighed (94%). Der blev identificeret samme slags bakterier i næsesvælget og i øreflåd som blandt børn andre steder i verden. S. pneumoniae var eneste bakterie som blev identificeret signifikant hyppigere i næsesvælget blandt AOM børn i forhold til aldersmatchede kontrolbørn. Chlamydiae, M. pneumoniae, adenovirus, RS-virus, parainfluenza- type 1,2 og 3 virus samt influenza- type A og B virus fandtes kun sjældent eller slet ikke. Derimod kunne entero- og rhinovirus detekteres signifikant hyppigere i næsesvælgsekret fra AOM børn (59%) i forhold til aldersmatchede kontrolbørn (33%) og i 29% af de undersøgte øreflådssekreter. Undersøgelsens hovedresultat er tidlig og massiv kolonisering af næsesvælget med mulige patogene bakterier blandt de undersøgte børn. Dette kan måske sammen med entero- eller rhinovirus infektion og periodevis infektion med andre virus være af betydning for den høje prævalens af OM blandt grønlandske børn. Afsnit 5.5 beskriver en undersøgelse over mulige sociomedicinske risikofaktorer i forbindelse med AOM, recidiverende AOM (rAOM: ≥ 5 episoder siden fødslen) og COM blandt de 591 børn som indgik i den epidemiologiske undersøgelse i afsnit 5.3. Tidlig første AOM episode var signifikant associeret med udvikling af rAOM. Den relative risiko for rAOM var således 8 gange større såfremt første AOM episode var opstået før 7 måneders alderen i forhold til efter 24 måneders alderen. Sammenlignet med studier fra andre områder af verden forekom AOM før l års alderen blandt en stor andel (40%) af børnene i Grønland og 41% af disse udviklede rAOM. Signifikant højere frekvens af AOM, rAOM eller COM forekom blandt børn, hvor begge forældre var født i Grønland, hvor forældre også havde haft OM problemer, og blandt børn som levede i meget trange leveforhold, var blevet selektivt ammet uden tilskud over en meget lang periode, eller hvor amning ikke kunne huskes. Køn, boligtype, boligstørrelse, boligisoleringskvalitet, pasningsforhold, udsættelse for passiv cigaretrygning og kostforhold kunne ikke associeres til AOM, rAOM eller COM i denne tværsnitsundersøgelse. Det konkluderes, at tidlig AOM debut forekommer hyppigt blandt børn i Grønland og en stor del af disse børn udvikler rAOM. Undersøgelsen bekræfter ligeledes, at AOM er en multifaktoriel sygdom som bestemmes af et antal genetiske og miljøbetingede faktorer. I afsnit 5.6 beskrives et hypotesegenererende studie, med henblik på at forklare den høje prævalens af tidlig AOM debut blandt børn i Nuuk. Hypotesen er baseret på en mulig sammenhæng mellem mannose-bindende lectin genotyper, tidlig Epstein Barr virus infektion og AOM, rAOM, eller næsesvælgkolonisering med mulige patogene bakterier. Imidlertid bekræfter undersøgelsen ikke disse hypotetiske sammenhænge. Endelig omhandler kapitel 6 forslag til fremtidige undersøgelser og kapitel 7 konkluderende bemærkninger.

