OBJECTIVE. Susceptibility to otitis media results from complex interactions among genetic factors of the host, exposure to pathogens, and environmental influences. The objective of this study was to study the role of single-nucleotide polymorphisms of regulatory elements of proinflammatory cytokine genes tumor necrosis factor-α−308, interleukin-1β+3953, and interleukin-6−174, in susceptibility to recurrent otitis media in childhood.
METHODS. A total of 505 children (296 otitis media susceptible, 209 nonsusceptible as control) were enrolled at 2 sites (Texas and Kentucky). DNA of the children was studied for specific single-nucleotide polymorphisms by restriction fragment length polymorphism assay and confirmed by gene sequencing.
RESULTS. In the overall study group, tumor necrosis factor-α−308 and interleukin-6−174 heterozygous or homozygous polymorphisms (high cytokine-producing genotypes) were significantly associated with otitis media susceptibility. The same association was found in a match-paired subgroup of 384 subjects. In the overall study group, there was a significant step-wise increase in otitis media susceptibility with increasing number of concomitant polymorphic genotypes. Simultaneous combination of tumor necrosis factor-α−308 and interleukin-6−174 polymorphisms further increased the risk for otitis media susceptibility. These 2 polymorphic genotypes also were associated with the increased risk for tympanostomy tube placement. Children who had tumor necrosis factor-α−308 polymorphism and were breastfed for <1 month or exposed to cigarette smoke were more likely to be otitis media susceptible.
CONCLUSIONS. Our data suggest that tumor necrosis factor-α−308 and interleukin-6−174 polymorphisms are associated with increased risk for otitis media susceptibility and placement of tympanostomy tubes. Environmental factors such as breastfeeding may modify the risk for otitis media susceptibility in polymorphic individuals.
Up to 19% of children are susceptible to recurrent otitis media (OM).1–3 The long-term sequelae of chronic and recurrent OM include delayed language development and learning difficulty. The cause of OM susceptibility is multifactorial and includes environmental, microbial, and host factors. Because OM susceptibility occurs in family clusters, genetic factors of the host have been thought to play a major role.
The genetic susceptibility factors of OM likely are multigenic with influence over a wide array of host immune responses. There is considerable polymorphism of the genes that encode cytokines and related molecules. Many polymorphisms of cytokine genes have been reported to influence resistance to infection and susceptibility to various diseases. Single-nucleotide polymorphisms (SNPs) of regulatory elements of cytokine genes may allow for dysregulated production of cytokines, which in turn influence innate and acquired immunity.4 High cytokine-producing SNPs of acute phase, proinflammatory cytokines tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6 may be proposed to have a role in promoting inflammation during viral-bacterial interaction that occurs in the naso-tubo-tympanum in the process that leads to OM. Indeed, specific high cytokine-producing SNPs, namely, TNF-α−308, IL-1β+3953, and IL-6−174, have been shown to be associated with susceptibility to frequent or severe infections.5–9 However, the role of these cytokine SNPs in OM susceptibility has not been well studied. Other factors that are known to affect susceptibility to OM include age, male gender, family history of OM susceptibility, allergic diathesis, child care attendance, exposure to cigarette smoke, lack of breastfeeding, use of oral pacifiers, and supine bottle feeding.1–3,10–14 The aim of the present study was to evaluate the association of the TNF-α−308, IL-1β+3953, and IL-6−174 polymorphisms with OM susceptibility and to correlate the polymorphisms with other host and environmental factors that contribute to OM susceptibility.
Study Population and Disease Definition
Children were enrolled in the outpatient general pediatrics clinics and the pediatric otolaryngology clinic at University of Texas Medical Branch (Galveston, TX) and at the Kentucky Pediatric Research Office (Bardstown, KY). This study was approved by the University of Texas Medical Branch Institutional Review Board. Informed consent was obtained from the parents of participating children.
OM-susceptible (also known as otitis-prone) children met at least 1 of the following criteria: first episode of acute OM before the age of 6 months, 3 or more episodes of acute OM within a 6-month period, 4 or more episodes of acute OM within a 12-month period, 6 or more episodes by 6 years of age, or history of tympanostomy tube placement for recurrent or persistent OM. Non–OM-susceptible control children included those who had 0 to 2 episodes of acute OM by 2 years of age. Children were excluded from the study when they had an anatomic or a physiologic defect of the ear or nasopharynx, were known to have immunologic abnormality, or had major medical conditions or treatment for chronic conditions.
