Association of the FBXO11 Gene with Chronic Otitis Media with Effusion and Recurrent Otitis Media the Minnesota COME/ROM Family Study

ORIGINAL ARTICLE Association of the FBXO11 Gene With Chronic Otitis Media With Effusion and Recurrent Otitis Media The Minnesota COME/ROM Family Study

Fernando Segade, PhD; Kathleen A. Daly, PhD; Dax Allred, BA; Pamela J. Hicks, BA; Miranda Cox, MS; Mark Brown, MS; Rachel E. Hardisty-Hughes, PhD; Steve D. M. Brown, PhD; Stephen S. Rich, PhD; Donald W. Bowden, PhD

Objective: The FBXO11 gene is the human homo- analysis. In univariate genetic analysis, 1 reference SNP logue of the gene mutated in the novel deaf mouse mu- (hereinafter rs) (rs2134056) showed nominal evidence tant jeff (Jf), a single gene model of otitis media. We have of association to COME/ROM (P=.02), and 2 SNPs ap- evaluated single nucleotide polymorphisms (SNPs) in the proached significance (rs2020911, P=.06; rs3136367, FBXO11 gene for association with chronic otitis media P=.09). In multivariable analyses, including known risk with effusion/recurrent otitis media (COME/ROM). factors for COME/ROM (sex, exposure to smoking, attending day care centers, no prior breastfeeding, and hav- Design: A total of 13 SNPs were genotyped across the ing allergies), the evidence of independent association 98.7 kilobases of genomic DNA encompassing FBXO11. was reduced for each SNP (eg, rs2134056, from P=.02 Data were analyzed for single SNP association using gen- to P=.08). In subsequent analyses using the Pedigree Dis- eralized estimating equations, and haplotypes were evalu- equilibrium Test, the association of FBXO11 SNP ated using Pedigree Disequilibrium Test methods. rs2134056 (P=.06) with COME/ROM was confirmed. In- corporating multiple SNPs in 2- and 3-locus SNP hap- Patients: The Minnesota COME/ROM Family Study, a lotypes, those haplotypes containing rs2134056 also ex- group of 142 families (619 subjects) with multiple af- hibited evidence of association of FBXO11 and COME/ fected individuals with COME/ROM. ROM (P values ranging from .03 to .10).

Main Outcome Measures: Genetic association of Conclusion: We have observed evidence consistent with COME/ROM with polymorphisms in FBXO11. an association between polymorphisms in FBXO11, the human homologue of the Jeff mouse model gene, and Results: The FBXO11 SNPs are contained in a single link- COME/ROM. age disequilibrium haplotype block. Ten of the 13 SNPs were sufficiently polymorphic in the sample to permit Arch Otolaryngol Head Neck Surg. 2006;132:729-733

HRONIC OTITIS MEDIA WITH ducted a linkage study (the Minnesota Author Affiliations: effusion (COME) and re- COME/ROM Family Study13) that re- Departments of Internal current otitis media vealed evidence for linkage to COME and Medicine (Drs Segade and (ROM) are relatively com- ROM on chromosome 10q and suggestive Bowden), Public Heath Sciences mon conditions, affect- evidence for linkage on chromosome 19q. (Ms Cox, Mr Brown, and ing 10% to 30% of children, and the rates Several case-control studies have identi- Dr Rich), and Biochemistry C 1-4 (Mr Allred, Ms Hicks, and of occurrence seem to be increasing. fied relationships between candidate genes Dr Bowden), and Center for These conditions often result in hearing and COME/ROM, including HLA anti- 5 6,7 14,15 16,17 Human Genomics (Drs Segade loss and middle ear sequelae. Disease gen, cytokine genes, and genes in- and Bowden, Mr Allred, and rates are highest in the preschool years and volved in the ability to clear infectious Ms Hicks), Wake Forest decline as children get older.1 These con- agents implicated in OM.18,19 To date, how- University School of Medicine, ditions are often treated with tympanos- ever, few detailed molecular genetic analy- Winston-Salem, NC; tomy tubes to ventilate the middle ear; this ses of genetic associations between candi- Department of Otolaryngology is the most common surgical procedure per- date genes and COME/ROM have been and Otitis Media Research formed in children in an ambulatory set- reported. Center, University of Minnesota ting.8 Epidemiologic studies2,9,10 have shown The FBXO11 gene is an intriguing can- School of Medicine, Minneapolis, (Dr Daly); and that a family history of these conditions in- didate for association with COME/ROM. Medical Research Council, crease a child’s personal risk for COME/ The deaf mouse mutant jeff represents a 11,12 20 Mammalian Genetics Unit ROM, and twin studies have demon- single gene model of COME, character- (Drs Hardisty-Hughes and strated a high degree of heritability for ROM ized by the development of OM in patho- Brown), Harwell, England. and otitis media (OM) duration. We con- gen-free conditions in the absence of any