46

10. REFERENCES Albeck H, Bille T, Fenger HJ, Narvestad U, Sørensen GS, Henle G, Henle W, Nielsen NH, Hansen JPH. Epstein-Barr virus infection and serological profile in Greenland Eskimo children. Acta Paediatr Scand 1985; 74: 691-96. Alho O, Koivu M, Sorri M, Rantakallio P. Risk factors for recurrent acute otitis media and respiratory infection in infancy. Int J Pediatr Otorhinolaryngol 1990; 19: 151-61. Alho O, Kilkku O, Oja H, Koivu M, Sorri M. Control of the temporal aspect when considering risk factors for acute otitis media. Arch Otolaryngol Head Neck Surg 1993; 119: 444-49. Anders EM, Hartley CA, Reading PC, Ezekowitz AB. Complement-dependent neutralization of influenza virus by a serum mannose-binding lectin. J Gen Virol 1994; 75: 615-22. Aniansson G, Alm B, Andersson B, Larsson P, Nylén O, Peterson H, Rigner P, Svanborg M, Svanborg C. Nasopharyngeal colonization during the first year of life. J Infect Dis 1992; 165 (suppl): S38-S42. Aniansson G, Alm B, Andersson B. Håkansson A, Larsson P, Nylén O, Peterson H, Rigner P, Svanborg M, Sabharwal H, Svanborg C. A prospective cohort study on breast-feeding and otitis media in Swedish infants. Pediatr Infect Dis J 1994; 13: 183-88. Arnvig J. Øre-, næse-, og halslidelser i Østgrønland. (In Danish). Annual report to the Chief Medical Officer in Greenland 1959: 1-5. Arola M, Ruuskanen O, Ziegler T, Mertsola J, Nanto-Salonen K, Putto-Laurila A, Viljanen MK, Halonen P. Clinical role of respiratory virus infection in .acute otitis media. Pediatrics 1990; 86: 848-55. Baxter JD, Ling D. Ear disease and hearing loss among the Eskimo population of the Baffin Zone. J Otolaryngol 1974; 3: 110-22. Baxter JD. Chronic otitis media and hearing loss in the Eskimo population of Canada. Laryngoscope 1977; 87: 1528-42. Baxter ID, Røjskjær C. A Greenland-Baffin zone project. J Otolaryngol 1979; 8: 315-25. Baxter JD. Observations on the evolution of chronic otitis media in the Inuit of the Baffin Zone, N.W.T.. J Otolaryngol 1982; 11: 161-66. Baxter JD. Clinical Research in the Canadian North. Acta Otolaryngol (Stockh) 1983; 95: 615 19. Baxter JD, Stubbing P, Goodbody L, Terraza O. The light at the end of the tunnel associated with the high prevalence of chronic otitis media among Inuit elementary school children in the Eastern Canadian Arctic is now visible. Arct Med Res 1992; 51: 29-31. Bastos I. Otitis media and hearing loss among children in developing countries. (Thesis) Lund: University of Lund, 1994: 1-116. Beal DD, Stewart KC, Fleshman JK. The surgical program to reduce the morbidity of chronic otitis media in the Alaskan Native. Acta Soc-Med Scand 1972; 6: 259-65. Beal DD, Nathanson SE, Chandler L, Springer S. Results of a surgical, otologic program in a remote Eskimo population. Alaska Medicine 1981; 23: 10-12. . Berg O, Adler-Nissen J. Housing and sickness in South Greenland: a sociomedical investigation. In: Proceedings of the III International Symposium on Circumpolar Health, eds. Shephard RJ & Hoh S. University of Toronto Press, Toronto and Buffalo 1976; 627-35. Bertelsen A. Grønlandsk medicinsk Statistik og Nosografi. Undersøgelser og erfaringer fra 30 års grønlandsk lægevirksomhed. III. Det sædvanlige grønlandske sygdomsbillede. (In Danish with English resume). Meddr Grønland 1940; 117: 139-59. International Journal of Circumpolar Health vol. 60; suppl. 2, 2001

