Epidemiology and Genetics of Hearing Loss and Tinnitus

7.1

Epidemiology of Sensorineural Hearing Loss

Worldwide, 16% of disabling hearing loss in adults is attributed to occupational noise exposure, ranging from 7% to 21% in various regions of the world ( ). At the time of that study, in the United States an estimated 9.4 million workers were exposed at levels more than 80 dB(A), of which 3.4 million were exposed to levels more than 90 dB(A). In a UK sample ( N =21,201; age ≤65 years), about 2% of subjects reported severe hearing difficulties, i.e., wearing a hearing aid or having great difficulty hearing a conversation in a quiet room. In men, the prevalence of this outcome rose steeply with age, from below 1% in those aged 16–24 years to 8% in those aged 55–64. In the United Kingdom overall some 153,000 men and 26,000 women aged 35–64 years were estimated to have severe hearing difficulties attributable to noise exposure at work ( ). analyzed the questionnaires answered by 9756 working people and 1685 nonworkers in a Swedish population. They found that 31% in the working population and 36% in the nonworking population reported either hearing loss or tinnitus or both. Nonoccupational (e.g., recreational) noise exposure was not taken into account, but the close numbers suggest that it likely accounts for an important part of the hearing loss (see chapter: Causes of Acquired Hearing Loss ).

The prevalence of audiometric hearing loss among all individuals (age ≥12 years) in the United States was estimated using an extrapolation from a nationally representative data set ( ). Pure-tone thresholds from people ( N =7490) aged from 12 years to well over 70 years from the 2001 through 2008 cycles of the National Health and Nutritional Examination Surveys (NHANES) were analyzed. A pure-tone average (PTA) of the hearing thresholds at 0.5, 1, 2, and 4 kHz of at least 25 dB HL (hearing level) in both ears as recommended by the World Health Organization was taken as an indication of hearing loss. From this sample, the authors extrapolated that 30.0 million (12.7%) of Americans 12 years and older had bilateral hearing loss from 2001 through 2008, and this estimate increased to 48.1 million (20.3%) when individuals with unilateral hearing loss were included. Overall, the prevalence of hearing loss increased with every age decade ( Fig. 7.1 ). The prevalence of hearing loss was lower in women than in men and in African-American versus Caucasian individuals across nearly all age decades.

More than 3 million people in Germany with employment subject to social insurance contributions to three health-insurance providers were included in the cohort studied by . During the study period (January 1, 2004–December 31, 2008), 283,697 cases of hearing loss were newly diagnosed. As expected, these incidence rates increased with age. Up to the age of 36 years the incidence of hearing loss in 4-year age groups was higher in women, but above that age it was higher for men ( Fig. 7.2 ).

In the period from 1996 to 1998, collected audiometric data from 50,723 of 82,141 unscreened invited subjects in Nord-Trondelag, Norway (age range 20–101 years, mean=50.2 years). The PTA of hearing thresholds at 0.5, 1, 2, and 4 kHz showed hearing impairment more than 25 dB in the worst ear in 32% of males and 23% of females. The overall prevalence of this hearing impairment was 18.8% for the better ear and 27.2% for the worse ear, respectively. In the same subjects, compared the effects of occupational noise and firearms noise. Reported noise exposure levels and observed threshold shifts were moderate among women. Threshold shifts averaged over both ears among subjects in the group of the highest 2% of exposure levels were 13 dB for 65-year-old men and generally largest at 3–4 kHz. The shifts induced by impulse noise were approximately 8 dB and strongest at 3–8 kHz among men aged 45–65 years. Comparable results for firearms noise were obtained from 3753 participants aged 48–92 years in the Beaver Dam study in Wisconsin ( ). After age and other factors were adjusted, men ( N =1538) who had regularly participated in target shooting (OR=1.57; CI=1.12–2.19) or who had done so in the past year (OR=2.00; CI=1.15–3.46) were more likely to have a marked high-frequency hearing loss than those who had not. The risk of having a marked high-frequency hearing loss increased 7% for every 5 years the men had hunted (OR=1.07; CI=1.03–1.12). Thirty-eight percent of the target shooters and 95% of the hunters reported they did not use hearing protection.