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 132, JULY 2006 WWW.ARCHOTO.COM 729

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 other inflammatory disease. Animals with the jeff gene this study. We chose SNPs to cover the entire FBXO11 ge- (hereinafter, Jeff animals) present with fluid and pus ac- nomic sequence, including the 5Ј promoter region, exons, in- cumulation in the middle ear cavity, with diffuse muco- trons, and the 3Ј-untranslated region, and included rs2020911, sal inflammation, disruption of the eustachian epithe- rs3136367, rs3136371, rs3771285, rs3732191, rs330787, lium, and reduction of the lumen owing to abnormal rs2651767, rs2134056, rs960106, rs2937345, rs874869, 20 rs4952896, and rs7582252 (where rs indicates 1 reference SNP). growths. The jeff mutation appears to be fully pen- Of these SNPs, rs7582252, rs2937345, and rs3136371 were not etrant and has been mapped to the distal region of mouse polymorphic and were not included in subsequent analyses. 20 chromosome 17. A missense mutation has been iden- We performed genotyping of the FBXO11 SNP polymor- tified in Fbxo11, a gene located on this region of mouse phisms by employing the MassARRAY SNP genotyping sys- chromosome 17 (R. E. Hardisty-Hughes and S. D. M. tem (Sequenom Inc, San Diego, Calif) using a standard pro- Brown, unpublished data, 2006). The human ortholog, tocol.23 Total genomic DNA was purified from whole blood FBXO11, is located on chromosome 2 and encodes a 141- samples obtained from the subjects using PUREGENE DNA iso- residue protein that belongs to the F-box family of genes. lation kit (Gentra Inc, Minneapolis, Minn). We quantitated DNA FBXO11 is characterized by the presence of a 50–amino using standardized fluorometric readings on a Hoefer DyNA acid sequence, the F-box, which functions in protein- Quant 200 fluorometer (Hoefer Pharmacia Biotech Inc, San Fran- 21 cisco, Calif). Each sample was diluted to a final concentration protein interactions. The FBXO11 gene therefore rep- of 5 ng/µL. resents a strong candidate gene for OM. This article ex- amines polymorphisms in the FBXO11 gene for association in the human disease pathway for COME/ROM. STATISTICAL GENETIC ANALYSES