47

Birch L, Iversen M, Elbrønd O, Lundkvist GR. A prospective epidemiological investigation of secretory otitis media and tubal dysfunction in children attending day-care centers. ORL J Otorhinolaryngol Relat Spec 1984; 46: 210-16. Bjerregaard P. Housing standards, social group, and respiratory infections in children of Upernavik, Greenland. Scand J Soc Med 1983; 11: 107-11. Bjerregaard P. Infectious diseases in Greenlanders of Upernavik. Scand J Prim Health Care 1985; 3: 163-69. Bjerregaard P, Young TK. The Circumpolar Inuit: health of a population in transition. Munksgaard 1998: 1-288. Bluestone CD. Epidemiology and pathogenesis of chronic suppurative otitis media: implications for prevention and treatment. Int J Pediatr Otorhinolaryngol 1998; 42: 207-23. Box QT, Cleveland RT, Willard CY. Bacterial flora of the upper respiratory tract. Am J Dis Child 1961; 102: 293-301. Branefors-Helander P, Dahlberg T, Nylén O. Acute otitis media. Acta Otolaryngol (Stockh) 1975; 80: 399-409. . Bretlau P, Kirkegaard J, Thomsen KA, Bohm J. Kronisk otitis media i nutidens Grønland (In Danish). In: Nuna Med 91 - en grønlandskmedicinsk konference. Kern P, Cordtz T. Nuuk Offset, Greenland, 1991; 136-38. Brodovsky D, Woolf C, Medd LM, Hildes JA. Chronic otitis media in the Keewatin district. In: Shephard RJ, Itoh S, eds. Proc III Int Symp Circumpolar Health. Toronto and Buffalo, Univ Toronto Press, 1976: 398-402. Brody JA, Overfield T, McAlister R. Draining ears and deafness among Alaskan Eskimos. Arch Otolaryngol 1965; 81: 29-33. Buchman CA, Doyle WJ, Skoner DP, Post JC, Alper CM, Seroky JT, Anderson K, Preston RA, Hayden FG, Fireman P, Ehrlich GD. Influenza A Virus - induced acute otitis media. J Infect Dis 1995; 172: 1348-51. Chang MJ, Rodriguez WJ, Mohla C. Chlamydia trachomatis in otitis media in children. Pediatr Infect Dis 1982; l: 95-97. Chief medical officer in Greenland. Retningslinier for tilrettelæggelsen af profylaktisk arbejde under sundhedsvæsenet i Grønland. (In Danish). Cirkulære nr. 109, 1976. Christensen V. Boligforhold og børnesygelighed. (Thesis) (In Danish). Munksgaard, Copenhagen 1956. Counter SA, Klareskov B. Hypoacusis among the Polar Eskimos of Northwest Greenland. Scand J Audiol 1990; 19: 149-60. Cranz D. Historie von Grönland, enthaltend die Beschreibung des Landes und der Einwohner. Barby und Leipzig 1770: 751-53. Curtis T, Iburg KM, Bjerregaard P. Family and child health in Greenland. (In Danish). DIKEs Grønlandsskrifter, 1997; 9: 1-64. DiSarno NJ, Barringer C. Otitis media and academic achievement in Eskimo high school students. Folia Phoniat 1987; 39: 250-55. Daly KA, Brown JE, Lindgren BR, Meland MH, Le CT, Giebink GS. Epidemiology of otitis media onset by six months of age. Pediatrics 1999; 103: 1158-66. Doyle WJ. A functiono-anatomic description of eustachian tube vector relations in four ethnic populations: an osteologic study. (Dissertation, microfilm). Ann Arbor: University of Michigan, 1977: 1-186. Duncan B, Ey J, Holberg CJ, Wright AL, Martinez FD, Taussig LM. Exclusive breast-feeding for at least 4 months protects against otitis media. Pediatrics 1993; 91: 867-72. Duval L, MacDonald S, Lugtig L, Mollins J, Tate R. Otitis media in the Keewatin: 20 years of experience 1970-1991. Arct Med Res 1994; 53 (S2): 676-79.

48

Ey JL, Holberg CJ, Aldous MB, Wright AL, Martinez FD, Taussig LM, Group Health Medical Associates. Passive smoke exposure and otitis media in the first year of life. Pediatrics 1995; 95: 670-77. Faden H, Brodsky L, Waz MJ, Stanievich J, Bernstein JM, Ogra PL. Nasopharyngeal flora in the first three years of life in normal and otitis-prone children. Ann Otol Rhinol Laryngol 1991; 100: 612-15. Falck G, Engstrand I, Gnarpe J, Gnarpe H. Association of Chlamydia pneumoniae with otitis media in children. Scand J Infect Dis 1998; 30: 377-80. Fiellau-Nikolajsen M. Tympanometry and secretory otitis media. (Thesis) Acta Otolaryngol (Stockh) Suppl. 1983; 394: 1-73. Fischer PB, Ellermann-Eriksen S, Thiel S, Jensenius JC, Mogensen SC. Mannan-binding protein and bovine conglutinin mediate enhancement of herpes simplex virus type 2 infection in mice. Scand J Immunol 1994; 39: 439-45. Fleming DW, Cochi SL, Hightower AW, Broome CV. Childhood upper respiratory tract infections: to what degree is incidence affectedby day-care attendance? Pediatrics 1987; 79: 55-60. Freijd A, Bygdeman S, Rynnel-Dagoo B. The nasopharyngeal microflora of otitis-prone children, with emphasis on H. influenzae. Acta Otolaryngol (Stockh) 1984; 97: 117-26. Freijd A, Oxelius V, Rynnel-Dagoo B. A prospective study demonstrating an association between serum IgG2 concentrations and susceptibility to otitis media in children. Scand J Infect Dis 1985; 17: 115-20. Froom J, Culpepper L. Otitis media in day-care children. J Fam Pract 1991; 32: 289-94. Garred P, Madsen HO, Kurtzhals JAL, Lamm LV, Thiel S, Hey AS, Svejgaard A. A diallelic polymorphism may explain variations of the blood concentration of Mannan-binding protein in Eskimos, but not in Black Africans. Eur J Immunogenet 1992; 19: 403-12. Garred P, Brygge K, Sørensen CH, Madsen HO, Thiel S, Svejgaard A. Mannan-binding protein-levels in serum and upper-airways secretions and frequency of genotypes in children with recurrence of otitis media. Clin Exp Immunol 1993; 94: 99-104. Garred P, Harboe M, Oettinger T, Koch C, Svejgaard A. Dual role of mannan-binding protein in infections: Another case of heterosis? Eur J Immunogenet 1994; 21: 125-31. Garred P, Madsen HO, Hofmann B, Svejgaard A. Increased frequency of homozygosity of abnormal mannan-binding-protein alleles in patients with suspected immunodeficiency. Lancet 1995; 346: 941-43. Giebink GS. The microbiology of otitis media. Pediatr Infect Dis J 1989; 8: S18-S20. Giles M, Asher I. Prevalence and natural history of otitis media with perforation in Maori school children. J Laryngol Otol 1991; 105: 257-60. Harker LA. Cholesteatoma: An incidence study. In: Cholesteatoma: First International Conference. McCabe BF, Sade J, Abramson M. Birmingham, Ala: Aesculapius Publishing CO. 1977; 308-9. Harsten G, Prellner K, Heldrup J, Kalm O, Kornfält R. Recurrent acute otitis media. A prospective study of children during the first three years of life. Acta Otolaryngol (Stockh) 1989; 107: 111-19. Hayman CR, Kester FE. Eye, ear, nose and throat infection in natives of Alaska. Northwest Med 1957; 56: 423-30. Henderson FW, Collier AM, Sanyal MA, Watkins JM, Fairclough DL, Clyde WA, Denny FW. A longitudinal study of respiratory viruses and bacteria in the etiology of acute otitis media with effusion. N Engl J Med 1982; 306: 1377-83. Howie VM, Ploussard JH, Sloyer J. The "otitis prone" condition. Am J Dis Child 1975; 129: 676-78. "