Clinical data were collected by interviewing the parents and reviewing medical charts. The data that were collected included gender, age, racial/ethnic background, history of episodes of acute OM, insertion of tympanostomy tubes, atopic diseases (asthma, eczema, hay fever, and food allergy), OM susceptibility in immediate family members, duration of breastfeeding, attendance at child care, and passive exposure to cigarette smoke in the household.
DNA was extracted from peripheral blood mononuclear cells or buccal epithelial cells of enrolled children. Polymerase chain reaction (PCR) was performed on the extracted DNA with the use of respective cytokine primer sets that spanned the SNP sites.6 The resultant PCR products were digested with polymorphic site-specific enzymes: NCo1 for TNF-α−308, Taq1 for IL-1β+3953, and Hsp92II for IL-6−174. All SNPs that were identified by PCR were confirmed by DNA sequencing. The laboratory staff who conducted the polymorphism assays were blinded for the OM-susceptible or nonsusceptible classification of the children. The study of SNPs yielded 3 genotypes for each cytokine: homozygous “normal” (low cytokine-producing) genotype, and homozygous and heterozygous polymorphic (high cytokine-producing) genotypes.
The study outcome variable of OM susceptibility was compared with genotype status with χ2 and Cochrane-Armitage trend tests using SAS statistical software (SAS Institute, Cary, NC). A logistic regression model was used to assess the predictive value of the 3 polymorphisms while controlling for individual- and study-level covariates and for additional posthoc analyses. All P values were considered significant at .05 with power of 80% to detect the difference.
The characteristics of enrolled children are summarized in Table 1. The overall study population (n = 505) was not balanced with respect to numbers of OM-susceptible and nonsusceptible control children (58.6% vs 41.4%, respectively). However, the distribution of race/ethnicity, age, and gender was not significantly different between the 2 groups (Table 1). After enrollment of the first 121 children for pilot study, the subsequent 384 children were enrolled by specifically matching the OM-susceptible and nonsusceptible children (192 children in each group) for race/ethnicity and gender at each study site (“match-paired” group). Most of the detailed analysis of the effect of polymorphic genotypes on OM susceptibility was performed on the match-paired groups to reduce the bias of race/ethnicity and gender, which may contribute to genetic diversity and OM susceptibility. The Texas site enrolled 72.0% and 84.9% of children in the overall and match-paired groups, respectively. The Kentucky site enrolled the remainder of the children.
SNPs in OM-Susceptible and Nonsusceptible Children
In the overall study group, the TNF-α−308, IL-1β+3953, and IL-6−174 heterozygous or homozygous polymorphic genotypes were found in 25.6%, 27.9%, and 39.0% children, respectively (data not shown). The homozygous polymorphic genotypes were not analyzed separately because of the relatively small number of children with this genotype. In the overall study group, the TNF-α−308 and IL-6−174 polymorphisms were associated with a significantly increased risk for OM susceptibility (odds ratio [OR]: 1.7 [P = .01] and 1.76 [P < .01], respectively; data not shown). Analysis of the match-paired group also showed a similar risk for OM susceptibility with TNF-α−308 and IL-6−174 polymorphisms (OR: 1.6 and 1.57, respectively; Table 2). All subsequent data analysis, therefore, was performed on the match-paired group.
Significantly more OM-susceptible boys (62%) were enrolled (and therefore the same number of male children in the matched control subjects). The distribution of polymorphic genotypes was equivalent among boys and girls (data not shown). Among the major racial/ethnic groups, white, black and Hispanic groups represented 41%, 29%, and 28%, respectively. These groups reflected largely the overall distribution in the general population at the 2 study sites. Use of logistic regression models showed that there was no interaction between race/ethnicity and polymorphic genotypes (P > .33 for all genotypes). However, posthoc subgroup analysis showed that among white children, the IL-6−174 polymorphism was significantly associated with OM susceptibility than nonsusceptibility (OR: 2.05; P = .03), with also a similar trend among black children (OR: 2.78; P = .07) but not among Hispanic children (OR: 0.86; P = .69; data not shown). The distribution of the TNF-α−308 and IL-1β+3953 polymorphisms was not significantly associated with OM susceptibility with respect to specific ethnic/racial groups.
History of tympanostomy tube placement was defined as increased susceptibility to OM. Fifty-three (27.6%) OM-susceptible children had a history of tympanostomy tube placement (Table 3). The rates of the TNF-α−308 and IL-6−174 genotypes were significantly higher in a step-wise manner from nonsusceptible, OM-susceptible without tubes, and OM-susceptible with tubes.