For a representative sample of 705 subjects from a rural Danish population aged 31–50 years, reported changes in hearing sensitivity over 5 years. Hearing deterioration was defined as an average at least 10 dB/5 years at 3–4 kHz in at least one ear and was present in 23.5% of the sample. The 41- to 50-year-olds had an OR=1.32 (CI=1.01–1.73) compared with the 31- to 40-year-olds. Males had an OR=1.35 (CI=1.03–1.76) compared with females. These example data suggest that a large percentage of the adult population has noise-induced hearing loss (NIHL), regardless being from a rural or urban environment.

7.2

Epidemiology of Age-Related Hearing Loss

determined the 10-year cumulative incidence of hearing impairment and its associations with education, occupation, and noise exposure history in a population-based cohort study of 3753 adults who were 48–92 years of age at the baseline examinations during 1993 – 95 in Beaver Dam, WI. Hearing thresholds were measured at baseline and at 2.5-year, 5-year, and 10-year follow-up examinations. Hearing impairment was defined as a PTA more than 25 dB HL at 0.5, 1, 2, and 4 kHz. Demographic characteristics and occupational histories were obtained by questionnaire. The 10-year cumulative incidence of hearing impairment was 37.2%. Age (5 year; HR=1.81), sex (men vs women; HR=2.29), occupation based on longest held job (production/operations/farming vs others; HR=1.34), marital status (unmarried vs married; HR=1.29), and education (<16 vs 16+ years; HR=1.40) were associated with the 10-year incidence, whereas a history of occupational noise was not. In this largely retired population, occupational noise exposure may have contributed to hearing impairments present at the baseline examination ( ), but there was no evidence of any residual effect on long-term risk of declining hearing sensitivity among people with normal hearing at the baseline examination. Even among those exposed to occupational noise at the baseline examination, there was no evidence of an effect. These results are consistent with the study by , which measured hearing repeatedly, and reported no difference in the rate of change between people with and without positive noise histories. ’s study suggests that, on a population basis, there is little evidence that prior occupational noise exposure plays an important role in the onset or progression of hearing impairment in older adults followed for 10 years. Note that these audiometric data do not reflect “hidden hearing loss” resulting from damage to inner hair cell ribbon synapses or loss of high-threshold auditory nerve fibers (see chapter: Hearing Problems, chapter: Types of Hearing Loss ).

7.3

Epidemiology of Tinnitus

Tinnitus prevalence in the general population was extracted from three review papers, the original publications contributing to those overviews, and from more recent papers. One review provided an in-depth reanalysis of a few large epidemiology studies ( ). The second study also covered some older epidemiology where different criteria for inclusion of tinnitus were used ( ). The third study presented a more general (but without a prevalence by age group) overview of a larger number of epidemiology studies ( ). All in all they covered 14 papers that illustrate an upward trend of tinnitus prevalence with age that is generally the same for all studies but where the absolute levels depend on the questions asked and the type of tinnitus included. This was previously reviewed more extensively in my book “The Neuroscience of Tinnitus” ( ). The prevalence of significant tinnitus across the adult lifespan is illustrated in Fig. 7.3 . Significant tinnitus has to be longer than 5 min in duration and not immediately (and transiently) following exposure to loud noise ( ). Sometimes, more stringent definitions, such as that tinnitus has to be bothersome, are used. This typically lowers the prevalence a few percentage points. This bifurcation can be seen in Fig. 7.3 . As these prevalence studies across the lifespan show, tinnitus is about twice as frequent in the elderly as in young adults.

Hearing loss, resulting from exposure to loud noise, is considered an important risk factor for developing tinnitus. Consequently, a history of recreational, occupational, or firearm noise exposure may be associated with increased likelihood of acquiring tinnitus. The relation between noise exposure and significant tinnitus, however, differs depending on the presence or absence of hearing impairment. It should be noted that clinical hearing impairment generally requires a loss of at least 25 dB over the audiometric frequencies (0.25–8 kHz). Especially, hearing loss for frequencies more than 8 kHz typically is accompanied by tinnitus ( ). Occupational noise exposure was more likely to correlate with significant tinnitus in participants with hearing impairment, while recreational noise exposure was more associated with increased occurrence of significant tinnitus in participants without (clinical) hearing impairment ( ). confirmed that occupation had a marked effect on tinnitus prevalence. In men, age-adjusted prevalence ORs of tinnitus (in relation to a reference population of teachers) ranged from 1.5 (workshop mechanics) to 2.1 (crane and hoist operators) in the 10 occupations with the highest tinnitus prevalence. In women, the most important contribution to the tinnitus prevalence was from the large group of occupationally inactive persons, with a prevalence OR of 1.5.