49

Ingvarsson L, Lundgren K, Ursing J. The bacterial flora in the nasopharynx in healthy children. Acta Otolaryngol (Stockh) 1982a; suppl 386: 94-96. Ingvarsson L, Lundgren K, Olofsson B, Wall S. Epidemiology of acute otitis media in children. Acta Otolaryngol (Stockh) 1982b; suppl 388: 1-52. Jerger J. Clinica1 experience with impedance audiometry. Arch Otolaryngol 1970; 92: 311-24. Johonnott SC. Differences in chronic otitis media between rural and urban Eskimo children. Clin Pediatr 1973; 12: 415-19. Kalm O, Johnson U, Prellner K, Ninn K. HLA antigens and recurrent acute otitis media. Acta Otolaryngol (Stockh) 1992; S492: 107-9. Kamme C, Lundgren K, Mårdh P. The aetiology of acute otitis media in children. Scand J Infect Dis 1971; 3: 217-23. Kaplan GJ, Fleshman JK, Bender TR, Baum C, Clark PS. Long-term effects of otitis media; a ten-year cohort study of Alaskan Eskimo children. Pediatrics 1973; 52: 577-85. . Karma P, Luotonen J, Pukander J, Sipilä M, Herva E, Grönroos P. Haemophilus influenzae in acute otitis media. Acta Otolaryngol (Stockh) 1983; 95: 105-10. Karma PH, Sipilä M, Pukander JS, Perälä ME. Occurrence of Cholesteatoma and Secretory. Otitis Media in Children. In: Cholesteatoma and Mastoid Surgery: Third Int Conf. Tos M, Thomsen J, Peitersen E. Kugler & Ghedini Publications, Amsterdam, Berkeley, Milano, 1989; 335-38. Kero P, Piekkala P. Factors affecting the occurrence of acute otitis media during the first year of life. Acta Paediatr Scand 1987; 76: 618-23. Klein BS, Dollete FR, Yolken RH. The role of respiratory syncytial virus and other viral pathogens in acute otitis media. J Pediatr 1982; 101: 16-20. Klein JO, Tos M, Casselbrant ML, Daly KA, Hoffman HJ, Ingvarsson L, Karma P, Van Cauwenberge PB. Epidemiology and natural history. In: Lim DJ, ed. Recent advances in otitis media. Report of the Fifth Research Conference. Ann Otol Rhinol Laryngol 1994; 103 (suppl 164): 9-12. Klein JO, Tos M, Casselbrant ML, Daly KA, Karma P, Van Cauwenberge PB, Ingvarsson L. Epidemiology and natural history. In: Lim DJ, ed.. Recent advances in otitis media. Report of the Sixth Research Conference. Ann Otol Rhinol Laryngol 1998; 107 (suppl 174): 9-13. Koch A, Mølbak K, Homøe P, Bretlau P, Melbye M. Respiratory tract infections in Geenlandic children: A prospective cohort study. Int J Circumpolar Health 1998; 57 (SI): 252-54. Kramer AH, McCullough DW. The prevalence of otitis media with effusion among Inuit children. Int J Circumpolar Health 1998; 57 (SI): 265-67. Kuklinska D, Kilian M. Relative proportions of Haemophilus species in the throat of healthy children and adults. Eur J Clin Microbiol 1984; 3: 249-52. Kunimoto M, Tamura S, Tabata T, Yoshie O. One-step typing of Epstein-Barr virus by polymerase chain reaction: Predominance of type l in Japan. J Gen Virol 1992; 73: 445-61. Kværner KJ, Tambs K, Harris JR, Magnus P. Distribution and heritability of recurrent ear infections. Ann Otol Rhinol Laryngol 1997a; 106: 624-32. Kværner KJ, Nafstad P, Hagen JA, Mair IWS, Jaakkola HK. Recurrent acute otitis media: the significance of age at onset. Acta Otolaryngol (Stockh) 1997b; 117: 578-84. Leach AJ, Boswell JB, Asche V, Nienhuys TG, Mathews JD. Bacterial colonization of the nasopharynx predicts very early onset and persistence of otitis media in Australian Aboriginal infants. Pediatr Infect Dis J 1994; 13: 983-89. Lee STS. Cholesteatoma in an Asian Population. Acta Otolaryngol (Stockh) 1991; 111: 536 41. Lehmann D. Epidemiology of acute respiratory tract infections, especially those due to 50

