Hearing impairment is one of the most common permanent disabilities in the western world. Advanced age is the most important risk factor for acquired hearing loss, and the prevalence will increase with increasing longevity. Age-related hearing loss is typically symmetrical and most pronounced in the high frequencies. Such hearing loss is permanent, curative treatment is not available, and hearing aids are the only remedy.
Although hearing ability normally declines with age, there is great individual variation in age of onset, progression, and severity.1 Men are consistently reported to be more severely affected than women in terms of earlier age of onset and more rapid progression of hearing loss.2 The most important environmental risk factor for hearing loss is noise exposure. Any exposure to noise of significant intensity and duration can increase the risk of permanent hearing damage. Sources of noise exposure include leisure activities (such as hunting) and occupational environments (such as industrial work spaces, blasting, and explosions).3 However, persons exposed to the same amount of environmental noise can show different severities of hearing loss, indicating that individual susceptibility plays a role.4 Genetic influence on individual variation in hearing thresholds has been reported. In a family study based on audiometry data from 2311 people, heritability estimates were 0.26, 0.27, and 0.32 for low-, middle-, and high-frequency hearing threshold levels, respectively.5 In a Danish study based on twins aged 70 years and older, the heritability of self-reported hearing ability was estimated as 0.40 for both men and women.6 A Finnish study of female twins aged 63 to 76 years reported a heritability of 0.75 in hearing thresholds averaged over the frequencies 0.5–4 kHz in the better ear.7
The relative contribution of genetic effects in various age groups has not been fully explored because most previous studies have included a narrow age span, or samples were too small to stratify by age. However, invariant heritability of hearing loss throughout life cannot automatically be assumed. The sum of genetic as well as environmental effects may accumulate or decrease during a lifetime. The underlying biologic and environmental factors causing hearing impairment may also differ across age. Only one previous study has examined age-specific heritability. In that study, the relative importance of genes compared with environmental effects decreased with age, but the study included a relatively small sample of male twins, and data only on high frequency hearing (3, 4, 6, and 8 kHz).8
We explored the relative importance of genetic and environmental effects in the etiology of individual differences in hearing and the extent to which such effects might vary with age.
From August 1995 to June 1997, the total adult populations of the 24 municipalities of Nord-Trøndelag County, Norway, were invited to take part in a health-screening survey, the Nord-Trøndelag Health Study. The invitation list was based on population files administered by the governmental Statistics Norway. The Nord-Trøndelag Hearing Loss Study was included as an integrated project, with participants from 17 of the 24 municipalities.9 The Hearing Loss Study included pure-tone audiometry. Participants ranged in age from 20 to 101 years (mean = 50; SD = 17). In one municipality, Levanger, the subjects were reinvited to the hearing examination after the survey was finished. The participation rate for all municipalities except Levanger was 69% (65% among men and 72% among women); for Levanger, overall participation rate was only 42%. Altogether 51,574 participants, including 5110 from Levanger, attended the hearing examination and signed an informed consent. Audiometric data are missing for 1252 (2%) subjects, including data on approximately 300 subjects that were lost due to computer breakdown or other technical problems. Sibling relationships were identified through public registries administered by the governmental Statistics Norway. The sex distribution of sibling pairs is shown in Table 1. The study was approved by the Norwegian Regional Committee for Medical Research Ethics and the Data Inspectorate.