Using data from the Epidemiology of Hearing Loss Study (1993–95, 1998–2000, 2003–05, and 2009–10) and the Beaver Dam Offspring Study (2005–08) in the United States, examined birth cohort patterns in the report of tinnitus for adults aged 45 years and older ( N =12,689 observations from 5764 participants). They found that tinnitus prevalence tended to increase in more recent birth cohorts compared to earlier birth cohorts. On average, participants in a given generation were significantly more likely to report tinnitus than participants from a generation 20 years earlier (OR=1.78, CI=1.44–2.21). This also may underlie the leveling off of tinnitus prevalence in Fig. 7.3 for the age group above 65 years and may thus refer back to the much lower prevalence in cohorts born before 1950.

The average prevalence of significant tinnitus by age group in the two Scandinavian countries, the United States and the United Kingdom (from Fig. 7.3 ), is shown in Fig. 7.4 . The most recent study covered 14,178 participants in the 1999–2004 National Health and Nutrition Examination Surveys ( ). The overall prevalences of the tinnitus in the sample groups were: the United Kingdom 10.1%, Sweden 14.2%, the United States 8.4%, and Norway 15.1%. One observes a tendency for the prevalence of tinnitus to level off in the seventh decade of life. In contrast, the prevalence for significant hearing loss (>25 dB HL, from 0.5–4 kHz) continues to increase ( ). This suggests that the prevalence of clinical hearing loss in the standard audiometric frequency range is not related to tinnitus prevalence.

A comprehensive new study ( ) described the incidence rate of clinical significant tinnitus in the United Kingdom. They identified 14,303 incident cases of significant tinnitus among 26.5 million person-years of observation. They found an incidence rate of 5.4 (CI=5.3–5.5) cases per 10,000 person-years. The incident rate peaked in the 60–69 years age group ( Fig. 7.5 ) and was the same for males and females. The incident rate increased with approximately 0.21 per year over the time span 2002–11.

7.4

Epidemiology of Smoking and Alcohol Consumption

Taken on their own, smoking and alcohol consumption have opposite effects on hearing loss. Nicotine affects the cochlea ( ) through its effects on antioxidant mechanisms and/or on the vasculature supplying the auditory system. Smoking is accompanied by a higher incidence of high-frequency hearing loss. Moderate alcohol consumption appears to have a protective effect on hearing.

found that current smokers were 1.69 times as likely to have a hearing loss as nonsmokers (CI=1.31–2.17). This relationship remained for those without a history of occupational noise exposure and in analyses excluding those with non-age-related hearing loss. Nonsmoking participants who lived with a smoker were more likely to have a hearing loss than those who were not exposed to a household member who smoked (OR=1.94; CI=1.01–3.74). Follow-up studies in the “Established Populations for Epidemiologic Studies of the Elderly” sample ( ) found among the 10,118 participants, 1406 (12.4%) that reported hearing problems at baseline. Of those with no baseline hearing problems and complete follow-up information ( N =8495), 1120 (13.2%) developed new hearing problems. Smoking was associated with higher prevalence and incidence rates of hearing impairment. In both cases the association was weak although statistically significant. Compared with participants without a history of smoking, those who had ever smoked were more likely to report hearing problems at baseline (OR=1.2; CI=1.0–1.3) and more likely to develop new hearing problems over the follow-up period (OR=1.6; CI=1.4–1.8).

Three studies measured the overall effect of both smoking and alcohol consumption on hearing loss in the same population. The first study ( ) collected audiometric data in 4083 subjects between 53 and 67 years from nine audiological centers across Europe. PTAs for 0.5, 1, and 2 kHz were adjusted for age and sex and tested for association with exposure to risk factors. Noise exposure was associated with a significant loss of hearing at frequencies of 1–8 kHz. Smoking significantly increased high-frequency hearing loss, and the effect was dose dependent. Moderate alcohol consumption was inversely correlated with hearing loss. Significant associations were found in the high as well as in the low frequencies.