Haemophilus infIuenzae, in Papua New Guinean children. J Infect Dis 1992; 165 (suppl 1): S20-S25. Lindberg K, Rynnel-Dagöö B, Sundqvist KG. Cytokines in nasopharyngeal secretions; evidence for defective IL-1∀ production in children with recurrent episodes of acute otitis media. Clin Exp Immunol 1994; 97 (3): 396-402. Ling D, McCoy RH, Levinson ED. The Incidence of Middle Ear Disease and its educational Implications Among Baffin Island Eskimo Children. Can. J. Publ. Health 1969; 60: 38590. Ling D, Katsarkas A, Baxter JD. Ear disease and hearing los s among eskimo elementary school children. Can J Publ Health 1974; 65: 37-40. Littauer J, Elm M. Skoleaudiometri og otitis i Umanak Distrikt. (In Danish). Ugeskr Laeger 1964; 126: 1363-65. Long SS, Henretig FM, Teter MJ, McGowan KL. Nasopharyngeal flora and acute otitis media. Infect Immun 1983; 41: 987-91. Lous J, Fiellau-Nikolajsen M. Epidemiology of middle ear effusion and tubal dysfunction. A one-year prospective cohort study comprising monthly tympanometry in 387 nonselected 7-year-old children. Int J Pediatr Otorhinolaryngol 1981; 3: 303-17. Lund E, Henrichsen J. Laboratory diagnosis, serology and epidemiology of Streptococcus pneumoniae. In: Bergan T. Norris JR, eds. Methods in Microbiology. Academic Press, London, New York & San Fransisco 1978; 12: 241-62. Lupin AJ. Chronic otitis media in the Keewatin district. In: Shephard RJ, Hoh S, eds. Proc III Int Symp Circumpolar Health. Toronto and Buffalo, Univ Toronto Press, 1976: 389-98. Madsen HO, Garred P, Kurtzhals JAL, Lamm LU, Ryder LP, Thiel S, Svejgaard A. A new frequent allele is the missing link in the structural polymorphism of the human mannan binding protein. Immunogenetics 1994; 40: 37-40. Madsen HO, Garred P, Thiel S, Kurtzhals JAL, Lamm LU, Ryder LP, Svejgaard A. Interplay between promotor and structural gene variants control basal serum level of mannan .. binding protein. J Immunol 1995; 155: 3013-20. Madsen M, Lindahl A, Osler M, Bjerregaard P. Børns sundhed ved skolestart 1988/89. (In Danish) DIKE, Copenhagen 1991. Malver O. Indtryk fra en rejse som ørelæge i Vestgrønland sommeren 1956. (In Danish). Annual report to the Chief Medical Officer in Greenland 1956: 1-5. Manning P, Avery ME, Ross A. Purulent Otitis Media: Differences between populations in different environments. Pediatrics 1974; 53: 135-36. Marx J, Osguthorpe ID, Parsons G. Day care and the incidence of otitis media in young children. Arch Otolaryngol Head Neck Surg 1995; 112: 695-99. Maw AR, Smith 1M, Lance GN. Lateral cephalometric analysis of children with otitis media with effusion: a comparison with age and sex matched controis. J Laryngo1 Otol 1991; 105: 71-77. Maynard JE. Otitis media in Alaskan Eskimo children: an epidemiologic review with observations on control. Alaska Medicine1969; 11: 93-99. McCullough D. Tympanoplasty results in the Canadian Arctic 1970-1980. In: Proceedings of 5th International Symposium on Circumpolar Health, Copenhagen 1981. Ed. Harvald B and Hansen JPH. Nordic Council for Arctic Medical Research 1982; 33: 364-65. McCullough D. Chronic otitis media in the Keewatin area of the Northwest Territories. J Otolaryngol1990; 19: 389-90. Melbye M, Ebbesen P, Levine PH, Bennike T. Early primary infection and high Epstein Barr virus antibody titers in Greenland Eskimos at high risk for nasopharyngeal carcinoma. Int J Cancer 1984; 34: 619-23.

