Status of the function of the ET is an important factor in the epidemiology of middle-ear disease.
The chapter includes
• The importance of otitis media as a major public health problem, which, in turn, stresses the need to understand the role played by the ET in the pathogenesis of this disease.
• The risk factors associated with otitis media that have been identified and are related to the ET in the pathogenesis of the disease, such as young age.
• The impact of otitis media on our healthcare budget and at the personal level for patients and their families.
The currently available data on the epidemiology of middle-ear disease is important to review since the Eustachian tube (ET) is involved in the pathogenesis of this disease and its complications and sequelae and therefore affects the prevalence, incidence, and natural history in specific age groups and differing populations. The epidemiology of middle-ear disease (e.g., acute otitis media, persistent middle-ear effusion, and otitis media with effusion), chronic suppurative otitis media, and cholesteatoma is presented in this context.
Middle-ear disease is a common cause of outpatient visits in the United States. In 2010, of the 132 million pediatric office visits, otitis media and ET disorders were the most common diagnosis, second only to routine infant or child health checkups.1,2 Infants and young children are most affected by this disease around the world.3–5 A recent study from Denmark (2014) reported two-thirds of preschool children had otitis media, despite reduction of known risk factors.6,7 The high prevalence of otitis media in infants emphasizes the importance in understanding the role that the ET plays in the pathogenesis of the disease in this highly susceptible age group. Studies revealed there was a dramatic increase in otitis media in children from the 1970s to the 1990s when the number of children in day-care facilities was on the rise.8–13 A study of Finnish children showed similar results, reporting increased rates of recurrent otitis media between 1978–1979 and 1994–1995, which paralleled a rise in the number of children in day-care facilities.3,10–13 A more recent study from Denmark reported in 2014, the rate of tympanostomy tube placement in children increased from 1997 to 2010, so that 3 in 10 children born in 2010 will undergo at least one insertion by their fifth birthday.14 Because exposure to upper respiratory tract infections increases with the number of contacts that infants and young children encounter13 and because their immune systems and the structure and function of their ET are immature, otitis media is now an all-too-frequent event (see Chapter 6, “Pathogenesis”).
There has been a recent impact of the pneumococcal conjugate vaccine in reducing the incidence of this most common pathogenic otic bacterium. As reported from Israel in 2015, PCV7 and PCV13 implementation was associated with rapid reduction of severe pneumococcal acute otitis media attacks.15 Another study from Spain by Ochoa-Gondar and colleagues (2015) also found that PCV7/13 had an aggregate effectiveness against vaccine-type pneumococcal infections was 72%.16 Although less common than in infants and young children, acute otitis media remains a health-care problem in older children, adolescents, and adults.10,17
Acute otitis media has been defined as the presence of a middle-ear effusion in which the signs and symptoms of acute infection are present, such as fever, irritability, and otalgia. Otitis media with effusion is characterized by the presence of relatively asymptomatic middle-ear effusion; however, some degree of hearing loss is usually present in both conditions. One study of 877 Boston children who were followed prospectively soon after birth found that 62% of them had acute otitis media by their first birthday and 17% had had more than three episodes. By age 3 years, 83% had had more than one episode and 46% had had more than three episodes.18 The incidence and prevalence of otitis media with effusion have also been found to be high in children. Of 103 children 2 to 6 years of age who were observed monthly over a 2-year period by otoscopy and tympanometry in a Pittsburgh day-care center, 53% had at least one episode of otitis media with effusion during the first year of the study and 61% during the second year; 30% had recurrent bouts, but 80% cleared within 2 months.19 In a similar study of 126 Pittsburgh schoolchildren 5 to 12 years of age, the incidence of otitis media with effusion was found to be much lower in children 6 years of age and older (Figure 2–1).20 Maturation of the structure (e.g., length) and function (i.e., active opening mechanism) of the ET and maturation of the immune system by age 6 years may be associated with the observed decrease in otitis media. Otitis media is also commonly diagnosed in adults. Finkelstein and colleagues reported on 167 consecutive adults who had otitis media with effusion in Israel between 1987 and 1990 and attributed most of these middle-ear effusions to the association of sinusitis.21