Air-conduction hearing thresholds were obtained by pure-tone audiometry with 5 Interacoustics AD25 automatic, self-administered audiometers with TDH-39 earphones and MX 41/AR cushions linked to a personal computer where the data were automatically stored. Hearing thresholds were determined in accordance with ISO 8253–1 (1989), with fixed frequencies using an automatic procedure. A maximum threshold of 120 dB was recorded, and threshold shifts exceeding this value (no response to the maximum signal) were treated as a 120 dB loss. The audiometric procedure included the standard frequencies 0.25, 0.5, 1, 2, 3, 4, 6, and 8 kHz. Semiportable, dismountable, sound-attenuation booths (Tegnèr T-booth 95 × 105 × 210 cm) were used in rooms specially selected to avoid background noise. Background noise was measured for a random sample of the rural examination rooms and all rooms in the urban areas. The results met the recommended standard for audiometric test administration (ISO8253–1, 1989), although background noise from shutting doors or walking in the room may have affected the low-frequency (0.25 and 0.5 Hz) results in some cases. Test-retest correlations for the audiometric results averaged over both ears were 0.82 for 0.25 kHz, 0.89 for 0.5 kHz, and 0.96–0.98 for the other frequencies.3 A more thorough description of the age distribution and audiometric data is described elsewhere.10 We used pure-tone average of the hearing thresholds at 0.5, 1, 2, and 4 kHz in the better ear as a measure of hearing acuity, as recommended by the World Health Organization.11 First, each hearing threshold measure was age-corrected by means of a linear-regression analysis. Age, age squared, and age cubed were entered as predictors in separate analyses with each of the 8 audiometric scores as dependent variables, and the residuals were kept as age-adjusted variables. The age-correction procedure was performed separately for men and women.
Model and Estimation
Similarity between siblings' hearing thresholds was assessed using polychoric correlations computed by means of the Mx software package.12 Polychoric correlations were calculated separately for sisters, brothers, and opposite-sex siblings to test for sex differences in the correlations. Sibling correlations were also calculated across 6 age strata (20–29, 30–39, 40–49, 50–59, 60–69, and 70 years and above) to test for age differences in the correlations. To test whether correlations were different by sex, we compared the likelihood of a saturated model in which correlations were allowed to be estimated independently across groups, to one in which they were constrained to be equal. To test whether correlations were different across age, we compared the likelihood of a model with a linear relationship between age and the magnitude of the correlations, to one where the correlations were constrained to be equal across age groups.
Heritability is a sample statistic representing the proportion of total phenotypic variance attributable to additive genetic differences, with values ranging from 0 to 1.0. For example, a heritability of 0.30 for a trait would indicate that 30% of the variation in that trait is explained by genetic differences between individuals. According to Mendelian laws of inheritance siblings share, on average, 50% of their segregating genes. This implies that they share half the phenotypic variation due to additively acting genes and one quarter of variance due to genetic dominance. It can be shown that, under most conditions, nonadditive genetic variance will amount to only a small fraction of total genetic variance for phenotypes influenced by many genes.13 Accordingly, assuming that siblings share 50% of genetic variance is usually a good approximation. Also, siblings by definition share the variance associated with environmental factors common to them, usually termed common environment or shared environment. Therefore, data from siblings alone do not allow separation of common environment and genetic effects. If we assume that common environmental effects do not contribute to similarity in siblings' hearing threshold level, then heritability can be estimated as twice the sibling correlation.
Pure-tone averages of hearing thresholds at 0.5, 1, 2, and 4 kHz are listed in Table 2. The data are stratified by age and displayed for men and women separately. Women had better hearing thresholds on average than men, and hearing ability declined with increasing age in both men and women.
The polychoric correlations of age-adjusted hearing thresholds' level between brothers, sisters, and opposite-sex siblings are shown in Table 1. The correlation for brothers (0.20 [95% confidence interval (CI) =0.17–0.24]) was slightly larger than the correlation for sisters (0.17 [0.14–0.21]).
Table 2 presents sibling correlations for 6 age strata. The correlations increase with increasing age, from 0.15 in the youngest age group (20–29 years) to 0.22 in the oldest (70+ years). The sibling correlation for the whole sample (including both sexes across all age groups) is 0.18 (0.16–0.20).
Based on the sibling correlations in Table 2, the upper limit of the heritability of hearing loss is 0.36.
Data from a population-based study, including audiometric data from 11,263 sibling pairs, show a substantial genetic contribution to individual variation in hearing thresholds.