Effect of Combinations of SNPs
In the match-paired group, 135 (35.2%) children had a “normal” homozygous genotype for all 3 cytokines (Table 4). A total of 175 (45.6%) children had a polymorphic genotype for at least 1 cytokine, 63 (16.4%) had combinations of polymorphic genotypes for any 2 cytokines, and 12 (3.1%) children had a combination of polymorphic genotypes for all 3 cytokines. In this match-paired group, there was no significantly increased step-wise risk for OM susceptibility among the children with 1, 2, or 3 combinations of the polymorphic genotypes (P = .09). However, analysis of the overall study group showed a significant step-wise increased risk for OM susceptibility with the increased combinations of polymorphic genotypes (P < .01).
Among the combinations of polymorphic genotypes for any 2 cytokines, only the combination of the TNF-α−308 and IL-6−174 polymorphisms was associated with increased susceptibility for OM (OR: 2.84) compared with that with the TNF-α−308 or the IL-6−174 polymorphism alone (OR: 1.6 and 1.57, respectively) in the match-paired group. The same association was seen with the overall group as well. It is interesting that when children with either the TNF-α−308 or the IL-6−174 polymorphism had a combination of the IL-1β+3953 polymorphism, the risk for OM susceptibility was not significant, suggesting a possible moderating effect of IL-1β+3953 polymorphism on OM susceptibility.
Independent Analysis of All Variables
Each study variable was analyzed for its ability to predict OM susceptibility by using multivariate logistic regression analysis (Table 5). TNF-α−308, IL-6−174, children’s history of atopic diseases, and family history of OM susceptibility were independently associated with increased risk for OM susceptibility. However, history of breastfeeding was independently associated with significantly decreased risk for OM susceptibility. The reduction in risk was greater with longer duration of breastfeeding (4 vs 6 months: OR: 0.74 vs 0.64, respectively). Gender, ethnicity/race, exposure to cigarette smoke, attendance at child care, and enrollment study site were not independently associated with OM susceptibility.
Additional Posthoc Analysis of Effect of Breastfeeding and Cigarette Smoke Exposure
There was no association between children’s history of atopic diseases or family history of OM susceptibility with polymorphic alleles of any of the 3 cytokines in the study children (data not shown). A total of 122 (31.8%) children were exposed to passive cigarette smoke. In this population, cigarette smoke exposure by itself was not associated with increased risk for OM susceptibility (32.3% and 31.3% exposure among OM-susceptible and nonsusceptible children, respectively). A logistic regression model showed that there was a marginal interaction between TNF-α−308, IL-1β+3953, and IL-6−174 polymorphisms and cigarette smoke exposure to predict OM susceptibility (P = .11, .12, and .52, respectively; data not shown). However, on additional posthoc subgroup analysis, it was shown that children who had the TNF-α−308 polymorphism and were exposed to cigarette smoke were more likely to be OM susceptible than nonsusceptible (OR: 2.9; P = .02). However, among children who were not exposed to cigarette smoke, the TNF-α−308 polymorphism was not associated with OM susceptibility (OR: 1.21; P = .51). Unlike the TNF-α−308 polymorphism, children who had the IL-6−174 polymorphism and were not exposed to cigarette smoke were more likely to be OM susceptible than nonsusceptible (OR: 1.74; P = .04) but not among those who were exposed to cigarette smoke (OR: 1.31; P = .46). It is interesting that children who were exposed to cigarette smoke and had the IL-1β+3953 genotype were more likely to be nonsusceptible than OM susceptible (OR: 0.45; P = .05).
A total of 265 (69%) children had been either not breastfed or breastfed for <1 month. A logistic regression model showed that the risk for OM susceptibility was lower among breastfed children in both polymorphic and nonpolymorphic genotypes for all 3 cytokines (data not shown). Additional posthoc subgroup analysis showed that children who had the TNF-α−308 polymorphism and were not breastfed or were breastfed for <1 month were at higher risk for being OM susceptible than nonsusceptible (OR: 2.19; P = .01).
IL-1β, IL-6, and TNF-α are acute-phase cytokines that promote acute inflammatory response after infection with viruses and bacteria. Our study is the first to show that TNF-α−308 and IL-6−174 polymorphic genotypes are independently associated with the risk for OM susceptibility. The role of these 2 cytokine genes was strengthened further by the observation that there was a significant step-wise increase in OM susceptibility with increasing number of concomitant polymorphic genotypes. Furthermore, simultaneous presence of the TNF-α−308 and IL-6−174 polymorphic genotypes added to the risk for OM susceptibility and for placement of tympanostomy tubes, a marker of more persistent or recurrent OM.