METHODS We examined each pedigree for evidence of an incorrect family relationship by using genome scan data and PREST software.24 For each of the SNPs genotyped in this study, mende- PARTICIPANTS lian inconsistencies in their genotype assignment were examined using PEDCHECK software.25 Any genotypes inconsistent with All families and individuals in this study were recruited to par- 13 mendelian inheritance (9 total [0.001%]) were converted to miss- ticipate in the genetic linkage study. Families who were excluded ing. Following pedigree and genotype correction, maximum- from the linkage study because they lacked an affected sibling pair likelihood estimates of allele frequencies were computed us- with sufficient DNA were evaluated in this study. Study partici- ing the largest set of unrelated individuals. All SNPs were then pants included families from previous studies of OM conducted tested for departures from Hardy-Weinberg proportions. With by the Otitis Media Research Center at the University of Minne- multiple SNPs genotyped in the FBXO11 gene, the haplotype sota,Minneapolis,themembersofthegeneralpublicwhoresponded block structure and linkage disequilibrium (LD) between mark- to fliers posted around the university’s Academic Health Center, ers was assessed. Estimates of LD (DЈ and r2) were computed and otolaryngology clinic patients. Prospective families from pre- also using the largest set of unrelated individuals and Dprime vious studies were identified using OM history data collected in software. For families in which no founder was genotyped, the the original study on the participant (proband), parents, and sib- first affected offspring was used to estimate the statistics. lings. Mothers from the families recruited from the general pub- Association between individual SNP and COME/ROM sta- lic and clinic patients were interviewed about the OM history of tus was performed using a series of generalized estimating equa- all family members to determine family eligibility. Data were col- tion models.26 We adjusted for the correlation between sub- lected at the study visit to determine phenotype (history of chronic jects within a pedigree in the analyses by assuming exchangeable OM or ROM), including findings from the otomicroscopic exami- correlation among siblings within a pedigree and computing nation, multifrequency tympanometry, reported OM history, and 13 the sandwich estimator of the variance. The sandwich estima- medical record. An individual was considered to be affected if tor is also denoted the robust or empirical estimator of the vari- at least 2 of these sources had abnormal findings, or if 1 source ance and is robust to misspecification of the correlation ma- showed abnormalities and middle ear or tympanometric findings trix because it estimates the within-pedigree correlation matrix were presumed to be definitive evidence of a history of COME/ from the first and second moments of the data. Broadly speak- ROM (eg, tympanosclerosis, atrophy, or tympanogram with high ing, this method is comparable to a logistic regression analysis static admittance) were present. Subjects were examined by a neu- with adjustment for familial correlation among family mem- rotologist, and those with obvious craniofacial anomalies were ex- bers. As such, the unaffected members of the pedigrees pro- cluded. Subjects were not examined by a dysmorphologist. Ana- vide allele counts that correspond to control genotype frequen- tomic and morphologic traits and environmental factors in OM cies. For each SNP, the 2-df overall test of genotypic association pathogenesis are highly interrelated, which made it difficult to im- was performed, and, if the overall genotypic association was pose specific criteria for inclusion or exclusion. Exclusion crite- significant, 3 individual contrasts defined by the a priori ge- ria included Down syndrome, cleft palate, or adopted siblings. netic models (dominant, additive, and recessive) were com- A total of 142 families (619 individuals) participated in the puted. If the overall genotypic association was not significant, candidate gene study. This includes 132 families who were in- 13 the a priori contrasts were examined after adjusting for the 3 cluded in the genome scan and linkage studies and 10 fami- comparisons using a Bonferroni adjustment, consistent with lies who were recruited using the same ascertainment criteria the Fisher protected least significant difference multiple- and included only in this candidate gene study. DNA ex- comparison procedure. Tests of association between SNP and tracted from blood (94%) or buccal (6%) samples was used to COME/ROM were computed adjusting for sex, number of smok- assess candidate genes. ers in the household, prior breastfeeding, presence of allergies, and past day care center attendance. GENOTYPING Haplotype associations between the COME/ROM and the FBXO11 SNPs were also assessed. Haplotypes were con- Thirteen single nucleotide polymorphisms (SNPs) in FBXO11, structed and the analyses of the 2- and 3-marker haplotypes selected from the dbSNP22 public database, were genotyped in were completed using a generalized estimating equation analy-

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 132, JULY 2006 WWW.ARCHOTO.COM 730