.


51

Meldorf G. Epidemiske sygdomme i Grønland: Influenza og epidemiske catarrhalske affektioner af luftvejs-slimhinderne. (In Danish). Meddr Grønland 1907; 33: 129-305. Melton V. Undersøgelser af Øre-, næse-, og halspatienter i Vestgrønland. (In Danish). Ugeskr Laeger 1955; 117: 846-48. Middaugh J, Ryan C, Metzler C, Dixon CL. Otitis media and hearing impairment among selected racial groups in Alaska. Epidemiology Office, Division of Public Health, Department of Health and Social Services, State of Alaska, 1993. Moffatt MEK, Tate R, Young TK, O'Neil JO, Woods W, Gillespie I. Hearing loss in adults in the Keewatin Region, NWT. Arct Med Res 1994; 53 (S2): 685-87. . Niclasen BVL, Møller SM, Christensen RB. Drug prescription to children living in the Arctic: An investigation from Nuuk, Greenland. Arct Med Res 1995; 54; Suppl I: 95-100 Pedersen CB. Incidence of chronic otitis media with cholesteatoma in children. Danish Med Bulletin 1983; 30: 204. Pedersen CB, Zachau-Christiansen B. Otitis media in Greenland children; acute, chronic and secretory otitis media in three- to eight-year-olds. J Otolaryngol 1986; 6: 332-5. alias Pedersen CB, Zachau-Christiansen B. Mellemørebetændelse hos grønlandske børn. (In Danish). Ugeskr Laeger 1986; 148: 3129-31. Pedersen CB, Zachau-Christiansen B. Chronic otitis media and sequelae in the population of Greenland. Scand J Soc Med 1988; 16: 15-19. alias Pedersen CB, Zachau-Christiansen B. Kronisk mellemørebetændelse og følger deraf i den grønlandske befolkning. (In Danish). Ugeskr Laeger 1986; 148: 3126-9 Poulsen IPV. Beretning om en otologisk konsultationsrejse foretaget i tiden 26/7 - 1/9 1961 til Thule distriktet og den nordlige del af den grønlandske vestkyst fra Upernavik til og med Egedesminde lægedistrikt. (In Danish). Annual report to the Chief Medical Officer in Greenland 1962: 1-6. Prellner K, Christensen P, Hovelius B, Rosen C. Nasopharyngeal carriage of bacteria in otitisprone and non-otitis-prone children in day-care centres. Acta Otolaryngol (Stockh) 1984; 98: 343-50. Prellner K, Kalm O, Harsten G, Heldrup J, Oxelius VA. Pneumococcal serum antibody concentrations during the first three yearS of life: a study of otitis-prone and non-otitisprone children. Int J Pediatr Otorhinolaryngol1989; 17: 267-79. Prellner K, Bohm J, Bretlau P, Hallberg T. Genetic factors as related to chronic otitis media in Polar Eskimos. In: Lim D, Bluestone CD, Klein J, Nelson ID, Ogra PL, eds. Proceedings of the Fifth International Symposium on Recent Advances in Otitis Media. Ft. Lauderdale 1991. New York: Decker Periodicals, 1993: 36-38. Ratnesar P. Chronic ear disease along the coasts of Labrador and Northern Newfoundland. J Otolaryngol1976; 5: 122-30. Reed D, Struve S, Maynard JE. Otitis media and hearing deficiency among Eskimo children: A cohort study. Am J Publ Health 1967; 9: 1657-62. Reed D, Dunn W. Epidemiologic studies of otitis media among Eskimo children. Publ Health Rep 1970; 85: 699-706. Reinhard KR, Huntley BE, Becker RA, Philip RN, Jackson H. Bacteriological studies on exudative otitis media occurring in six communities of Alaskan natives. Acta Otolaryngol (Stockh) 1970; suppl 260: 1-36. Richardson VF, Larcher VF, Price JF. A common congenital immunodeficiency predisposing to infection and atopy in infancy. Arch Dis Child 1983; 58: 799-802. 52