The role of TNF-α−308 and IL-6−174 polymorphisms in OM susceptibility is consistent with their role in other infectious diseases. The most widely studied polymorphism is the TNF-α−308 allele, which has been shown to be associated with susceptibility to or increased severity of infectious diseases such as cerebral malaria,5 septic shock,6 and periodontitis.7 TNF-α−174AA/AG polymorphism is associated with severe bacterial sepsis in neonates.8 Therefore, the association between TNF-α−308 and IL-6−174 genotypes and OM susceptibility that was shown in the present study likely represents the overall role of these polymorphisms in predisposition to increased susceptibility to many infectious diseases and not necessarily to OM alone. Additional studies should be performed to confirm our results in other populations.
The polymorphism of cytokine genes was not the only factor that was associated with OM susceptibility, as OM is a multifactorial disease that involves complex interactions among environmental factors, infectious agents, and the host genetic factors. Numerous risk factors for OM development have been identified. Environmental factors include passive exposure to cigarette smoke,12,13 child care attendance,1,2,11 lack of breastfeeding,12,13,15 and the use of oral pacifiers.16 Our study suggests that breastfeeding itself is a factor that protects against OM susceptibility even when children have TNF-α−308 and IL-6−174 polymorphisms. Breastfeeding may protect against OM by providing passive protective factors in the oropharynx and the gut as well as by modifying the developing immune system of the infant that may render protection throughout life.17
In our study population, exposure to cigarette smoke was not an independent risk factor for OM susceptibility. However, in a subgroup analysis, we showed that exposure to cigarette smoke further increased the susceptibility for OM in children who had the TNF-α−308 polymorphism, whereas the IL-6−174 polymorphism increased the susceptibility for OM in unexposed children. Cigarette smoke is thought to contribute to airway inflammation through release of many inflammatory mediators, including TNF-α.18 Cigarette smoke could worsen respiratory tract inflammation during respiratory infection, particularly among people with the TNF-α−308 polymorphism.
Similar to other studies, our data show an association among atopy, family history of OM susceptibility, and the child’s increased susceptibility for OM.1,10,11,19–21 However, there was no association among atopy, family history of OM susceptibility, and the polymorphisms of cytokine genes. This observation suggests that OM susceptibility that is associated with atopy may involve other immunoregulatory genes. T helper lymphocyte (Th)1/Th2 cytokine imbalance has been speculated to be important in atopic individuals with OM susceptibility,22
The role of genetic factors in OM susceptibility has been recognized in studies of family clusters1 and in twins and triplets.23,24 Ramet et al25 studied a gene that controls surfactant protein-A, a lipid–protein complex that plays a role in innate host defense in the lung and also is expressed in eustachian tubes. They found that the frequency of specific surfactant protein-A haplotypes were found more frequently in OM-susceptible children. Straetemans et al26 found increased rates of Fcγ-RIIa-R/R131 polymorphisms in OM-susceptible Dutch children, but this was statistically not significant. Fcγ-RIIa is a leukocyte immunoglobulin G receptor that plays a role in immunoglobulin G–facilitated phagocytosis of bacteria. In a genome scan for susceptibility loci in OM-susceptible families, Daly et al27 found contribution of genes in 3 distinct chromosomal regions (10q, 19q, and 3p) with potential gene–gene interaction. Although these regions of linkage contain several putative positional candidate genes, the exact genes and encoded proteins that are involved in OM susceptibility have not been determined. The genes that encode TNF-α, IL-1β, and IL-6 are not in these chromosomes.
With respect to previous studies of cytokine gene polymorphisms in OM susceptibility, Joki-Erkella et al28 studied TNF-α−308, IL-1α−889, IL-1β−511, IL-1β+3953, and IL-1 receptor antagonist genes in 20 OM-susceptible Finnish families, comprising 63 individuals, and 400 random healthy blood donors as controls. In contrast to our findings, no association was seen between recurrent acute OM (AOM) and the polymorphisms of studied cytokine genes, including TNF-α−308 and IL-1β+3953. Their study design differed as follows. First, a smaller number of individuals were studied. Second, the subjects were genetically related; therefore, the focus was more on a heritable pattern of OM susceptibility. Because AOM is multifactorial, it is not surprising that a Mendelian-type inheritance was not seen. Third, the role of IL-6−174 was not investigated. Finally, the control subjects were recruited from adult blood bank donors whose childhood history of AOM was not obtained and who may have been OM-susceptible individuals. Because OM is a multifactorial disease, the genes of OM susceptibility may play different roles in different populations, particularly when comparing a homogeneous Finnish population with a heterogeneous multiracial population of the United States.