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 sis as described herein, except the quasi-likelihood was weighted positive for COME/ROM, or definitive examination find- by the probability for each possible haplo-genotype for an in- ings), and 60% were unaffected. Most were white (94%) dividual. Each individual was entered in the generalized esti- and non-Hispanic (99%). Risk factors were prevalent, most mating equation analysis once for each haplo-genotype possi- participants had attended day care centers, about 20% bility, weighted by the haplo-genotype probability. Thus, the reported allergies, and exposure to smokers was com- weight for each individual sums to 1. The weighted generalized estimating equation analyses were completed as demon (Table 1). scribed herein, using the sandwich estimator of the variance to account for the within-cluster correlation. GENOTYPING AND LD STRUCTURE To test for allelic association while accounting for potential population stratification, the Pedigree Disequilibrium Test The FBXO11 gene consists of 23 exons spanning more (PDT) was employed.27 The PDT was performed using both than 98 kilobases (kb) of genomic sequence arranged in single SNPs as well as 2- and 3-marker haplotypes within a tail-to-tail configuration with the MSH6 guanine/ FBXO11. The PDT operates on the same principle as the trans- thymine mismatch repair gene, resulting in an overlap- mission disequilibrium test: alleles transmitted to cases create ping sequence of 32 base pairs in their 3Ј untranslated genotype frequencies for cases, whereas alleles that are not trans- regions (Figure).28 Thirteen SNPs were genotyped across mitted comprise allele frequencies for controls. The PDT is more the 98.7 kb of genomic DNA encompassing the FBXO11 powerful than the transmission disequilibrium test, and it al- gene in 619 individuals from 142 families in the Minne- lows analysis of transmission of alleles with all available family data. The PDT is a valid test of association even in the pres- sota COME/ROM Family Study. The Figure also shows ence of population substructure and it maintains the analysis the location and distribution of the SNPs relative to the of transmission within families.27 genomic structure of the gene. FBXO11 is a complex gene, and the 13 SNPs were chosen to systematically cover the genomic region with a density of 1 SNP per 7.6 kb. A large RESULTS gap of 31.5 kb exists between exons 2 to 15 in which no SNPs were identified that were polymorphic in the study PARTICIPANTS subjects. The current set of 13 SNPs is, however, contained in a single LD haplotype block. Only 1 of the SNPs A total of 142 families (619 individuals) participated in is in an exon (rs3136371), but this SNP is in the non- the candidate gene study. Of the participants in this study, coding 3Ј untranslated region. Bioinformatic analysis did 40% were classified as affected (2 or more sources of data not identify SNPs in other exons. After genotyping, all SNPs were evaluated for Hardy-Weinberg equilibrium and Table 1. Descriptive Data for the Minnesota COME/ROM pairwise LD. The SNPs with minor allele frequencies Family Study greater than 0.10 exhibited genotypic frequencies that were consistent with Hardy-Weinberg proportions. DЈ Variable Value values were greater than 0.95 for all SNPs with minor al- Pedigrees, No. 142 lele frequencies greater than 0.10, suggesting that the en- Total subjects, No. (% female) 619 (53) tire FBXO11 gene is encompassed by a single LD haplo- Smokers in the home, mean ± SD (median) 0.75 ± 0.85 (0.0) [618] type block (data not shown). [No. of subjects] Single SNP association analysis was performed using Exposed to smoking in household, No. (%) 618 (49) general estimating equation methods. Ten of the 13 SNPs Attending day care centers, No. (%) 617 (52) were sufficiently polymorphic in the sample (ie, they had Prior breastfeeding, No. (%) 591 (57) Ͼ Allergies, No. (%) 618 (22) minor allele frequencies 0.10) to permit analysis. In analysis of the COME/ROM phenotype summarized in Abbreviation: COME/ROM, chronic otitis media with effusion/recurrent Table 2, SNP rs2134056 (located in the distal portion otitis media. of intron 1; Figure) exhibited nominal evidence of asso-

FBXO11 rs874869 rs960106 rs330787 rs7582252 rs4952896 rs2937345 rs2134056 rs2651767 rs3732191 rs3771285 rs3136371 rs3136367 rs2020911

1 23 4 5 6 7 8-11 12 13 14 15 16 17 18 19 20 21 22 23 3.5 kb16.5 kb 22 kb 14 kb 5.7 kb 6 kb 2.8 kb 9 kb 4.5 kb 2 kb

10 9 8 7 6 5

MSH6

O kb 56 kb 63 kb 73 kb 75 kb 86 kb 89 kb 95 kb 100 kb 103 kb 106 kb

Figure. Genomic map of FBXO11. The gray boxes represent coding exons, and the ruler at the bottom shows the relative location and spacing of genotyped single nucleotide polymorphisms in kilobases (kb). FBXO11 and the neighboring MSH6 (homologue of MutS 6) genes are oriented in a tail-to-tail fashion.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 132, JULY 2006 WWW.ARCHOTO.COM 731

