- Email: [email protected]
Visual and hearing impairment and retirement in older adults: A population-based cohort study
Accepted Manuscript Title: Visual and hearing impairment and retirement in older adults: a population-based cohort study Authors: Bamini Gopinath PhD Gerald Liew PhD George Burlutsky MApplStats Catherine M. McMahon PhD Paul Mitchell MD, PhD PII: DOI: Reference:
S0378-5122(17)30006-3 http://dx.doi.org/doi:10.1016/j.maturitas.2017.03.318 MAT 6796
To appear in:
Maturitas
Received date: Revised date: Accepted date:
2-1-2017 7-3-2017 10-3-2017
Please cite this article as: Gopinath B, Liew G, Burlutsky G, McMahon CM, Mitchell P, Visual and hearing impairment and retirement in older adults: a population-based cohort study, Maturitas (2017), http://dx.doi.org/10.1016/j.maturitas.2017.03.318 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights Older adults with moderate to severe hearing loss had greater odds of being retired
•
Use of a hearing aid was associated with significantly lower mean retirement age
•
Concurrent visual loss and hearing loss was linked to lower mean retirement age
Ac ce pt e
d
M
an
us
cr
ip t
•
1
Page 1 of 22
Visual and hearing impairment and retirement in older adults: a population-based cohort study
ip t
Running title: Sensory impairment and retirement
cr
Bamini Gopinath, PhD1 Gerald Liew, PhD1
us
George Burlutsky, MApplStats1 Catherine M McMahon, PhD2,3
1
an
Paul Mitchell, MD PhD1
Centre for Vision Research, Department of Ophthalmology and Westmead Millennium
M
Institute, University of Sydney, NSW, Australia.
HEARing Co-operative Research Centre, NSW, Australia.
3
Centre for Language Sciences, Linguistics Department, Macquarie University, Sydney,
Ac ce pt e
NSW, Australia.
d
2
Correspondence to: Bamini Gopinath
Centre for Vision Research
Westmead Hospital, Westmead, NSW, 2145, Australia. Telephone: 61-2-86273368 Fax: 61-2-98458345 Email: [email protected]
Abstract: 249; Text: 2495; Tables: 3
2
Page 2 of 22
Abstract Objective: Impaired vision and hearing are common among older adults and has been shown to reduce functional independence and adversely affect quality of life. This cohort study investigated the cross-sectional and temporal association between objectively measured dual
ip t
sensory impairment (DSI) and retirement from employment. Study design: 2409 Blue Mountains Eye Study participants aged 55+ years at baseline were
cr
included for analyses.
us
Main outcome measures: Visual impairment was defined as visual acuity less than 20/40 (better eye), and hearing impairment as average pure-tone air conduction threshold >25 dB
an
HL (500-4000 Hz, better ear). Employment status was questioned at each examination over 10 years.
M
Results: At baseline, 650 (27.0%) were employed and 1759 (73.0%) were retired. Crosssectional analysis showed that participants with moderate to severe hearing loss compared to
d
those with normal hearing had greater odds of being retired, multivariable-adjusted OR 1.96
Ac ce pt e
(95% CI 1.05-3.66). Participants with concurrent presenting visual impairment and moderate to severe hearing loss versus participants with no sensory loss had significantly lower mean retirement age, 57.1 versus 58.7 years (multivariable-adjusted p-value=0.04). Participants with any hearing loss at baseline had significantly higher odds of being retired by the 10-year follow-up, age-sex adjusted OR 1.82 (95% CI 1.00-3.30), this became marginally nonsignificant after adjusting for all other covariates, multivariable-adjusted OR 1.74 (95% CI 0.95-3.17). No associations were observed between DSI and incidence of retirement. Conclusions: Sensory impairment in older adults was independently associated with the decision to retire.
3
Page 3 of 22
Keywords: Blue Mountains Eye Study; dual sensory impairment; employment; hearing loss;
Ac ce pt e
d
M
an
us
cr
ip t
and vision loss.
4
Page 4 of 22
1. Introduction Impaired vision and hearing are common among older adults [1,2] and can occur separately or concurrently [3]. Visual loss adversely affects functional independence, mental health and cognition, reduces quality of life, and increases mortality risk [4-6]. Age-related
ip t
hearing loss is more frequent and is associated with a greater risk of depression, and impairs quality of life and the ability to conduct activities of daily living [2,7-12]. Moreover,
cr
previous literature shows that the presence of more than one sensory impairment increases
us
morbidity relative to single sensory impairments [13-15].