Risk Factors Related to the ET
The risk factors that are considered to be associated with otitis media may be related to ET dysfunction. Table 2–1 lists the factors that have been shown to be related to increased risk and others that are thought to be associated. Of the 17 possible risk factors listed, a case can be made for most of them being related, at least to some degree, to dysfunction of the tube (see Chapter 5, “Pathophysiology”). Recently Casselbrant and colleagues (2015) reported that comorbidities, such atopy, asthma, gastroesophageal reflux disease, and upper respiratory tract infections, and others, collected by history predicted the occurrence of chronic otitis media with effusion.22
Infants have a high rate of otitis media because the ET is shorter in infants than in older children and adults,23,24 they have difficulty in the active opening of the tube by swallowing,25–27 and they have immature immunity. Some infants who are “otitis prone” may have subtle evidence of a defect in immunity.28,29 Because the tube is short, aspiration, insufflation, or reflux nasopharyngeal secretions into the middle ear could be enhanced. Also, preterm infants have otitis media, and the lower the gestational age the higher the frequency, which is statistically significant in a report from 2014.30 Failure of the opening mechanism may lead to middle-ear underpressures with subsequent aspiration of organisms from the nasopharynx into the middle ear. As noted clinically during otoscopy and tympanometry, infants insufflate their middle ears during crying (related to the short floppy tube), which is most likely a physiologic compensatory mechanism to maintain middle-ear pressure. Even though proof is lacking, nasopharyngeal secretions are probably also insufflated into the middle ear during crying, especially when an upper respiratory tract infection occurs, which then results in otitis media.
An early age at onset has been associated with frequently recurrent disease in infants. In a study from Boston, Teele and colleagues reported that the peak incidence for first episodes of acute otitis media occurred at 6 months of age.18 Age at the first episode of acute otitis media was significantly and inversely associated with the risk of one or more or two or more episodes of acute otitis media in the 12 months after initial diagnosis. In a study from Norway, an association with allergy in the family appeared to be a predisposing factor for early onset.31 Universal infant hearing screening in The Netherlands identified risk factors using a parental questionnaire for selecting infants diagnosed or suspected of otitis media.32
Genetic Predisposition and Ethnic Groups
Now that data are available from two studies in twins that have demonstrated a genetic predisposition for the occurrence of otitis media, the hereditary defect could be related to structural or functional abnormalities of the ET (or immunity or both) in possibly all populations.31,33 Many patients are encountered in whom siblings, parents, and grandparents are affected. In a very comprehensive study, Casselbrant and colleagues recruited 168 like-sex twins and seven sets of triplets at birth in Pittsburgh and, for their first 2 years of life, followed them prospectively every month and whenever an upper respiratory tract infection intervened.33 The investigators, who were validated for the presence or absence of middle-ear effusion using otoscopy, clinically evaluated the subjects at each visit. Home visits were made when the children could not be brought to the clinical site (Children’s Hospital of Pittsburgh Otitis Media Research Center). Of the 143 sets of twins and triplets in whom the zygosity was known, 67 were dizygotic and 76 were monozygotic. Figure 2–2 shows the rate of otitis media in the dizygotic and monozygotic twins. An estimate of heritability for the average proportion of time with middle-ear effusion by 24 months of age was 75%. In a follow-up of these children, Casselbrant and colleagues reported that from ages 2 to 5 years, the cumulative effect remained significant for the entire first 5 years of life.34 The investigators concluded that a genetically determined predisposition for development of otitis media is likely.
In the other study, Kvaerner and colleagues gathered data from 2750 pairs of twins born in Norway between 1967 and 1974 by obtaining questionnaires related to lifetime prevalence of self-reported recurrent otitis media.31 They evaluated the rates in monozygotic and dizygotic twin pairs and the relative contribution of genes and environment to the variability in the predisposition to develop otitis media. Kvaerner and colleagues estimated additive genetic and dominance factors for otitis media at 74% in women, with the remaining 26% explained by environmental factors. For men, 45% was attributed to genetic factors and the remaining 55% about equally to either common familial or individual environmental factors.