Rosenfeld RM, Moura RL, Bluestone CD. Predictors of Residual-Recurrent Cholesteatoma in Children. Arch Otolaryngol Head Neck Surg 1992; 118: 384-91. Ruuskanen O, Arola M, Heikkinen T, Ziegler T. Viruses in acute otitis media: increasing evidence for c1inical significance. Pediatr Infect Dis J 1991; 10: 425-27. Røjskjær C, Andersen M. Grønlands Audiologi. Nordisk Audiologi 1963; 12: 99-135. Røjskjær C. Audiology in Greenland: an audiological program in a remote area. Audiology 1974; 13: 408-13. . Saarinen DM. Prolonged breast feeding as prophylaxis for recurrent otitis media. Acta Paediatr Scand 1982; 71: 567-71. Sandvej K, Peh SC, Andresen B, Pallesen G. Identification of potential hot-spots in the carboxy-terminal part of the Epstein-Barr virus (EBV) BNLF-1 gene in both malignant and benign EBV associated diseases: High frequency of a 30 bp deletion in Malaysian and Danish peripheral T-cell lymphomas. Blood 1994, 84: 4053-60. Sarkkinen HK, Halonen PE, Arstila PP, Salmi AA. Detection of Respiratory Syncytial, Parainfluenza Type 2, and Adenovirus antigens by radioimmunoassay and enzyme immunoassay on nasopharyngeal specimens from children with acute respiratory disease. J Clin Microbiol 1981; 13: 258-65. Sarvas H, Rautonan N, Sipinen S, Makela O. IgG subclasses of pneumococcal antibodies effect of allotype G2m(n). Scand J Immunol 1989; 29: 229-37. Schaefer O. Otitis media and bottle-feeding. Can J Publ Health 1971; 62: 478-89. Schwartz R, Rodriguez WJ, Mann R, Khan W, Ross S. The nasopharyngeal culture in acute otitis media: a reappraisal of its usefulness. JAMA 1979; 241: 2170-73. Shapiro RS. Otologic findings in an Inuit population of cleft palate children. J Otolaryngol 1988; 17: 101-2. Smidt CM. Otolaryngology and audiology in Greenland with a short review of Greenland medical history. In: Harvald B, Hansen JPH, eds. Circumpolar Health 1981. Proceedings of the 5th International Symposium on Circumpolar Health. Nordic Council for Arctic Med Res Report 1982; 33 (SI): 133-36. Sorri M, Alho O-P, Oja H. Dynamic multivariate modelling: day care and consultation rate for acute otitis media. Acta Otolaryngol (Stockh) 1996; 116: 299-301. Stenström C, Bylander-Groth A, Ingvarsson L. Eustachian tube function in otitis-prone and healthy children. Int J Pediatr Otorhinolaryngol 1991; 21: 127-38. Stenström C, Ingvarsson L. Otitis-prone children and controls: a study of possible predisposing factors. 1. Heredity, family background and perinatal period. Acta Otolaryngol (Stockh) 1997a; 117: 87-93. Stenström C, Ingvarsson L. Otitis-prone children and controls: a study of possible predisposing factors. 2. Physical findings, frequency of illness, allergy, day care and parental smoking. Acta Otolaryngol (Stockh) 1997b; 117: 696-703. Stewart JL. Otitis media in the first year of life in two Eskimo communities. Ann Otol Rhinol Laryngol 1989; 98: 200-1. Storgaard M, 0stergaard L, Jensen JS, Farholt S, Larsen K, Ovesen T, Nødgaard H, Andersen PL. Chlamydia pneumoniae in children with otitis media. Clin Infect Dis 1997; 25: 1090-93. Strangert K. Otitis media in young children in different types of day-care. Scand J Infect Dis 1977; 9: 119-23. Stahlberg M, Ruuskanen O, Virolainen E. Risk factors for recurrent otitis media. Pediatr Infect Dis J 1986; 5: 30-32. Summerfield JA, Sumiya M, Levin M, Turner MW. Association of mutations in mannan binding protein gene with childhood infection in consecutive hospital series. BMJ 1997;