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 Table 2. Single Nucleotide Polymorphism (SNP) Association Analysis of FBXO11 SNPs With COME/ROM Using the Generalized Estimating Equation

P Value*

SNP rs No. MAF General Association, No Covariates General Association, Sex General Association, All Covariates† rs2020911 0.324 .055 .053 .14 rs3136367 0.261 .09 .07 .13 rs3771285 0.005 .25 .30 .36 rs3732191 0.065 .22 .32 .52 rs330787 0.381 .53 .54 .56 rs2651767 0.166 .52 .51 .27 rs2134056 0.231 .02 .02 .08 rs960106 0.402 .19 .20 .41 rs874869 0.443 .17 .20 .31 rs4952896 0.254 .13 .12 .16

Abbreviation: COME/ROM, chronic otitis media with effusion/recurrent otitis media; MAF, minor allele frequency; rs, reference SNP. *Boldface indicates PϽ.05; italics, .05ϽPϽ.10. †Analysis with all covariates: sex, exposure to smoking, attending day care centers, prior breastfeeding, and having allergies.

(P=.06). Haplotype analysis was performed using a slid- Table 3. Single Nucleotide Polymorphisms (SNPs) ing window approach, so that combinations of 2- and and Haplotype Analysis of FBXO11 SNPs 3-adjacent SNPs were tested sequentially. Consistent with Using the Pedigree Disequilibrium Test* the single SNP analysis, the 2- and 3-adjacent SNP combinations that include rs2134056 (eg, rs2651767 and SNP rs Number 1 Marker 2 Markers 3 Markers rs2134056; P=.03) reveal evidence for association with rs2020911 COME/ROM. rs3136367 rs3136371 rs3771285 COMMENT rs3732191 rs330787 FBXO11 is a member of the FBXO subfamily of proteins rs2651767 .03 that have an F-box but no recognized substrate-binding rs2134056 .06 .10 region.21 Although the specific function of the FBXO11 rs960106 .06 rs2937345 protein is unknown, a missense mutation in the mouse rs874869 Fbxo11 gene (R. E. Hardisty-Hughes and S. D. M. Brown, rs4952896 unpublished data, 2006) in the jeff mouse model of rs7582252 chronic proliferative OM makes human FBXO11 a candidate for involvement in COME/ROM. Evidence for mu- Abbreviation: rs, reference SNP. Ͻ tation of Fbxo11 leading to OM in the jeff mouse model *Bold face indicates P .05. Data are P values. illustrates the power of modern genetic methods to identify novel pathways that, based on biological understand- ciation to COME/ROM (P=.02). Two SNPs in FBXO11 ing alone, might not be implicated in the pathophysi- (rs2020911, P=.06; rs3136367, P=.09) approached sig- ologic traits of a disorder such as COME/ROM. nificance. Because these initial analyses did not adjust for Some comment as to the relevance of the jeff mouse recognized COME/ROM risk factors (covariates), we per- model to COME/ROM seems appropriate. For example, formed further analyses. Addition of sex as a covariate in susceptibility to OM in both the Jeff mouse and the study the analysis did not significantly change the evidence for subjects could be due to subtle craniofacial abnormali- association. In an analysis that included other recognized ties leading to eustachian tube dysfunction. Hearing loss risk factors for COME/ROM (sex, smoking exposure, at- in the Jeff mouse20 was of a mixed nature. The raised tending day care centers, prior breastfeeding, and having thresholds obtained for the cochlear nerve response were allergies), evidence of association was reduced for each SNP beyond what would be expected for a purely conductive (eg, rs2134056; significance reduced from P=.02 to P=.08). loss. Also, some Jeff mice showed an abnormally low en- docochlear potential indicating involvement of the stria PDT ANALYSIS vascularis in the hearing loss. Humans with middle ear disease often show sensorineural components to their In addition to single SNP associations, there is the pos- hearing loss, potentially as a consequence of damage to sibility that a combination of SNPs (ie, a haplotype) con- the lateral wall of the cochlea. Also, in other animal mod- fers increased risk for COME/ROM. We performed hap- els, cochlear disease has been seen in experimentally in- lotype association analysis with the SNP data using the duced OM, which is consistent with the raised endoco- PDT. Table 3 shows the results of analyses with PՅ.10. chlear potentials seen in Jeff mice.28,29 In single SNP analysis with a PDT, the rs2134056 SNP The craniofacial abnormality in Jeff mice was evident showed some evidence of association with COME/ROM on the C3H/HeH background (the laboratory strain on

