Philip Newall

... of Education and Arts, The University of Newcastle, NSW 2308, Australia, +61 2 4921 7347, +61 2 4921 7386 david.wigney@newcastle ... Stumer, Hickson and Worrall (1996) similarly found that a 95% prevalence of hearing impairment amongst elderly people living in residential ...

Tympanometric peak pressure, peak compensated static acoustic admittance (peak Ytm) and acoustic stapedius reflex (ASR) thresholds were obtained for a representative sample of 1565 older Australians who were participants in the Blue Mountains Hearing Study (BMHS). No significant age or gender effects were found for tympanometric peak pressure. Peak Ytm measures, however, decreased with age in the left ear only across all age groups and were consistently higher for men than for women. After allowing for hearing loss, the effect of age on ASR thresholds was inconsistent. An increase in ASR thresholds with age was observed at selected frequencies but only when measured contralaterally, and these changes were not clinically significant. Overall, our findings suggest that current normative data for peak Ytm is too restricted for application in the older population, but there is insufficient evidence to warrant alternative normative data for the ASR threshold range in this same population.

Prompted by a recent paper in this journal (Snik et al., 1995), this communication compares the gain and frequency response preferred by a group of 21 severely and profoundly hearing-impaired children (37 test ears) to the NAL-RP and the DSL 3.0 hearing aid prescriptions. The results support the findings of Snik et al. (1995) that the NAL-RP procedure provides an adequate prescription of amplification on average. The individual discrepancies between prescribed and preferred characteristics imply that fine-tuning with an evaluation tool is necessary for some children. It is uncertain whether previous listening experience affects listeners' preferences, and it is suggested that further studies be carried out to address this issue.

There have been few recent estimates of the prevalence of tinnitus from large population-based samples of older persons. Our study aimed to assess the prevalence and characteristics of prolonged tinnitus in a representative sample of 2015 adults aged 55-99 years, residing in the Blue Mountains, west of Sydney, Australia, during 1997-99. All participants underwent a detailed hearing examination by an audiologist, including comprehensive questions about hearing. After age adjustment, subjects reporting tinnitus had significantly worse hearing at both lower and higher frequencies (p < 0.001). This difference was more marked in younger than in older subjects (p < 0.05). Overall, 602 subjects (30.3%) reported having experienced tinnitus, with 48% reporting symptoms in both ears. Tinnitus had been present for at least 6 years in 50% of cases, and most (55%) reported a gradual onset. Despite tinnitus being described as mildly to extremely annoying by 67%, only 37% had sought professional help, and only 6% had received any treatment.

Explanations for associations found between sensory and cognitive function remain unclear. To assess in an older Australian population: (1) the correlation between sensory and cognitive function across groups with a narrow age range; (2) any independent association between sensory and cognitive impairment. We examined 3,509 non-institutionalised residents aged 50+ years in the second cross-sectional survey of the Blue Mountains Eye Study (1997-2000). Visual impairment was defined for best-corrected visual acuity (VA) < 6/12 in the better eye, moderate to severe hearing loss for hearing threshold > 40 dB (better ear), and likely cognitive impairment for Mini-Mental State Examination (MMSE) < 24 points. We found likely cognitive impairment in 3.3%, vision impairment in 2.7% and moderate to severe hearing loss in 10.5% of this population. Correlation between VA or hearing threshold and MMSE score increased with age. After adjusting for age, weak but significant correlation was evident in the normal ageing sample between vision and MMSE (r = 0.12 with vision items included and r = 0.11 with vision items excluded, both p < 0.0001), and between hearing thresholds and MMSE (r = -0.12, p < 0.0001). After adjusting for age, sex, education and history of stroke, persons with vision impairment had a lower mean MMSE score than those with normal vision, regardless of whether vision-related items were included (27.1 vs. 28.6, p < 0.001) or excluded (19.8 vs. 21.0, p < 0.001). Similarly, persons with moderate to severe hearing loss had a lower mean MMSE score than those without hearing loss (28.1 vs. 28.7, p < 0.001). Persons with likely cognitive impairment also had lower mean VA and higher mean hearing threshold than those without, after adjustment. We have documented an age-related correlation between sensory and cognitive function in a normal ageing sample. The association between sensory impairment and likely cognitive impairment remained significant after excluding vision-related MMSE items and adjusting for confounding factors. Our data suggest that age-related decline and the effect of visual impairment on the measurement of cognition only partly explain the association between sensory and cognitive impairments in older persons.