Even though the study by Kvaerner and colleagues was neither prospective nor clinically assessed like the one by Casselbrant and colleagues,34 together these two reports provide convincing data to support our clinical impression that there is a strong influence of heredity in the susceptibility for otitis media. In a more recent longitudinal study of 1373 same-sex English twin pairs, Rovers and colleagues further confirmed the genetic predisposition for otitis media.35
In another, possibly related study, Todd and Todd reported on their evaluation of 115 nonsyndromic patients, aged 5 to 20 years, who had cardiac anomalies and reported a twofold higher prevalence of middle-ear pathology in individuals who had conotruncal cardiac malformations (transposition of the great vessels, tetralogy of Fallot, aortic stenosis) when compared with those patients with nonconotruncal anomalies (atrial septal defect, tricuspid atresia, atrioventricular canal).36 The investigators concluded that individuals who have conotruncal cardiac anomalies are otitis prone, and this may be related to a neural crest–determined branchial field defect, which may be related to a deletion in chromosome 22q11. Even though the methodology of this study has been criticized, the findings, if confirmed by a study that prospectively evaluates infants who have cardiac anomalies for the occurrence of otitis media, may be important in localizing a genetic defect related to ET dysfunction.37
Certain ethnic groups, such as Native Americans,38 North American Inuits,39 and Australian Aborigines,40,41 have an extraordinarily high rate of middle-ear disease that progresses to chronic suppurative otitis media. Despite the lack of proof of a genetic predisposition in these high-risk populations, it is likely that heredity is an important factor in the pathogenesis, and at least one study found differences in the bony anatomy of the ET between certain racial populations.42 Another study found abnormal ET function in members of a Native American tribe (see High-Risk Populations in the Cholesteatoma section).43
Black Versus White Children in the United States
In the past, it has been assumed that black children in the United States have a lower incidence of otitis media than white children. This assumption was based on several studies, such as the prevalence study by Kessner and colleagues, in which they reported evidence of middle-ear pathology in 35.6% of 112 white children and 19% of black children in Washington, DC.44 Another study from Cleveland by Marchant and colleagues reported that the incidence of otitis media was 62% in young black infants compared with 86% in white infants (p = .017).45 The report from the US Division of Health Care Statistics found that the number of visits for otitis media between 1975 and 1990 made by black children was lower than for white children6; however, two reports in two different populations from Pittsburgh contradict these earlier findings. In the study reported by Casselbrant and colleagues, black infants have the same incidence of otitis media as white infants (Figure 2–3).46 In the other study from Pittsburgh, Paradise and colleagues prospectively followed 2253 infants from soon after birth until age 2 years and reported that one or more episodes of middle-ear effusion occurred in 48% by age 2 months and 91% by 24 months of age and that black infants were affected to the same or a greater degree as white infants, especially among infants of a lower socioeconomic status (Table 2–2).3 A recent study by Fleming-Dutra and colleagues (2014) in the United States of race and otitis media reported the percentage of all visits resulting in a diagnosis of otitis media was 30% lower in black children compared with those who were nonblack.47
Number or Percentage
Mean number of visits
1st year of life
2nd year of life
Percentage subjects with middle-ear effusion at earliest visit after 2 mo
Percentage subjects developing one episode of MEE between age 2 mo and age:
Mean cumulative percentage of days with MEE
1st year of life
2nd year of life
Mean number of days of antimicrobial treatment for otitis media
1st year of life
2nd year of life
Mean number of days of antimicrobial treatment for all reasons
1st year of life
2nd year of life
Percentage subjects undergoing tympanostomy tube insertion
1st year of life
2nd year of life
Adapted from Paradise JL et al.3 MEE = middle-ear effusion.