Hearing and vision difficulties hinder access to environmental information and may
an
also become an obstacle to communication [16]. Given that communication in the workplace is important, the presence of one or more sensory impairments could impact on an
M
individual’s ability to perform their work duties and these experiences could influence older workers' labour force participation [17]. Recently, the HUNT study found that that the degree
d
of objectively measured hearing loss and self-perceived hearing disability was associated
Ac ce pt e
with early retirement [18]. Conversely, the Epidemiology of Hearing Loss Study [17] showed that hearing impairment was not associated with the 15-year retirement rate. The Health and Retirement Study found that self-report of poor hearing or vision was linked to an earlier (~1year) than expected retirement age [19]. However, studies on the combined effect of vision and hearing loss on the decision to retire are currently lacking. In this cohort of adults aged 55 years and over, we aimed to: 1) Establish whether objectively measured hearing loss, self-perceived hearing handicap and/ or hearing aid use is independently associated with the prevalence and 10-year incidence of retirement from employment; and 2) Determine if dual sensory impairment (DSI) at baseline is independently associated with retirement 10 years later.
5
Page 5 of 22
2. Methods 2.1 Study sample The Blue Mountains Eye Study (BMES) is a population-based cohort study of common eye diseases and other health outcomes in a suburban Australian population located west of
ip t
Sydney. Study methods and procedures have been described elsewhere [20]. Baseline examinations of 3654 residents aged >49 years were conducted during 1992-4 (BMES-1,
cr
82.4% participation rate). Of the baseline participants, 2335 (75.1% of survivors) returned for
us
5-year follow-up examinations during 1997-9 (BMES-2), and 1952 participants (53.4% of the original cohort, or 76.6% of survivors) returned for 10-year follow-up examinations during
an
2002-4 (BMES-3). Hearing was measured at BMES-2 and BMES-3 only; i.e. in BMES-2, 2956 participants had audiometric testing performed. In this report, baseline is referring to
M
BMES-2. The study was approved by the Human Research Ethics Committee of the University of Sydney and was conducted adhering to the tenets of the Declaration of
Ac ce pt e
d
Helsinki. Signed informed consent was obtained from all participants at each examination.
2.2 Audiological assessment
Pure-tone audiometry was performed by audiologists in sound-treated booths, using TDH-39 earphones and Madsen OB822 audiometers (Madsen Electronics, Denmark). Bilateral hearing impairment was determined as the pure-tone average of audiometric hearing thresholds at 500,1000, 2000, and 4000 Hz (PTA0.5-4kHz) in the better ear, defining any hearing loss as PTA0.5-4kHz > 25 dB HL; mild hearing loss as PTA0.5-4kHz > 25-40 dB HL; and moderate to severe hearing loss as PTA0.5-4kHz >40 dB HL. An audiologist asked questions about history of any self-perceived hearing problem, and if a hearing aid had been provided. Specifically, participants were asked: ‘Do you or have you ever worn a hearing aid?’ (Yes/No/Don’t know). The Hearing Handicap Inventory for 6
Page 6 of 22
the Elderly - Shortened version (HHIE-S), developed by Ventry and Weinstein [21], was also administered. The HHIE-S includes 10 questions and a response of ‘yes’ is given 4 points; ‘sometimes’ is given 2 points, and ‘no’ is given 0 points. Scores ≥8 were taken to indicate the
ip t
presence of a hearing handicap [22].
2.3 Assessment of visual impairment
cr
Monocular distance logMAR (logarithm of the minimum angle of resolution) visual
us
acuity was measured with forced-choice procedures using a retroilluminated chart with automatic calibration to 85 cd/m2 (Vectorvision CSV-100TM; Vectorvision Inc, Dayton,
an
Ohio) according to the Early Treatment Diabetic Retinopathy Study protocol [20]. This was conducted with habitual correction (presenting visual acuity, with current eyeglasses, if worn)
M
and after subjective refraction (best-corrected visual acuity). For each eye, visual acuity was recorded as the number of letters read correctly from 0 to 70. For the present study, any
d
visual impairment was defined as presenting or best-corrected visual acuity of the better eye
Ac ce pt e
less than 39 letters (<20/40). DSI was defined as concurrent visual (either presenting or bestcorrected) and hearing impairment, as determined using the above definitions.
