Abstract
Objective:
To estimate the national prevalence of asymmetric hearing among adults through applying two distinct audiometric criteria.
Study Design:
National cross-sectional survey.
Setting:
Ambulatory examination centers within the National Health and Nutrition Examination Survey (NHANES).
Patients:
Non-institutionalized adults in the United States from the 2001 to 2012 cycles of NHANES aged 20 years and older with pure tone audiometric and tympanometric data (n = 6,190).
Intervention:
Standardized protocol for pure tone audiometry and tympanometry.
Main Outcome Measure:
Proportion of asymmetric hearing according to two distinct audiometric criteria. One criterion (American Academy of Otolaryngology–Head and Neck Surgery [AAO-HNS]) specifies asymmetry as a difference between pure tone averages (PTA) greater than 15 dB, and the other (Veterans Affairs [VA]) specifies asymmetry as a difference greater than/equal to 20 dB across two contiguous frequencies or 10 dB across three contiguous frequencies. Analyses included sampling weights to account for the epidemiologic survey’s complex sampling design.
Results:
Using a definition from the AAO-HNS, overall prevalence was 2.77 and 9.46% when calculating the PTA with 0.5 to 4kHz and 4 to 8kHz, respectively. In contrast, through a working definition used within the VA, overall prevalence was 25.05% across 0.5 to 8kHz. Estimates differed across sex and age, with men and older age cohorts exhibiting higher prevalence.
Conclusions:
A nationally-representative sample of US adults indicates higher prevalence of asymmetric hearing among men and older adult cohorts. There is currently no standard audiometric criterion for defining asymmetry, and prevalence estimates vary markedly depending on which audiometric criteria is used. Given the potentially high prevalence of asymmetry depending on criterion, clinicians should also consider other supplementary clinical data when recommending medical referral.
Keywords: Asymmetric hearing loss, Epidemiology, National Health and Nutrition Examination Survey, Prevalence
Asymmetric hearing can present itself in both medical and audiologic visits. Patients with asymmetric hearing are often referred to otolaryngologists to evaluate for middle ear and retrocochlear pathologies before audiologic rehabilitation with hearing aids or other strategies are considered. Although the finding of an asymmetry is significant from the perspective of clinical management, there are no population-based studies on the prevalence of asymmetric hearing among adults. Understanding the prevalence may be relevant for informing medical decisions and management protocols.
Audiometric evaluations provide valuable clinical data, however, there is currently no standard criterion for defining asymmetric hearing. In the present study, we applied two different audiometric definitions for asymmetry to estimate prevalence using nationally representative data from the United States. One definition comes from the American Academy of Otolaryngology–Head and Neck Surgery (AAO-HNS), whose position statement defines asymmetry as a difference in pure tone averages (PTA) that is greater than 15 dB between ears (1). The precise frequencies comprising the PTA are unspecified for this definition. The second definition was based on a working criterion currently used by audiologists in the US Department of Veterans Affairs (VA) when determining need for medical referral: Differences greater than/equal to 20 dB at two contiguous frequencies or greater than/equal to 10 dB at three contiguous frequencies between ears (2). While the former definition from the AAO-HNS is a more rigid definition of asymmetry based on a fixed average, the latter used by audiologists in the VA system is a more pragmatic definition that accounts for a broader frequency range. The current study proceeded to estimate the national prevalence of asymmetric hearing among adults through applying the two distinct audiometric criteria. A secondary aim explored prevalence differences across demographics.
METHODS
We analyzed audiometric data for adults aged 20 years and older across the 2001 to 2012 cycles of the National Health and Nutrition Examination Survey (NHANES), an ongoing and biannual epidemiological survey of a representative sample of the US noninstitutionalized population (3). Data from NHANES are publicly available and managed by the Centers for Disease Control and Prevention (CDC) (3). The analytic cohort (n = 6,190) comprised participants with complete audiometric data (0.5–8 kHz), those demonstrating test-retest reliability at 1 kHz within 10 dB, as well as tympanometric data recorded within normative ranges (4) to reduce the likelihood of including participants with potentially transient conductive hearing loss in our analyses. When no responses were indicated at the highest presentation level (120 dB), 125 dB was assigned accordingly.
For estimates using the AAO-HNS definition, we explored two different PTA calculations. One is defined by the World Health Organization (0.5–4 kHz) and the other is an average across high-frequencies (4–8 kHz). In applying the criterion used by VA audiologists, we analyzed thresholds across the full spectrum, 0.5 to 8 kHz. All analyses were conducted using a statistical software to identify participants meeting the audiometric criteria for asymmetric hearing with the two definitions (StataCorp. 2011. Stata Statistical Software: Release 12. College Station, TX: StataCorp LP.). We applied sampling weights in accordance with recommended guidelines (5) for reporting nationally representative data and stratified all estimated proportions by sex and age decade to show the prevalence of asymmetric hearing. In accordance with Johns Hopkins University School of Medicine policy, this investigation with public CDC data was exempt from review by the institutional review board (IRB).