In the Blue Mountain Hearing Study ( ) of 2956 participants (aged >50 years) alcohol consumption and smoking status were measured by using an interviewer-administered questionnaire. Logistic regression was used to obtain ORs with 95% CIs that compared the chances of having hearing loss in participants, who did or did not smoke or consume alcohol, after adjusting for other factors previously reported to be associated with hearing loss. Cross-sectional analysis demonstrated a significant protective association between the moderate consumption of alcohol (>1 but ≤2 drinks/day) and hearing function in older adults (compared with nondrinkers), OR=0.75 (CI=0.57–0.98). Smokers who were not exposed to occupational noise had a significantly higher likelihood of hearing loss after adjusting for multiple variables, OR=1.63 (CI=1.01–2.64). The interaction between smoking and noise exposure was not significant.

evaluated the association between smoking, passive smoking, alcohol consumption, and hearing loss in 164,770 adults aged between 40 and 69 years who completed a speech-in-noise hearing test (the Digit Triplet Test). Current smokers were more likely to have a (speech-in-noise) hearing loss than nonsmokers (OR=1.15, CI=1.09–1.21). Nonsmokers who reported passive exposure to tobacco smoke were also more likely to have a hearing loss (OR=1.28, CI=1.21–1.35). Those who consumed alcohol were less likely to have a hearing loss than lifetime teetotalers. The association was similar across three levels of consumption by volume of alcohol (lightest 25%, OR=0.61, CI=0.57–0.65; middle 50%, OR=0.62, CI=0.58–0.66; heaviest 25%, OR=0.65, CI=0.61–0.70). They defined one unit of alcohol equal to 8 g, consequently one glass of wine and beer were rated as 2.5 units, and one shot of strong liquor at 1 unit. The lightest consumption level was less than 118.4 g/week (not including teetotalers; i.e., <6 glasses of wine or beer), middle level 118.4–196.8 g/week (6–10 glasses), and the highest consumption level more than 196.8 g/week (>10 glasses). Regardless of the dose level, alcohol consumption was associated with a protective effect.

The incidence of tinnitus also depends on the smoking and alcohol consumption history, much in the same way as hearing loss does. described the 10-year cumulative incidence of tinnitus and its risk factors. Participants ( N =2922, aged 48–92 years) who did not report tinnitus at the baseline study (1993–95) were followed for up to 10 years. In addition to audiometric testing data on tinnitus, health, and other history were obtained via questionnaire. Potential risk factors were assessed with discrete-time proportional hazards models. They found that “the 10-year cumulative incidence of tinnitus was 12.7%. The risk of developing tinnitus was significantly associated with a history of ever smoking (HR=1.40), and among women, hearing loss (HR=2.59). Alcohol consumption (HR=0.63 for ≥141 grams/week vs. <15 grams/week) was associated with decreased risk.”

There have, so far, not been studies into the interaction of smoking and alcohol consumption, i.e., by separating smokers in a drinking and nondrinking group, on hearing loss or tinnitus in the same subjects.

7.5

Epidemiology of Diabetes

As we have seen in Chapter 6 , Causes of Acquired Hearing Loss, uncontrolled diabetes may cause hearing loss. The epidemiological studies reviewed here detail the risks. found greater amounts of hearing loss in less than 50 years old adult diabetes subjects compared to controls. Significant hearing differences were present at all frequencies for Type 2 diabetes subjects, but for Type 1 subjects (for definitions see chapter: Causes of Acquired Hearing Loss ), differences were found ≤1 kHz and ≥10 kHz. Over age 50 years, there were significant associations between hearing at low frequencies for Type 1 only. found that the association between diabetes and hearing impairment was independent of known risk factors for hearing impairment, such as noise exposure, ototoxic medication use, and smoking. They phrased this as: “The ORs, adjusted for low- or mid-frequency and high-frequency hearing impairment, were 1.82 (CI=1.27–2.60) and 2.16 (CI=1.47–3.18), respectively.” reported for adults aged 25–69 years, tested between 1971 and 2004, that “the adjusted prevalence ratios of hearing impairment for persons with diabetes vs. those without diabetes was 1.17 (CI=0.87–1.57) for the NHANES 1971–1973 and 1.53 (95% CI, 1.28–1.83) for NHANES 1999–2004.” This suggests a clear increase in hearing loss prevalence over the 25-year span between the two surveys.