Heritability estimates are time- and population-specific, and our results cannot automatically be generalized to another population or to future generations. Data from siblings alone do not allow the effect of environmental transmission from parents to offspring to be separated from genetic effects. If we assume that shared environmental effects do not contribute to similarity in siblings regarding hearing level, and that all genetic effect is additive, heritability can be estimated as twice the sibling correlation. Such estimates represent the upper limit of the heritability. If there is an undetected effect of family environment, then this heritability estimate will be somewhat inflated. However, previous studies do not suggest that effects due to shared familial environments are important to variation in hearing ability.6,7 A strength of the study is that the nonclinical nature of the sample allows both normal-hearing and hearing-impaired participants to contribute to the results.
Heritability estimates from previous studies range from 0.28 to 1.0.5–8,14–16 This wide range may be attributable to random fluctuation, and to differences in study design, measurement of hearing, statistical approaches, and study population characteristics. We find sibling correlations from 0.15 to 0.22, in accordance with the sibling correlations reported in the Framingham cohort5 but slightly lower than the sibling correlations reported by Raynor et al.14
Heritability estimates from twin studies are based on the difference between monozygotic (MZ) twins (who are genetically identical) and dizygotic (DZ) twins (who share, on average, 50% of their segregating genes). The sibling correlations in our study are within the range of the DZ twin correlations reported in a Swedish study of male twins7 and a Finnish study of female twins.8 Despite the similarity of DZ twin correlations from these twin studies and the sibling correlations from family studies, the twin studies reported higher heritability estimates. This could be due to nonadditive genetic effects, which inflate heritability estimates in twin studies but deflate the upper-limit heritability based on sibling data alone. When MZ correlations exceed the double value of the DZ correlations (which was the case in the twin studies of hearing), nonadditive genetic effects are indicated.
We find a trend of sibling correlations increasing with age. However, the numbers are relatively small, especially in the oldest age groups. The majority of observations are in the youngest ages where there is little variation in hearing ability. The only previous study examining age-specific heritability is the Swedish study of 250 MZ and 397 DZ male twin pairs aged 36–80 years. There the trend was opposite—a decrease of genetic influence with age.8 The hearing measure in that study was a mean of the high-frequency hearing thresholds (3, 4, 6, and 8 kHz) averaged over both ears. Like our sibling correlations, the DZ twin correlations increased with age from 0.13 in the youngest age group (35–45 years) to 0.28 in the oldest age group (65 years and older), whereas the MZ twin correlations decreased from 0.72 in the youngest age group to 0.52 on the oldest age group. This correlation pattern could indicate nonadditive genetic effects and not necessarily heritability decreasing with age.
We find only minor sex differences in the sibling correlations, with a slightly higher correlation in brothers. This is contrary to 2 recent family studies that reported higher sibling correlations in sisters.5,14 However, those studies were based on relatively small samples. In one study, the confidence intervals for the male and female sibling correlations are overlapping.14 Confidence intervals are not reported in the other study, but the difference in correlation estimates for brothers and sisters is small.5 A twin study with information on self-rated hearing abilities in male and female twin pairs reported no sex differences in heritability estimates.6
Our study supports previous work suggesting a substantial genetic contribution to individual variation in hearing. In a recent study of 70 candidate genes, only one gene was reported to be associated with age-related hearing loss.17 Several genetic regions have been identified that did not meet the criterion of genome-wide significance, and the findings were not replicated.18,19 Despite great optimism following the improvements of methods in gene finding, the genetic basis for the substantial heritability of hearing loss remains unknown. Population genetic studies are nonetheless important because they may facilitate the dissection of phenotypic variation and allow the interplay between genes and environment to be unraveled more clearly.
1. Mitchell P, Gopinath B, Wang JJ, et al.. Five-year incidence and progression of hearing impairment in an older population. Ear Hear. 2011;32:251–257.
2. Karlsmose B, Lauritzen T, Parving A. Prevalence of hearing impairment and subjective hearing problems in a rural Danish population aged 31–50 years. Br J Audiol. 1999;33:395–402.