Microvascular disease may contribute to hearing loss. We tested this hypothesis by examining the relation of retinal microvascular abnormalities and hearing loss. This was a cross-sectional, population-based study. We performed retinal photography and pure-tone audiometry on 1511 older individuals (ages 54+ years) from the Blue Mountains Eye Study population. Examination of the retinal microvasculature provides a noninvasive means of assessing systemic microvascular changes associated with cardiovascular disease, especially since reliable methods of quantifying retinal microvascular abnormalities from retinal photographs have recently been developed. Trained graders masked to audiometry data assessed retinal photographs for signs of retinal microvascular damage, for example, focal arteriolar narrowing, arteriovenous nicking, and retinopathy (microaneurysms, retinal hemorrhages), and used a computer-assisted method to measure arteriolar and venular calibers. We defined moderate or profound hearing loss as pure-tone average air-conducted hearing thresholds in the better ear worse than 40 dB at 0.5, 1, 2, and 4 kHz. We used odds ratios (OR), 95% confidence intervals (CI), and probability values to present the associations between retinal microvascular abnormalities and hearing loss; an OR >1 indicates that persons with the microvascular abnormality are more likely to have hearing loss, a CI that includes 1.00 (e.g., 0.85 to 2.34) indicates no statistically significant association (i.e., the association may have occurred due to chance), whereas a value of p < 0.05 indicates that an association is unlikely to have occurred due to chance. The proportion of women and men with moderate or profound hearing loss was 10.4% (90/862) and 15.4% (100/649), respectively. After adjustment for age, sex, blood pressure, and other risk factors for hearing loss, retinopathy in women was associated with hearing loss (adjusted OR, 2.10; 95% CI, 1.09 to 4.06, p = 0.002) but not in men. Associations were stronger for moderate or worse low-frequency (0.25 to 1.0 kHz) hearing loss in women (adjusted OR, 3.00; 95% CI, 1.25 to 7.19, p = 0.0005) but were absent for high-frequency hearing loss. In men, no retinal microvascular abnormalities were associated with hearing loss at either low or high frequencies. Retinopathy, a sign of retinal microvascular damage, was associated with hearing loss in women but not in men. Our results provide modest support to the hypothesis that microvascular disease may play a role in age-related hearing loss in women, particularly low-frequency losses.

To assess the association between hearing impairment and health-related quality of life (HRQOL) in an older population, using the self-administered 36-item Short-Form Health Survey (SF-36). Participants of the Blue Mountains Hearing Study (BMHS, N = 2956) attended a comprehensive interview and hearing examination in which both self-reported and measured hearing impairments were assessed. Hearing impairment was defined as the pure-tone average of air-conduction hearing thresholds >25 decibels hearing level (dB HL) for the four frequencies (0.5 to 4.0 kHz) in the better ear. Of the 2431 participants with complete data (mean age, 67.0 yr), 1347 (55.4%) did not have measured hearing loss, whereas 324 (13.3%) had unilateral (285 mild, 22 moderate, 17 severe) and 760 (31.3%) had bilateral hearing impairment (478 mild, 207 moderate, 75 severe). After adjusting for demographic and medical confounders, bilateral hearing impairment was associated with poorer SF-36 scores in both physical and mental domains (fall in physical component score, PCS of 1.4 points, p = 0.025; fall in mental component score, MCS of 1.0 point, p = 0.13), with poorer scores associated with more severe levels of impairment (PCS p(trend) = 0.04, MCS p(trend) = 0.003). Participants with bilateral hearing impairment who habitually used hearing aids had a slightly better PCS (mean, 43.1; standard error [SE], 0.9) than those with the same impairment who did not have hearing aids or who only used them occasionally (mean, 41.2; SE 0.5), although this finding was not statistically significant (p = 0.055). Persons with self-reported hearing loss had significantly poorer HRQOL than corresponding persons without, but persons with unilateral or high-frequency hearing loss did not have significantly different HRQOL scores than their corresponding counterparts. This study quantifies the associated disease burden of age-related hearing impairment on health-related quality of life in a population-based cohort of older persons.

To identify potential and modifiable risk factors for tinnitus in a population of older adults. Cross-sectional study. Detailed questionnaires were interviewer-administered in a representative sample of 2015 persons aged 55+ yr, living in an area west of Sydney, Australia. Air- and bone-conduction audiometric thresholds were measured from 250 to 8000 Hz and from 500 to 4000 Hz, respectively. TEOAE and SOAE were measured for both ears. After adjusting for multiple variables in a Cox proportional hazards model, factors that significantly increased the risk of tinnitus were poorer hearing and cochlear function, self-reported work-related noise exposure, and history of middle ear or sinus infections, severe neck injury or migraine. Interventions aimed at reducing age-related hearing loss, particularly by reducing excessive work-related noise exposure, and the effective, timely treatment of ear-related infections, may all decrease the risk of tinnitus.