reported the relationship between Type 2 diabetes and the prevalence, the 5-year incidence, and the progression of hearing impairment in a representative, older, Australian population (the Blue Mountain study). They found that age-related hearing loss was present in 50% of diabetic participants ( N =210) compared with 38.2% of nondiabetic participants ( N =1648), OR=1.55 (CI=1.11–2.17), after adjusting for multiple risk factors. Diabetes duration and hearing loss were positively correlated. After 5 years, incident hearing loss occurred in 18.7% of participants with, and 18.0% of those without diabetes, adjusted OR=1.01 (CI=0.54–1.91). Progression of existing hearing loss (>5 dB HL) was significantly greater in participants with newly diagnosed diabetes (69.6%) than in those without diabetes (47.8%) over this period, adjusted OR=2.71 (CI=1.07–6.86). Type 2 diabetes was associated with the larger prevalent, but not incident hearing loss in this older population.

conducted a longitudinal population-based cohort study (1993–95 to 2009–10). Follow-up examinations were obtained from 87.2% ( N =1678; mean baseline age 61). The 15-year cumulative incidence of hearing impairment was 56.8%. Adjusting for age and sex, current smoking (HR=1.31, p =0.048) and poorly controlled diabetes mellitus (HR=2.03, p =0.048) were associated with greater risk of hearing impairment. People with better-controlled diabetes mellitus were not at greater risk.

7.6

Epidemiology of Otitis Media

Conductive hearing loss resulting from acute otitis media or otitis media with effusion is mostly intermittent mild to moderate hearing loss in infants and young children. studied a population-based cohort of 32,786 participants who had their hearing tested by pure-tone audiometry in primary school and again at ages between 20 and 56 years. Hearing loss was diagnosed in 3066 children; the remaining sample had normal childhood hearing. Significantly reduced adult hearing thresholds in the whole frequency range were found in those diagnosed with childhood hearing loss caused by otitis media with effusion ( N =1255; 2 dB), chronic suppurative otitis media (CSOM; N =108; 17–20 dB), or hearing loss after recurrent acute otitis media (rAOM; N =613; 7–10 dB) compared those with normal childhood hearing. The effects of CSOM and hearing loss after rAOM on adult hearing thresholds were larger in participants tested in middle adulthood (ages 40–56 years) than in those tested in young adulthood (ages 20–40 years). concluded that CSOM as well as rAOM in childhood are associated with adult hearing loss.

7.7

Epidemiology of Auditory Neuropathy Spectrum Disorder

Auditory nerve myelinopathy and/or deficits in synchrony of neural discharges are the most probable underlying pathophysiological mechanisms of auditory neuropathy spectrum disorder (ANSD) (see chapter: Types of Hearing Loss ). collected all available published evidence on the prevalence of auditory neuropathy in the “well-baby” population in the Netherlands. The population-based prevalence in children in population hearing screening was found to vary between 0.006% (SD=0.006) and 0.03% (SD=0.02). The false-negative rate, based on otoacoustic emission testing in newborn hearing screening programs, caused by missed children with auditory neuropathy, was estimated between 4% and 17%. investigated in the period from 2002 to 2011, 9419 infants whose hearing ability was uncertain or who had risk factors for hearing loss were investigated, and 352 were diagnosed with sensorineural hearing loss (SNHL). Of these 352 children, 18 (5.1%) were diagnosed with ANSD, suggesting that is not an extremely rare hearing disorder. Auditory neuropathy, as defined in the audiology/otolaryngology literature, occurs frequently and is responsible for approximately 8% of newly diagnosed cases of hearing loss in children per year ( ).

7.8

Genetics of Sensorineural Hearing Loss

Genetic deafness that affects about 0.1% of individuals by severe or profound deafness at birth or during early childhood, i.e., in the prelingual period can be distinguished in syndromic and nonsyndromic deafness ( ). Syndromic deafness is associated with other defects and contributes to about 30% of the cases during early childhood and may be conductive, sensorineural, or mixed. Nonsyndromic hereditary deafness is classified by the mode of inheritance; DFNX, DFNA, and DFNB refer to deafness forms inherited on the X chromosome-linked, autosomal (i.e., any chromosome other than a sex chromosome) dominant, and autosomal recessive modes of transmission, respectively. About 80% of the cases of prelingual nonsyndromic deafness are DFNB forms, whereas most of the late-onset forms are DFNA forms. Prelingual nonsyndromic deafness is almost exclusively sensorineural ( ).