Prevalence of Tinnitus in Community-Dwelling Japanese Adults

Background Several studies have reported the prevalence of tinnitus among general populations; however, most of these studies were conducted in Europe or the United States. We estimated the prevalence of tinnitus among the general adult population in Japan. Methods The subjects were participants in the Takayama Study, a population-based cohort study. In 2002, a total of 14 423 adults (6450 men and 7973 women) aged 45 to 79 years responded to a self-administered questionnaire that inquired about history of tinnitus, which was defined as episodes lasting longer than 5 minutes, excluding those occurring immediately after noise exposure. Respondents were also asked about the loudness and severity of tinnitus. Results Overall, 11.9% of the subjects reported having tinnitus; the percentage was somewhat higher among men (13.2%) than women (10.8%). The prevalence of tinnitus increased with age in both sexes. Approximately 0.4% of the overall population reported that tinnitus had a severe effect on their ability to lead a normal life. Medical history of hypertension or ischemic heart diseases, use of steroid or antihypertensive medication, and employment as a factory worker or machine operator were associated with tinnitus status in both men and women. Conclusions Tinnitus is relatively common in Japan. Although the use of various definitions of tinnitus in different studies makes it difficult to compare prevalence among populations, the present prevalence estimate was similar to those in studies in Europe and the United States.


INTRODUCTION
Tinnitus is defined as sound perceived in the ear or head for which there is no acoustic source external to the head. Thus, tinnitus is an entirely subjective experience that can only be described by patient reports. 1 Tinnitus prevalence has been reported in several studies of general populations. [2][3][4][5][6][7][8][9][10][11][12] These studies indicate that approximately 10% to 15% of the population experience tinnitus. 11 However, most of these studies were conducted among populations in Europe and the United States. In Japan, only 1 study to date has examined the prevalence of tinnitus among the general population. 12 However, the study subjects were limited to elderly adults: the prevalence of tinnitus was 18.6% among 1320 residents of a community in Japan who were aged 65 years or older. In that study, participants were asked to report any history of tinnitus in the past year. The Comprehensive Survey of Living Conditions in Japan investigated the prevalence of tinnitus in the population. 13 In that survey, participants were asked whether they had felt unwell or uncomfortable due to ill health or injury in the past few days. If they answered affirmatively, they were asked to indicate the symptoms they had experienced from a list of 42 general symptoms, including tinnitus. According to the 2007 survey, tinnitus was reported by 3.8% of the entire adult population. The percentage among those aged 65 years or older was 7.4%. Thus, the prevalence of tinnitus appears to vary with the definition used. In the present study, we used the definition of tinnitus employed in previous studies to determine its prevalence among adults living in a community in Japan. We also attempted to identify the factors associated with tinnitus status.