3. Tambs K, Hoffman HJ, Borchgrevink HM, Holmen J, Engdahl B. Hearing loss induced by occupational and impulse noise: results on threshold shifts by frequencies, age and gender from the Nord-Trondelag Hearing Loss Study. Int J Audiol. 2006;45:309–317.
4. Davis RR, Kozel P, Erway LC. Genetic influences in individual susceptibility to noise: a review [review]. Noise Health. 2003;5:19–28.
5. Gates GA, Couropmitree NN, Myers RH. Genetic associations in age-related hearing thresholds. Arch Otolaryngol Head Neck Surg. 1999;125:654–659.
6. Christensen K, Frederiksen H, Hoffman HJ. Genetic and environmental influences on self-reported reduced hearing in the old and oldest old. J Am Geriatr Soc. 2001;49:1512–1517.
7. Viljanen A, Era P, Kaprio J, Pyykko I, Koskenvuo M, Rantanen T. Genetic and environmental influences on hearing in older women. J Gerontol A Biol Sci Med Sci. 2007;62:447–452.
8. Karlsson KK, Harris JR, Svartengren M. Description and primary results from an audiometric study of male twins. Ear Hear. 1997;18:114–120.
9. Tambs K, Hoffman HJ, Borchgrevink HM, Holmen J, Samuelsen SO. Hearing loss induced by noise, ear infections, and head injuries: results from the Nord-Trondelag Hearing Loss Study. Int J Audiol. 2003;42:89–105.
10. Engdahl B, Tambs K, Borchgrevink HM, Hoffman HJ. Screened and unscreened hearing threshold levels for the adult population: results from the Nord-Trondelag Hearing Loss Study. Int J Audiol. 2005;44:213–230.
11. Mathers C, Smith A, Concha M. Global burden of hearing loss in the year 2000. 2011. Available at: http://www.who.int/healthinfo/statistics/bod_hearingloss.pdf
12. Neale MC, Boker SM, Xie G, Maes HH. Mx: Statistical Modeling. 5 ed. VCU Box 9000126. Richmond, VA: Department of Psychiatry; 1999.
13. Falconer DS. Introduction to Quantitative Genetics. London: Longman; 1981.
14. Raynor LA, Pankow JS, Miller MB, et al.. Familial aggregation of age-related hearing loss in an epidemiological study of older adults. Am J Audiol. 2009;18:114–118.
15. Viljanen A, Kaprio J, Pyykko I, et al.. Genetic and environmental influences on hearing at different frequencies separately for the better and worse hearing ear in older women. Int J Audiol. 2007;46:772–779.
16. Wingfield A, Panizzon M, Grant MD, et al.. A twin-study of genetic contributions to hearing acuity in late middle age. J Gerontol A Biol Sci Med Sci. 2007;62:1294–1299.
17. Van Laer L, Van Eyken E, Fransen E, et al.. The grainyhead like 2 gene (GRHL2), alias TFCP2L3, is associated with age-related hearing impairment. Hum Mol Genet. 2008;17:159–169.
18. DeStefano AL, Gates GA, Heard-Costa N, Myers RH, Baldwin CT. Genomewide linkage analysis to presbycusis in the Framingham Heart Study. Arch Otolaryngol Head Neck Surg. 2003;129:285–289.
19. Huyghe JR, Van LL, Hendrickx JJ, et al.. Genome-wide SNP-based linkage scan identifies a locus on 8q24 for an age-related hearing impairment trait. Am J Hum Genet. 2008;83:401–407.
The Nord Trøndelag Health Study (HUNT), of which the Hearing Loss Study is a part, was conducted in collaboration with the National Institute of Public Health, Oslo, The National Health Screening, Oslo, Nord-Trøndelag County Council, and the Norwegian University of Technology and Science, Trondheim. The Nord-Trøndelag County Health Officer and the Community Health Officer in Levanger and in other municipalities provided organizational and other practical support. We also thank the NTHLS team for their diligence.