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 132, JULY 2006 WWW.ARCHOTO.COM 732

©2006 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/29/2021 which the mutant was originally maintained, and the back- young children in relation to testing technique, age group, and the presence or ground used in previous studies).20 When on a different absence of middle-ear effusion. Ear Hear. 2003;24:38-47. 6. Schilder AGM. Zielhuis GA, Van Den Broek P. The otological profile of a cohort genetic background (C57BL/6), the penetrance of the ab- of Dutch 7.5-8 year olds. Clin Otolaryngol. 1993;18:48-54. normal face phenotype diminishes and is not always co- 7. Tos M, Stangerup SE, Holm-Jensen S, Sorenson CH. Spontaneous course incident with the presence of OM (R. E. Hardisty- of secretory otitis and changes of the eardrum. Arch Otolaryngol. 1984;110: Hughes and S. D. M. Brown, unpublished data, 2006). 281-289. For these reasons, OM in the jeff mouse model does not 8. Kozak LJ, Hall MJ, Pokras R, Lawrence L. Ambulatory Surgery in the United States, 1994: National Survey of Ambulatory Surgery. Hyattsville, Md: National Center for seem to be solely due to a structural abnormality that re- Health Statistics; 1997. Advance Data From Vital and Health Statistics, No. 283. sults in dysfunction of the eustachian tube. 9. Black N. The aetiology of glue ear: a case-control study. Int J Pediatr We have performed a systematic evaluation of FBXO11 Otorhinolaryngol. 1985;9:121-133. in a well-characterized population of COME/ROM fami- 10. Sipila M, Karma P, Pukander J, Timonen M, Kataja M. The Bayesian approach to the evaluation of risk factors in acute and recurrent acute otitis media. Acta lies from the COME/ROM Family Study. Multiple poly- Otolaryngol. 1988;106:94-101. morphisms have been genotyped in an effort to provide 11. Kvaerner KJ, Tambs K, Harris JR, Magnus P. Distribution and heritability of re- a comprehensive survey of gene variation in the gene. The current ear infections. Ann Otol Rhinol Laryngol. 1997;106:624-632. SNP genotyping data are consistent with the FBXO11 12. Casselbrant ML, Mandel EM, Fall PA, et al. Heritability of otitis media: a twin/ being contained in a single linkage disequilibrium hap- triplet study. JAMA. 1999;282:2125-2130. 13. Daly KA, Brown WM, Segade F, et al. Chronic and recurrent otitis media: a ge- lotype block. Univariate, multivariate, and haplotype nome scan for susceptibility loci. Am J Hum Genet. 2004;75:988-997. analysis using a variety of analytical approaches provide 14. Kalm O, Johnson U, Prellner K, Ninn K. HLA frequency in patients with recurrent evidence consistent with the genetic involvement of hu- acute otitis media. Arch Otolaryngol Head Neck Surg. 1991;117:1296-1299. man FBXO11 in COME/ROM. We note that the magni- 15. Barton JC, Bertoli LF, Acton RT. HLA-A and -B alleles and haplotypes in 240 in- tude of the significance for association is not dramatic, dex patients with common variable immunodeficiency and selective subclass de- ficiency in central Alabama. BMC Med Genet. 2003;4:3. which is consistent with the observation that the FBXO11 16. Joki-Erkkila¨ VP, Puhakka H, Hurme M. Cytokine gene polymorphism in recur- region was not detected in our previous genome scan for rent acute otitis media. Arch Otolaryngol Head Neck Surg. 2002;128:17-20. linkage. However, with a strong a priori hypothesis, these 17. Ra¨met M, Lo¨fgren J, Alho OP, Hallman M. Surfactant protein-A gene locus as- results suggest additional analyses of FBXO11 in other sociated with recurrent otitis media. J Pediatr. 