Infants and children with craniofacial malformations, such as those with cleft palate and Down syndrome, have a high rate of otitis media that is attributed to abnormalities of ET function.48–51
Risk Factors for Inflammation
Risk factors for inflammation most likely adversely affect ET function. Infection secondary to the common cold can result in tubal obstruction, which has an increased frequency when infants and young children are in child day care, have older siblings, or are in the lower socioeconomic strata.3 The winter and spring seasons are associated with a higher rate of viral respiratory tract infections than the fall and winter months, which increases the risk of high negative middle-ear pressures and otitis media (Figure 2–4).19,52,53 Early onset of otitis media (in the first few months of life) as a risk factor for recurrent and chronic disease49 could be related to a genetic predisposition or environmental factors, such as increased exposure to upper respiratory infections, or both. Also, inflammation owing to upper respiratory allergy has been shown to cause tubal obstruction.54–56
Passive Smoking and Ambient Air Pollution
Even though conclusive data are lacking, exposure to smoke in the household has been shown to be a risk factor, which, like smoke from wood-burning stoves, could interfere with normal ET function (clearance and immune defense). Smoke exposure can result in goblet cell hyperplasia and mucus hypersecretion in the respiratory tract, ciliostasis, and decreased mucociliary transport57 and may play a role in altering the immune defenses of the respiratory tract. The availability of a biochemical marker—salivary, serum, or urine cotinine—has made documentation of passive exposure to tobacco smoke more reliable than that provided by history alone. Cotinine concentrations were related to the number of smokers in the household.58 High concentrations of serum cotinine have been associated with an increased incidence of acute otitis media and increased duration of middle-ear effusion following an acute episode.59 Smoking has been associated with impairment of ciliary beat frequency and the occurrence of otitis media.60 Brauer and associates (2006) reported that traffic-related air pollution also is associated with occurrence otitis media.61
Pacifiers, Sleeping Position, and Gastroesophageal Reflux
Currently, there is no explanation for the observed increased incidence of otitis media owing to the use of a pacifier,62,63 but one study found that pacifiers do not contain high numbers of organisms, which would make them unlikely candidates for a source of transfer of potentially pathogenic organisms.64 A study was reported by Warren and colleagues that used data from questionnaires from parents of over 1000 infants during their first year of life, which also showed an increased rate in otitis media associated with the use of the pacifier.65 One reason for the association between the use of pacifiers and the risk of developing otitis media, especially in a child daycare center,66 is that high negative pressure that develops within the pharynx can prevent the ET from opening, especially when nasal obstruction is present, such as when a viral upper respiratory tract is present (the Toynbee phenomenon) (see Chapter 5). Subsequent aspiration or insufflation of infected nasopharyngeal secretions into the middle ear may occur (Figure 2–5).21 Support for the hypothesis that the Toynbee phenomenon may be related to the pathogenesis of otitis media with the use of a pacifier is the study by Lamp, in which he identified a child who had chronic middle-ear effusion and was a habitual thumb-sucker, whose effusion resolved with the cessation of the habit.67
A study in infants of the effect of bottle feeding in the supine position by Tully and colleagues may also be related to the Toynbee phenomenon.68 The investigators reported the occurrence of flat tympanograms when the infants were fed in the supine position but not in the semi-upright position. When the infants had flat tympanograms while in the supine position, the tympanograms returned to normal within 15 minutes when the infants were placed in the semi-upright position. The investigators thought that the flat tympanograms were the result of milk entering the middle ear. But if this is the correct interpretation, then why do mothers only rarely observe milk coming through the non-intact tympanic membrane when infants have functioning tympanostomy tubes in place? A more likely explanation for the flat tympanograms is that they were related to extreme underpressures within the middle ear, as a result of the extraordinary degree of negative pressure infants develop during sucking (the Toynbee phenomenon). Also, in the supine position, the ET is less patent than in the semi-upright position,69 which may make it less likely for negative pressure to be equilibrated.
The prone sleeping position has been identified as a risk factor for otitis media that also could be related to the Toynbee phenomenon. Gannon and colleagues assessed 14,000 infants for sleeping patterns in the United Kingdom and reported an odds ratio of 1.159 to 2.65 for otitis media in those infants who slept in the prone position compared with those who did not.70 Hunt and colleagues reported that in a study that involved 3000 young infants, the rate of otitis media was greater in those infants who slept in a prone position than in those who slept supine.71 They postulated that the supine position could improve nasal drainage during sleep. Even though a definitive explanation for this finding is lacking, the possibility of the Toynbee phenomenon seems reasonable. In infants, nasal obstruction is more likely in the prone than in the supine position, which could adversely affect ET function during swallowing, especially when the child has accumulated nasal and nasopharyngeal secretions. Either development of middle-ear negative pressure or positive nasopharyngeal pressures could result in middle-ear effusion, especially during an episode of an upper respiratory tract viral infection.