METHODS
This survey was conducted among participants of the Takayama Study, which is a population-based cohort study initiated in 1992. The design and methodology of that study has been described elsewhere. 14,15 All nonhospitalized residents aged 35 years or older in Takayama City, Gifu were invited to participate in the study. A total of 31 552 persons, yielding a participation rate of 85.3%, completed a baseline questionnaire that included questions on demographic characteristics, smoking and drinking habits, diet, exercise, occupation, and medical and reproductive histories. In July 2002, we sent a questionnaire that inquired about history of tinnitus to those who were younger than 70 years of age at the baseline in 1992. We obtained residential registry data on death and moves between September 1992 and March 2000, but not after that period. Among 22 435 individuals, after the exclusion of those known to have died or moved, 14 975 responded to the questionnaire. The details are described elsewhere. 16 The response rate was 66.7%.
To identify tinnitus cases, the following question was asked: "Have you ever had tinnitus lasting longer than 5 minutes? Do not include when this happened immediately after very loud sounds." This was the criterion proposed by Coles. 3 Each participant was asked to choose from 3 responses: "I have never had tinnitus," "I have tinnitus," or "I have had tinnitus." Those individuals reporting a history of tinnitus were asked to provide information concerning loudness, severity, and age at onset. Tinnitus severity was assessed by the answers to 2 questions, one on the regularity of tinnitus during waking hours and the other on the extent to which tinnitus affected the respondent's ability to lead a normal life. These questions were selected from questionnaires developed by Coles 3 and Erlandsson. 17 We also asked participants to provide demographic information and details of their smoking status, exercise, and medical histories. Exercise was assessed by asking respondents about the average hours per week spent performing various kinds of activities during the past year. The activities were categorized into strenuous sports (jogging, bicycling on hills, tennis, racquetball, swimming, laps, or aerobics), vigorous work (moving heavy furniture, loading or unloading trucks, shoveling, weight lifting, or equivalent manual labor), and moderate activity (brisk walking, golfing, bowling, bicycling on level ground, or gardening). The time per week spent at each intensity was multiplied by its corresponding energy expenditure requirement, expressed in metabolic equivalent (METs), and summed to yield a MET score (METs·h/week). The details, including the analysis of the validity of the method, are described elsewhere. 18 Informed consent was obtained from each subject, and the study was approved by the ethical board of the Gifu University Graduate School of Medicine.
A total of 552 subjects (224 men and 328 women) did not answer the questions about tinnitus. The remaining 14 423 (6450 men and 7973 women) were included in the present analysis. After taking into account the participation rate (85.3%) in the Takayama Study, the ultimate response rate to the tinnitus questionnaire was approximately 55%.
The prevalence of tinnitus was calculated according to age and sex. Logistic regression models were used to examine associations of tinnitus with marital status, years of education, body size, smoking and drinking habits, exercise, and occupational and medical histories. Information about alcohol intake and occupation was obtained at baseline only. To evaluate the effect of nonresponse on the estimation of prevalence, we examined baseline data from the 2002 questionnaire to determine whether variables shown to be associated with tinnitus differed between respondents and nonrespondents.

RESULTS
The tinnitus status of the respondents is shown in Table 1. Overall, 11.9% of subjects reported that they had recurrent tinnitus. The percentage was somewhat higher among men (13.2%) than women (10.7%). The percentage reporting a history of tinnitus was 16.0%. The prevalence of tinnitus increased with age in both sexes. The peak was at age 70 to 79 years: 15.5% of men and 14.5% of women in this age group reported tinnitus. Table 2 shows the loudness distribution, regularity of tinnitus during waking hours, and extent of the inability to lead a normal life among those reporting current tinnitus. Approximately 50% reported low levels of sound associated with tinnitus. However, 20% to 30% reported an awareness of tinnitus during waking hours. Approximately 0.4% of the total population reported that tinnitus severely affected their ability to lead a normal life. Table 3 shows the age-adjusted ORs of tinnitus according to sociodemographic and clinical factors. Years of education, exercise, and some occupational and medical factors were significantly associated with tinnitus status. Among the factors that were significantly associated with tinnitus, years of education, exercise, and occupational history of mining/ quarrying/rock crushing/cement manufacturing significantly differed among male respondents and nonrespondents after controlling for age; in comparison with respondents, nonrespondents were less likely to have attained a high level (≥15 years) of education (11.7% vs 11.1%, respectively) and less likely to been exercisers (67.7% vs 65.3% for ≥7.5 METs·h/wk). They were more likely to be have been engaged in mining/quarrying/rock crushing/cement manufacturing (1.6% vs 2.1%). In women, the distributions of factors associated with tinnitus were similar between respondents and nonrespondents. The details of the baseline characteristics of nonrespondents and respondents have been described elsewhere. 15