RESULTS
Across both definitions, prevalence of asymmetric hearing was generally greater among men and increases with age (Table 1). Prevalence estimates varied markedly between the AAO-HNS and VA definitions. Overall estimates with criterion from the AAO-HNS were 2.77% (PTA: 0.5–4 kHz) and 9.46% (PTA: 4–8 kHz). When stratified by sex and age decade, prevalence estimates ranged 0.97 to 8.31% and 2.82 to 19.92% with PTA calculations from 0.5 to 4 kHz and 4 to 8 kHz, respectively. The overall prevalence estimated with a criterion used within the VA system was 25.05% and ranged 10.37 to 43.68% across demographics.
TABLE 1.
Prevalence of asymmetric hearing in the United States, by sex and age decade: National Health and Nutrition Examination Survey (NHANES),a 2001–2012 (n = 6,190)
| Asymmetric Hearing Per AAO-HNS Definitionb | Asymmetric hearing per working definition used by the VAc | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| PTA: 0.5–4 kHz | PTA: 4–8kHz | 0.5–8 kHz | |||||||
| Age (yrs) | % Overall (95% CI) | % Males (95% CI) | % Females (95% CI) | % Overall (95% CI) | % Males (95% CI) | % Females (95% CI) | % Overall (95% CI) | % Males (95% CI) | % Females (95% CI) |
| 20–29 | 1.32d (0.53, 2.11) | 1.66d (0.42, 2.91) | 0.97d (−0.94, 2.88) | 5.09 (3.64, 6.55) | 6.85 (4.14, 9.55) | 3.30 (0.85, 5.75) | 13.87 (11.37, 16.36) | 17.28 (13.35, 21.22) | 10.37 (6.65, 14.09) |
| 30–39 | 1.03 (0.30, 1.75) | 0.62d (−0.01, 1.26) | 1.41d (0.16, 2.67) | 5.31 (3.44, 7.19) | 7.91 (4.72, 11.11) | 2.82 (1.07, 4.58) | 17.93 (15.15, 20.72) | 22.19 (17.45, 26.92) | 13.85 (10.01, 17.69) |
| 40–49 | 2.75 (1.40, 4.11) | 3.57 (1.02, 6.11) | 2.08 (0.64, 3.52) | 9.23 (6.79, 11.66) | 14.02 (9.14, 18.90) | 5.27 (2.93, 7.60) | 25.84 (22.20, 29.48) | 34.33 (28.52, 40.13) | 18.83 (14.55, 23.11) |
| 50–59 | 4.51 (3.03, 5.99) | 4.61 (2.06, 7.16) | 4.41 (2.34, 6.49) | 13.76 (10.24, 17.28) | 19.04 (13.78, 24.30) | 9.02 (5.14, 12.91) | 33.37 (28.86, 37.89) | 42.99 (36.82, 49.17) | 24.74 (19.37, 30.12) |
| 60–69 | 4.05 (2.32, 5.79) | 4.95 (1.84, 8.06) | 3.33 (1.26, 5.40) | 13.97 (11.28, 16.66) | 19.92 (14.15, 25.69) | 9.20 (6.41, 11.98) | 35.52 (31.41, 39.64) | 43.07 (35.92, 50.23) | 29.47 (24.72, 34.22) |
| 70–79 | 5.43 (3.78, 7.48) | 8.31 (4.42, 12.19) | 3.45 (1.33, 5.57) | 16.66 (13.28, 20.03) | 19.50 (13.85, 25.15) | 14.70 (9.37, 20.03) | 38.64 (35.07, 42.20) | 43.68 (37.91, 49.45) | 35.16 (29.02, 41.30) |
| ≥80 | 5.12 (2.40, 7.84) | 7.92 (1.37, 14.48) | 3.50d (1.03, 5.98) | 14.87 (11.22, 18.53) | 14.93 (9.19, 20.67) | 14.84 (9.19, 20.50) | 33.68 (28.20, 39.17) | 42.27 (35.34, 49.20) | 28.74 (21.00, 36.48) |
| Overall | 2.77 (2.20, 3.34) | 9.46 (8.31, 10.61) | 25.05 (23.46, 26.64) | ||||||
NHANES is an ongoing, biannual epidemiological survey of a representative sample of the US noninstitutionalized population (3). Audiometric data collection follows a standardized operation across cycles. Inclusion criteria in analytic cohort include participants with recorded thresholds across 0.5 to 8 kHz, test-retest reliability at 1 kHz within 10 dB, and tympanometric data recorded within normative ranges (4); 125 dB was assigned when no responses were indicated at the highest presentation level (120 dB).
American Academy of Otolaryngology–Head/Neck Surgery’s (AAO-HNS) definition of asymmetry: difference more than 15 dB across pure tone averages (PTA) between ears (1).