Another possible risk factor may be that children who have gastroesophageal reflux have otitis media owing to inflammation of the ET or middle-ear mucosa, or both, if the secretions from the stomach are present in the nasopharynx and enter the tube. Indeed, Tasker and colleagues reported that 91% of 65 children had increased pepsin or pepsinogen levels in their middle-ear aspirates, when compared with their sera, at the time of myringotomy and tympanostomy tube placement.72 The gastric contents could enter the middle ear through the ET by insufflation (crying), by aspiration (associated with middle-ear negative pressure secondary to inflammation of the tube), or during closed-nose swallowing (the Toynbee phenomenon).
Breast-feeding Lack and Otitis Media
In a review of 24 studies from the United States and Europe, Bowatte and colleagues (2015) concluded after a meta-analysis that breast-feeding protects against acute otitis media until age 2 years, but protection was greatest for exclusive and longer-duration breast-feeding.73 Also, according to McIsaac and colleagues (2015), breast-feeding was a means to prevent infant morbidity and mortality in Aboriginal Canadians, which included between 5.1% and 10.6% of otitis media.74 There is no current reason why breast-feeding confers a lower risk for single and recurrent episodes of otitis media, especially when exclusive breast-feeding is for 4 or more months,18,75,76 but it is unlikely to be associated with ET dysfunction. Human milk secretory immunoglobulin A antibody to nontypable Haemophilus influenzae has been thought to convey protection, but this hypothesis has been challenged.77,78 However, if the benefit is related to immunity, there could be an advantage in mucosal immunodefense within the tubal lumen in the prevention of otitis media. Likewise, the presence of impairment in immunity may not be directly related to dysfunction of the tube, but the tube may be secondarily involved.
Other Risk, Prognostic, or Pathogenetic Factors
With regard to men being more susceptible to otitis media than women, we assessed ET function in both sexes of children and found no difference.18,79–81 Currently, there is no known explanation, but this sex predilection may be immunologic in origin. It is unlikely that immature or impaired immunity is related to dysfunction of the tube, although there is an immunodefense system in the mucosa of the lumen of the tube (and middle ear). Human immunodeficiency virus (HIV) infection has been associated with an increased prevalence of otitis media, and the incidence density of recurrent middle-ear disease also significantly correlates with disease progression.82
Recently, a study of New Zealand children reported (2014) that household crowding was associated with hospitalizations for otitis media.83 A study of 1006 children attending child day-care centers in 2014 showed that diet of sweet pastries and jam was associated with increased risk for pneumococcal otitis media and that frequent consumption of fruit and berries was associated with decreased risk.84
Sinusitis has been associated with otitis media in children. Honjo and colleagues examined 131 ears of 83 patients, 10 to 20 years of age, and found that sinusitis was associated with 49% of refractory otitis media cases.85 The use of nasotracheal tubes and nasogastric tubes has been associated with nosocomial middle-ear infection that is most likely due to dysfunction of the ET.86
Decreased volume of the mastoid gas cell system has been associated with increased incidence of otitis media in children, that is, those children with underdeveloped mastoids.87 Some investigators think that the mastoids are underdeveloped on the basis of heredity,88 others believe that the poor development is secondary to otitis media in infancy and early childhood when the mastoid becomes aerated (environmental etiology),89 and still others think that both heredity and environmental factors control the growth and development of the mastoid gas-cell system.90 My opinion is that ET dysfunction in early childhood impedes mastoid aeration, which may then impair middle-ear aeration. (See Chapter 4, “Physiology.”)