2001;138:266-268. relevant populations. 18. Lee HY, Kang S, Ohcho S, et al. Investigation of the role of surfactant protein-D in the pathogenesis of otitis media. In: Lim DJ, Bluestone CD, Casselbrant M, Bakaletz LO, Giebink GS, Klein JO, Ogra PL, Bluestone MB, eds. Proceedings of Submitted for Publication: August 5, 2005; final revi- the Eighth International Symposium on Recent Advances in Otitis Media, Fort sion received March 8, 2006; accepted March 18, 2006. Lauderdale, Fla, June 3-7, 2003. Hamilton, Ontario: BC Decker; 2005:120. Correspondence: Donald W. Bowden, PhD, Center for 19. Roy S, Knox K, Segal S, et al. MBL genotype and risk of invasive pneumococcal disease: a case control study. Lancet. 2002;359:1569-1573. Human Genomics, Wake Forest University School of 20. Hardisty RE, Erven A, Logan K, et al. The deaf mouse mutant Jeff ( Jf)isasingle Medicine, Medical Center Boulevard, Winston-Salem, NC gene model of otitis media. J Assoc Res Otolaryngol. 2003;4:130-138. 27157 ([email protected]). 21. Kipreos ET, Pagano M. The F-box protein family. Genome Biol. 2000;1(5): Financial Disclosure: None reported. REVIEWS3002.1-REVIEWS3002.7. doi:10.1186/gb-2000-1-5-reviews3002. Funding/Support: This research was supported by the Accessed April 18, 2006. 22. Data Base of Single Nucleotide Polymorphisms. National Center for Biotechnol- National Institutes of Health grants NIDCD R01 DC03166 ogy Information homepage. www.ncbi.nlm/nih.gov/SNP. Accessed April 20, 2006. and NIH P30 DC04660 to Dr Daly. 23. Sutton BS, Langefeld CD, Williams AH, et al. Association of proopiomelanocor- tin gene polymorphisms with obesity in the IRAS Family Study. Obes Res. 2005; 13:1491-1498. REFERENCES 24. McPeek MS, Sun L. Statistical tests for detection of misspecified relationships by use of genome- screen data. Am J Hum Genet. 2000;66:1076-1094. 1. Daly KA. Definition and epidemiology of otitis media with effusion. In: Roberts 25. O’Connell JR, Weeks DE. PedCheck: a program for identification of genotype in- JE, Wallace I, Henderson F, eds. Otitis Media, Language, and Learning in Young compatibilities in linkage analysis. Am J Hum Genet. 1998;63:259-266. Children. Baltimore, Md: Paul H Brooks Publishing; 1997:3-41. 26. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous 2. Teele DW, Klein JO, Rosner B; Greater Boston Otitis Media Study Group. Epide- outcomes. Biometrics. 1986;42:121-130. miology of otitis media during the first seven years of life in children in greater 27. Monks SA, Kaplan NL, Weir BS. A comparative study of sibship tests of linkage Boston: a prospective cohort study. J Infect Dis. 1989;160:83-94. and/or association. Am J Hum Genet. 1998;63:1507-1516. 3. Daly KA, Hunter LL, Giebink GS. Chronic otitis media with effusion. Pediatr Rev. 28. Paparella MM. Interactive inner ear/middle ear disease, including perilymphatic 1999;20:85-93. fistula. Acta Otolaryngol Suppl. 1991;485:36-45. 4. Lanphear BP, Byrd RS, Auringer P, Hall CB. Increased prevalence of recurrent 29. Ichimiya I, Yoshida K, Hirano T, Suzuki M, Mogi G. Significance of spiral liga- otitis media among children in the United States. Pediatrics. 1997;99:E1. ment fibrocytes with cochlear inflammation. Int J Pediatr Otorhinolaryngol. 2000; 5. Sabo DL, Paradise JL, Kurs-Lasky M, Smith CG. Hearing levels in infants and 56:45-51.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 132, JULY 2006 WWW.ARCHOTO.COM 733