DISCUSSION
The overall prevalence of tinnitus in adults aged between 45 and 79 years was 11.9%. Although only 0.4% of adults reported that tinnitus severely affected their ability to lead a normal life, this finding should still be regarded as important because of the poor results of tinnitus treatment. 19 The prevalence of tinnitus in the general adult population has been estimated in a number of countries (Table 4). In a study conducted in Japan, 12 tinnitus was present in 15.3% and 18.9% of adults aged 65 to 69 years and 70 to 79 years, respectively. Prevalence in the corresponding age groups was somewhat lower in the present study. However, a much lower prevalence (5.2% for those aged 45-79 years) was found in the Comprehensive Survey of Living Conditions in Japan, 13 in which only individuals who indicated that they felt unwell or  4,9,10 or 70 to 79 years, 5,11 with a subsequent decline in older age groups. A somewhat higher prevalence among men than women has been reported in some 2,7,11,12 but not all studies. 5,6,9,10 High noise levels can cause hair cell damage. 20 Workrelated exposure to noise has been reported to be associated with tinnitus, 8, 21 and we confirmed that workers exposed to occupational noise were more likely to have tinnitus. An association between history of cardiovascular disease and tinnitus has been reported in several studies. 8,9,12 The mechanism for this association is unclear, but the function of the auditory nervous system may be affected by cardiovascular disease. 19,22 The observed associations of tinnitus with histories of ischemic heart disease and hypertension provide additional evidence that cardiovascular disorders may be a potential risk factor of tinnitus. Numerous drugs and chemicals, including antimicrobials, diuretics, antineoplastic drugs, and salicylates and other nonsteroidal anti-inflammatory drugs, have been reported to be potentially ototoxic. 23 A history of asthma and use of steroids were associated with tinnitus in the present study. Because asthma involves inflammatory mediators common in the pathogenesis of cardiovascular disease, 24 it is possible that asthma causes tinnitus. Steroids are not classified as ototoxic medications. Thus, it is not steroid use itself, but rather the clinical indications for steroid administration, including asthma, that might play a role in tinnitus. Furthermore, the use of steroids as a treatment for tinnitus or hearing loss, which often accompanies tinnitus, 20 may be another explanation for the observed association between steroid use and tinnitus.
The risk factors for tinnitus are not well known. Many theories have been proposed to explain the pathophysiologic basis of tinnitus, 20 and there are likely to be many possible mechanisms, due to the potential effects of previous disease history and medication use on tinnitus. Although exercise was associated with tinnitus in men, there is no plausible biological explanation for exercise as a cause of tinnitus. Because of the cross-sectional nature of the present data, Prevalence of Tinnitus in Japan 302 no causal inferences can be drawn regarding observed associations. Perhaps a diagnosis of tinnitus or experience of related symptoms prompted some men to begin exercise programs.
Because we could not thoroughly identify decedents due to lack of information in the city registry, the response rate in the present study might be underestimated. However, it is unlikely that such underestimation was large. Therefore, we should consider the possibility that nonresponse bias affected the results. It was an advantage that we were able to compare the characteristics of respondents and nonrespondents using information from a baseline questionnaire of this cohort. The observed associations of tinnitus with level of education and mining/quarrying/rock crushing/cement manufacturing suggest that we might have underestimated prevalence among men. However, the association of exercise with tinnitus suggests an overestimation of prevalence. Nonetheless, the magnitude of these associations and the distributions of the variables in respondents and nonrespondents suggests that the effect of nonresponse bias was not large. It must be kept in mind, however, that the comparison of the baseline characteristics of nonrespondents and respondents may or may not be valid, depending on changes in these variables after the baseline survey. Self-reporting of tinnitus is another limitation of the study. However, because tinnitus is by its nature subjective, there is no objective measure to prove its existence or verify reported severity.
In conclusion, we found that 11.9% of community-dwelling Japanese adults aged 45 to 79 years had tinnitus, and approximately 0.4% experienced tinnitus that was severe enough to reduce their ability to lead a normal life. Although the differing definitions of tinnitus used in studies make it difficult to compare prevalence among populations, the present prevalence study was similar to estimates in studies of European and US populations. Further studies should be encouraged to thoroughly identify the causes, characteristics, and impact of tinnitus in affected individuals.