Dysfunction of the ET has also been reported to be associated with deviation of the nasal septum91,92; trauma induced by nasogastric and nasal endotracheal tubes93,94; trauma to the palate, pterygoid bone, or tensor veli palatini muscle; injury to the trigeminal nerve or, more specifically, to the mandibular branch of this nerve95; and trauma associated with surgical procedures, such as palatal or maxillary resection for tumor.96 Neoplastic disease, either benign or malignant, that invades the palate, pterygoid bone, or tensor veli palatini muscle can cause otitis media,97 which can be caused by failure of the opening mechanism of the tube.95,98,99
Apparently, children with cystic fibrosis have less otitis media than those without the disease. It is the collective experience of many otolaryngologists that very few affected children require tympanostomy tube insertion for recurrent or chronic disease. Haddad and colleagues reported that of 70 infants and children with cystic fibrosis evaluated at the Children’s Hospital of Pittsburgh, 7 (10%) had a point prevalence of otitis media.100 When compared with another Pittsburgh study of children without cystic fibrosis, Casselbrant and colleagues found a point prevalence of 22–33% during the winter months.19 Even though the populations and methodologies were not strictly comparable, I would agree that children with cystic fibrosis probably have less otitis media despite the almost universal occurrence of sinusitis in this population. The explanation for this enigma may be related to the frequent use of antibiotics in these children, but, if so, why do they have the high rate of recurrent and chronic sinusitis? Another study by Todd and Martin revealed that temporal bone pneumatization—assessed by computed tomography (CT)—in 350 children with cystic fibrosis had larger volumes of mastoid pneumatization compared with reference data from cadaver specimens; sinus pathology was present in all developed sinuses in the CT scans of children with cystic fibrosis.101 Even though controversy still lingers, most current investigators are convinced that recurrent or chronic otitis media early in life is associated with hypoaerated mastoids and many have concluded that the better the ET function is, the larger the mastoid gas volume, and vice versa. (See Chapter 4.) Todd and Martin speculated that ET function may be better in patients with cystic fibrosis and that the gene for differentiation of the good function of the tube may be linked to the gene determining cystic fibrosis. This population appears to be a rich research area related to possible genetic differences and ET compared with individuals without cystic fibrosis. There is some speculation that the difference is related to chloride channels. (See Chapter 11, “Future Directions.”)
Complications and Sequelae
Complications and sequelae of middle-ear disease remain a major health problem. The World Health Organization estimates 20,000 individuals die around the world from complications of otitis media, presumably mostly in developing nations. A recent review of hospital admissions in the United States found complications that occurred in children, such as mastoidis, meningitis, venous thrombosis, and brain abscess, resulting in 21 deaths.102
Chronic Suppurative Otitis Media
Chronic suppurative otitis media is a major health problem in many populations around the world, affecting diverse racial and cultural groups living not only in temperate climates but also in climate extremes ranging from the arctic circle to the equator.103 From a review of approximately 50 reports published during the past 30 years, there appear to be four groups of populations based on the prevalence of the disease (Table 2–3). The populations in which the prevalence of chronic suppurative otitis media (defined in the published reports as chronic perforation with and without suppuration but not cholesteatoma) has been reported to be the highest are the Inuits of Alaska (30–46%),104–106 Canada (7–31%),38,107–112 and Greenland (7–12%)113–116; Australian Aborigines (12–33%);39,117–125 and certain Native Americans (e.g., Apache and Navajo tribes) (4–8%).38,126–130 Apparently, these North American Indian tribes have higher rates than others.131 Populations with moderately high rates are certain natives of the South Pacific islands, such as the Solomon Islands (4–6%),132 New Zealand Maori (4%),133,134 Malaysians (4%),135 and Micronesians (4%)136–138 (in contrast to these high rates, in some islands of the South Pacific, natives of Melanesia have an extremely low rate, less than 1%), as well as some African populations, such as Sierra Leone (6%),139 Gambia (4%),140 Kenya (4%),141 and Tanzania (2–3%).142–144 However, not all reports from Africa have documented these relatively high rates. One study from Nigeria reported less than 1% with the disease145; another study found a 4% rate.146 One study of South African rural blacks also found a rate of less than 1%.147 Populations with relatively low rates of chronic suppurative otitis media are Korea (2%),148 India (2%),149 and Saudi Arabia (1.4%).150 Studies from highly industrialized nations have reported the lowest rates (none or less than 1%), such as the United States,20,46,151 Finland,152 the United Kingdom,153–155 and Denmark.156 In one adult population in Great Britain, the rate has been reported to be 3.1%.155
But with the widespread use of tympanostomy tubes in these countries, chronic suppurative otitis media occurs as a not uncommon complication in infants and children.157 A recent report from Orji and colleagues (2015) found the prevalence of this complication of otitis media in a developing country has been reduced in past 20 years, but there has been shift in the epidemiology toward the adult population.158 Shaheen and Nahar reported in 2014 that in Bangladesh the rate of chronic suppurative otitis media was higher in rural primary-school children (6.02%) than in urban primary-school children (2.07%).159