2.4 Assessment of retirement status
Employment status was determined from the question: ‘Are you retired or still employed?’ For incidence analysis, we included those participants who reported being employed at baseline, and if any of these participants retired at the follow-up examinations then they were considered as an incident (new) case of retirement. If they reported being ‘retired’ then they were asked ‘If retired, how old were you when you retired?’ They were asked to provide their retirement age in years. Participants were also asked whether they receive a pension and if they answered ‘yes’ they were asked what type of pension did they 7
Page 7 of 22
receive with the following options: age pension, veteran pension, invalid pension, or blind pension.
2.5 Assessment of covariates
ip t
A face-to-face interview with trained interviewers was conducted, and comprehensive data including information about medical history, demographic factors, socio-economic
cr
characteristics, and lifestyle behaviours were obtained from all participants. The medical
us
history included cardiovascular or other systemic disease, and medications used. A past history of angina, diabetes, myocardial infarction, and stroke was determined by responses to
an
a question: “Has a doctor advised you that you have any of the following conditions?” Selfrated health was assessed by asking: “For somebody your age, would you say your health is
2.6 Statistical analysis
d
M
excellent, very good, good, fair, or poor?” Low self-rated health was defined as fair or poor.
Ac ce pt e
SAS statistical software performed all analysis (v9.2, SAS Institute, NC), including ttests, χ2-tests and linear regression. The cross-sectional association between hearing related measures and DSI with retirement status was examined using logistic regression models to estimate odds ratios (OR) and 95% confidence intervals (CI). Multiple regression models were first adjusted for age (entered as a continuous variable) and sex, and then further adjusted for confounders that were found to be significantly associated with retirement status i.e. self-rated health, presence of ≥3 chronic conditions (e.g. hypertension, diabetes, angina, and stroke). Analysis of covariance was used to calculate differences in mean retirement age between those with and without sensory impairment(s), after adjusting for the aforementioned covariates. For longitudinal analysis of baseline sensory impairment and 10-year incidence of retirement we used Cox discrete-time modelling. 8
Page 8 of 22
3. Results 3.1 Hearing-related measures and retirement status Of the 2956 participants who had hearing data at baseline (i.e. at BMES-2), 547 did not
ip t
have information on visual acuity and/or retirement status and thus, were excluded, leaving 2409 participants for cross-sectional analysis. Table 1 shows the study characteristics of
cr
participants included in cross-sectional analysis. At baseline, 650 (27.0%) were employed
us
and 1759 (73.0%) were retired. The majority were receiving the age pension. Just under a third had any hearing impairment while over 1 in 5 participants had presenting visual
an
impairment (Table 1). Cross-sectional data showed that among those with any hearing impairment, 697 (89.2%) were retired compared to 1062 (65.2%) with no hearing loss who
M
were retired (p <0.0001). The highest retirement frequency was observed among participants with moderate to severe hearing loss, 93.6% (n=221). In multiple logistic regression analysis
d
(Table 2), participants with moderate to severe hearing loss versus those with normal hearing
Ac ce pt e
had 96% higher likelihood of being retired. Non-significant associations were observed between self-perceived hearing handicap and hearing aid use and the likelihood of retirement (Table 2).
Analysis of baseline data showed that those with moderate to severe hearing loss had a significantly lower adjusted mean retirement age of 57.2 years compared to a mean retirement age of 58.6 years among those with no hearing loss (multivariable-adjusted pvalue = 0.009). We found no significant differences in adjusted mean retirement ages between those with and without self-perceived hearing handicap (data not shown). However, persons who used a hearing aid versus those who did not use an aid had a significantly lower adjusted mean retirement age, 57.3 versus 58.7 years, multivariable-adjusted p=0.004.
9
Page 9 of 22
Longitudinal or incidence analysis (using Cox discrete-time modelling) included all participants who reported being employed at baseline and who also had complete sensory information i.e. 650. Of the 650 who were employed at baseline, 455 were followed-up 10 years later. Participants who were followed up versus those lost-to follow up, had similar
ip t
proportions of hearing loss (17.3% versus 18.2%, p=0.90), presenting visual impairment (11.1% versus 18.2%, p=0.23) and best corrected visual impairment (2.9% versus 3.0%,
cr
p=0.98) at baseline. Of those who were followed up, 190 participants (41.8%) retired during
us
the 10-year follow-up period. After adjusting for age and sex, participants with any hearing loss had significantly higher odds of being retired by the 10-year follow-up, OR 1.82 (95%
an
CI 1.00-3.30), but this became marginally non-significant after adjusting for all other covariates, multivariable-adjusted OR 1.74 (95% CI 0.95-3.17). No other associations were
Ac ce pt e
3.2 DSI and retirement status
d
retirement.