IL-1β has a role similar to TNF-α and IL-6 in acute inflammation. The IL-1β+3953 polymorphism is associated with an increased risk for advanced periodontitis.7 In our study, the IL-1β+3953 polymorphism does not increase the risk for OM susceptibility and seems to be protective even when children are exposed to cigarette smoke. Furthermore, the IL-1β+3953 polymorphism inhibits the OM-promoting actions of TNF-α−308 or IL-6−174 polymorphism when it is present in combination. We speculate that the IL-1β+3953 polymorphism leads to enhanced production of IL-1β, a Th1 cytokine, thereby preventing a Th2 response that may lead to atopy and OM susceptibility.22
It is unlikely that screening for these cytokine genes would be of much practical value in identifying the susceptible population in a routine clinical setting because of the relative commonality of these polymorphic cytokine genes in non–OM-prone individuals and because of the multifactorial cause of OM. Instead, research on the role of these cytokine genes may allow development of cytokine-modulating therapies that would prevent the development of OM after viral upper respiratory infection. Additional studies also can be designed to examine the mechanisms of immunity that is induced by vaccines against OM pathogens to prevent the OM-prone condition in a genetically susceptible population.
We have shown an association between the TNF-α−308 and IL-6−174 polymorphisms and OM susceptibility. There was an additional role of these polymorphic alleles in specific racial/ethnic groups and among children who were exposed to cigarette smoke. The IL-1β+3953 genotype is protective in children who are exposed to cigarette smoke and also counteracts the susceptibility that is conferred by the TNF-α−308 and IL-6−174 polymorphisms. Breastfeeding protected against OM susceptibility regardless of gene polymorphisms. In children who are genetically susceptible to OM, environmental modifications such as breastfeeding or reduction in cigarette smoke exposure may help to reduce OM occurrences; however, this requires additional investigations.
This work was supported by the National Institutes of Health grants R01 DC 5841 (to Dr Chonmaitree). The study was conducted at the General Clinical Research Center at the University of Texas Medical Branch at Galveston, funded by grant M01 RR 00073 from the National Center for Research Resources, National Institutes of Health.
We thank S. Ahmad, MD, A. Syal, MD, K.L. Najera, BS, M. Tran, BS, and M. Spalding, RN, for assistance with recruitment of study subjects and Ronald Deskin, MD, for facilitating subject recruitment from otolaryngology clinic. We also thank L. Yin, MD, W. Song, MD, and R. Serna, BS, for assistance in the laboratories and L. Zhang, MS, for statistical analysis of data.
- Accepted July 12, 2006.
- Address correspondence to Tasnee Chonmaitree, MD, Division of Pediatric Infectious Disease and Immunology, Department of Pediatrics, University of Texas Medical Branch, Galveston, TX 77555-0371. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵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
- ↵Teele DW, Klein JO, Chase C, Menyuk P, Rosner BA, the Greater Boston Otitis Media Study Group. Otitis media in infancy and intellectual ability, school achievement, speech, and language at age 7 years. J Infect Dis.1990;162 :685– 694
- ↵Knight J. Polymorphisms in tumor necrosis factor and other cytokines as risks for Infectious diseases and the septic syndrome. Cur Infect Dis Rep.2001;3 :427– 439
- ↵Harding D, Dhamrait S, Millar A, et al. Is interleukin-6–174 genotype associated with the development of septicemia in preterm infants? Pediatrics.2003;112 :800– 803
- ↵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– 1166
- ↵Duffy LC, Faden H, Wasielewski R, Wolf J, Krystofik D. Exclusive breastfeeding protects against bacterial colonization and day care exposure to otitis media. Pediatrics. 1997;100(4). Available at: www.pediatrics.org/cgi/content/full/100/4/e7
- ↵Ey JL, Holberg CJ, Aldous MB, Wright AL, Martinez FD, Taussig LM. Passive smoke exposure and otitis media in the first year of life. Pediatrics.1995;95 :670– 677
- ↵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– 872
- ↵Niemela M, Uhari M, Mottonen M. A pacifier increases the risk of recurrent acute otitis media in children in day care centers. Pediatrics.1995;96 :884– 888
- ↵Bernstein JM, Ballow M, Xiang S, O’Neil K. Th1/Th2 cytokine profiles in the nasopharyngeal lymphoid tissues of children with recurrent otitis media. Ann Otol Rhinol Laryngol.1998;107 :22– 27
- ↵Kvaerner KJ, Tambs K, Harris JR, Magnus P. Distribution and heritability of recurrent ear infections. Ann Otol Rhinol Laryngol.1997;106 :624– 632
- ↵Joki-Erkkila V, Puhakka H, Hurme M. Cytokine gene polymorphism in recurrent acute otitis media. Arch Otolaryngol Head Neck Surg.2002;28 :17– 20
- Copyright © 2006 by the American Academy of Pediatrics