53

314: 1229-32. Sundermnan J, Dyer H. Chronic ear disease in Australian Aborigines. Med J Aust 1984; 140: 708-11. Sung RYT, Ling JM. Fung SM, Oppenheimer SJ, Crook DW, Lau JTF, Cheng AFB. Carriage of Haemophilus influenzae and Streptococcus pneumoniae in healthy Chinese and Vietnamese children in Hong Kong. Acta Pædiatr 1995; 84: 1262-67. Super M, Thiel S, Lu J, Lewinsky RJ, Turner MW. Association of low levels of mannan binding protein with a common defect in opsonisation. Lancet 1989; 2: 1236-39. Sørensen CH, Andersen LP, Tos M, Thomsen J, Holm-Jensen S. Nasopharyngeal bacteriology and secretory otitis media in young children. Acta Otolaryngol (Stockh) 1988; 105: 126-31. Sørensen CH, Bille M, Bretlau P, Rygaard J. IgD-cells in nasopharyngeal tonsils from Inuit children with recurrent upper-airway infections. Unpublished manuscript. Teele DW, Klein JO, Rosner B. Epidemiology of otitis media during the first seven years of life in children in Greater Boston: a prospective cohort study. J Infect Dis 1989; 160: 83 94. The McGrath Project: report of a demonstration - study on the prevention of upper respiratory disease. In: Duncan DRL, Lomprey N, eds.. State of Alaska, Department of Health and Welfare, Division of Public Health and the Children's Bureau, U.S. Dept. of HEW. 1962: 1-83. Timmermans FJW, Gerson S. Chronic granulomatous otitis media in bottle-fed Inuit children. Can Med Assoc J 1980; 122: 545-47. Todd N, Bowman C. Otitis media at Canyon Day, Arizona. Arch Otolaryngol1985; 111: 606 8. Tos M. Epidemiology and spontaneous improvement of secretory otitis. Acta Otolaryngol (Belg) 1983; 37: 31-43. Tos M. Incidence, Etiology and Pathogenesis of Cholesteatoma in Children. Adv Oto-Rhino Laryng 1988; 40: 110-17. Tower EA. Chronic otitis media in Alaskan Natives: 1954 - 1979. An historical perspective. Alaska Medicine 1979; 21: 48-52. Tower EA. Otitis media surgery - Long-term effects. In: Proceedings of 5th International Symposium on Circumpolar Health, Copenhagen 1981. Ed. Harvald B and Hansen JPH. Nordic Council for Arctic Medical Research 1982; 33: 360-63. Tschopp CF. Chronic Otitis Media and Cholesteatoma in Alaskan Native Children. In: Cholesteatoma: First International Conference. McCabe BF, Sade J, Abramson M. Birmingham, Ala: Aesculapius Publishing CO. 1977; 290-92. Turk DC. The pathogenecity of Haemophilus Influenzae. J Med Microbiol 1984; 18: 1-16. Uhari M, Mäntysaari K, Niemelä M. A meta-analytic review of the risk factors for acute otitis media. Clin Infect Dis 1996; 22: 1079-83. Vinther B, Elbrønd O, Pedersen CB. A population study of otitis media in childhood. Acta Otolaryngol (Stockh), 1979; suppl360: 135-37. Vinther B, Elbrønd O, Pedersen CB. Otitis Media in Childhood. Socio-medical aspects with special reference to day-care and housing conditions. Acta Otolaryngol (Stoekh) 1982; suppl 386: 121-23. . WHO 1992. International Statistical Classification of Diseases and Health Related Problems. Danish edition. Sundhedsstyrelsen og Munksgaard, Copenhagen 1993. Woods W, Moffatt MEK, Young TK, O'Neil J, Gillespie I. Hearing loss and otitis media in Keewatin children, NWT. Arct Med Res 1994; 53: (S2): 693-96.

54