Definition of asymmetry used by audiologists in the US Department of Veterans Affairs (VA) when determining need for medical referral: more than or equal to 20 dB at two contiguous frequencies or more than or equal to 10 dB at three contiguous frequencies between ears (2).
Prevalence estimates based on small sample size (n<10).
DISCUSSION
Prevalence estimates of asymmetric hearing differ substantially depending upon the audiometric definition used. Our findings also suggest that asymmetries are more prevalent across higher frequencies. When applying a definition from the AAO-HNS, the overall estimated prevalence is 2.77% (95% confidence interval: 2.20, 3.34) and 9.46% (8.31, 10.61) when using PTA’s across 0.5 to 4 and 4 to 8 kHz, respectively. These are lower than the overall prevalence estimates when applying the definition used by VA audiologists to determine recommendations for medical referral, 25.05% (23.46, 26.64). Differences in estimates between asymmetry definitions are likely attributed to the one used by VA audiologists accounting for a broader frequency range. These observations highlight that the lack of a standard audiometric definition for asymmetry contributes to highly varied prevalence estimates among adults in the US, which is consistent with a previous report investigating the prevalence of asymmetric hearing among children (6).
A limitation of the present investigation is the lack of bone conduction thresholds in NHANES. To the authors’ best awareness, there are currently no nationally representative datasets that include both air and bone conduction audiometric thresholds. This may contribute to prevalence estimates that are also inclusive of some asymmetric hearing attributed to conductive etiologies. Despite restricting our analyses to participants with tympanometric data recorded within the normative range to address this, we are limited in our ability to fully distinguish between sensorineural and conductive etiologies of asymmetry.
Nevertheless, a strength of the current investigation is the use of nationally representative data, which to the authors’ best knowledge provides a first report of the prevalence of asymmetric hearing among the noninstitutionalized US adult population. However, given that there is currently no universally agreed upon audiometric criterion and that prevalence estimates can be as high as 25.05% (23.46, 26.64), clinicians should reconsider relying on audiometric data in isolation and ultimately also consider other supplementary information (e.g., speech audiometry, case history, etc.) when determining recommendations for further medical evaluations. Future research may consider pursuing investigations that contribute to establishing a standard criterion for defining asymmetric hearing.
Funding:
J.J.S.: Supported by the Johns Hopkins School of Nursing. J.B.: None declared related to submitted work. N.S.R.: Supported by the Cochlear Center for Hearing and Public Health at the Johns Hopkins Bloomberg School of Public Health. J.A.D.: NIH/NIA K01AG054693. F.R.L.: None declared related to submitted work. A.M.G.: None declared related to submitted work. This manuscript is supported in part by funding from the Eleanor Schwartz Charitable Foundation and the Cochlear Center for Hearing and Public Health at the Johns Hopkins Bloomberg School of Public Health
Conflicts of Interest:
J.J.S. declares no conflicts. J.B. declares no conflicts. N.S.R. reports being a scientific board member (non-financial) for SHOEBOX Ltd., Ottawa, Ontario, Canada and a consultant to Helen of Troy. J.A.D. declares no conflicts. F.R.L. reports being a nonprofit board member for Access HEARS, a consultant for Boehringer Ingelheim and Amplifon, receiving speaker honoraria from Caption Call, receiving reimbursed travel from Cochlear Ltd., and being the director of a research center at the Johns Hopkins Bloomberg School of Public Health funded in part by a philanthropic gift from Cochlear Ltd.. A.M.G. reports being a consultant to Cochlear Ltd and Auditory Insight.
REFERENCES
- 1.American Academy of Otolaryngology-Head and Neck Surgery. Position Statement: Red Flags-Warning of Ear Disease. Available at: https://www.entnet.org//content/position-statement-red-flags-warning-ear-disease Accessed September 9, 2018.
- 2. Personal communication during expert audiologist consensus meeting convened for the planning of the Aging & Cognitive Health Evaluation in Elders Randomized Controlled Trial, Tampa, FL, Dec 2014, Participants: Harvey Abrams, Terry Chisolm, Cynthia Compton-Conley, Larry Humes, Rachel McArdle, Gurjit Singh, Helena Solodar, Michael Valente.
- 3.Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Available at: https://www.cdc.gov/nchs/nhanes/about_nhanes.htm. [Accessed July 2, 2018]. [Google Scholar]
- 4.Margolis RH, Heller JW. Screening tympanometry: criteria for medical referral. Audiology 1987;26:197–208. [DOI] [PubMed] [Google Scholar]
- 5.Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey Examination Protocol and Analytic Guidelines. Hyattsville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Available at: [https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx] [accessed July 2, 2018]. [Google Scholar]
- 6.Ross DS, Visser SN, Holstrum WJ, Qin T, Kenneson A. Highly variable population-based prevalence rates of unilateral hearing loss after the application of common case definitions. Ear Hear 2010;31:126–33. [DOI] [PubMed] [Google Scholar]