M
observed between severity of hearing loss, hearing handicap or hearing aid use and incident
Participants with both presenting visual impairment and any level of hearing loss had the highest retirement rate i.e. 94.5% (n=257) compared to those with no presenting visual impairment and normal hearing function – 62.5% (n=830). Participants with no presenting visual impairment but who had moderate to severe hearing loss had greater odds of reporting retirement compared to those with no visual and hearing impairment (Table 3), multivariableadjusted OR 2.52 (95% CI 1.13-5.64). Similarly, participants with no best-corrected visual impairment but with moderate to severe hearing loss had increased likelihood of reporting retirement compared to those with no sensory loss (Table 4), multivariable-adjusted OR 2.15 (95% CI 1.10-4.21). No other significant associations were observed between best-corrected visual impairment and/or hearing loss and retirement (Table 4). 10
Page 10 of 22
Analysis of covariance (ANCOVA) showed that participants who had no presenting visual impairment but who had moderate to severe hearing loss had significantly earlier mean retirement age compared to those with no sensory loss, 57.1 versus 58.7 years (multivariableadjusted p value=0.02). Similarly, those with concurrent presenting visual impairment and
ip t
moderate to severe hearing loss versus participants with no hearing or vision loss had significantly earlier mean retirement age, 57.1 versus 58.7 years (multivariable-adjusted p
cr
value=0.04). No significant associations were observed between best-corrected visual
us
impairment and/or hearing loss and adjusted mean retirement age (data not shown). Additionally, no significant associations were observed between DSI and 10-year incident
M
4. Discussion
an
retirement (data not shown).
The present study provides new insight into the societal impacts of sensory
d
impairments among older adults. We show that older adults with moderate to severe hearing
Ac ce pt e
impairment compared to those with normal hearing had a greater likelihood of being retired at baseline examination. Further, those who used a hearing aid reported a significantly lower adjusted mean retirement age (~1-year). Older adults who had concurrent presenting visual impairment and moderate to severe hearing loss compared to those with no impairment had lower adjusted mean retirement age. Hearing loss at baseline showed a marginally nonsignificant association with the decision to retire by the 10-year follow-up. Study participants with moderate to severe hearing loss (>40 dB HL) versus those with normal hearing not only had a higher likelihood of reporting retirement, but also retired around one to two years earlier. These data concur with other studies which have shown that the degree of acquired hearing impairment is associated with earlier retirement from the labour force [18,23]. Exchange of information which constitutes an important aspect of the 11
Page 11 of 22
work environment, can be seriously impaired in individuals with severe hearing impairment [24], thus, influencing a person’s decision to retire early. Further, we show that hearing aid use was associated with earlier retirement age. Hearing aid use is a marker of more severely impaired hearing, hearing handicap (or high communication needs), or potentially a marker
ip t
of frailty and aging [25], thus, explaining its association with earlier retirement. Our findings are not unexpected given that hearing is an important component of the
cr
face-to-face communicative and interactive processes that occur in the workplace setting
us
[17]. Despite this, there is evidence that workers with severe hearing impairment, defined as having difficulty hearing normal conversations even while wearing a hearing aid, were
an
approximately half as likely to report receiving employer-initiated accommodations (e.g. telephone amplifiers, FM or infrared sound systems) compared to workers with other
M
disabilities [17,26]. One potential reason could be that ‘denial’ or non-acceptance of hearing loss, and the stigma associated with impaired hearing [27], could lead to reluctance to
Ac ce pt e
relevant accommodations [17].
d
disclose a hearing loss and without this disclosure, the employer may not be able to provide
In our study, we found that older adults who had moderate to severe hearing loss and no visual impairment (either best-corrected or presenting) had approximately 2-fold higher odds of reporting retirement at baseline. In regards to retirement age, those with moderate to severe hearing loss reported an earlier retirement age regardless of whether or not they had accompanying presenting visual impairment. Therefore, these findings suggest that the crosssectional association with early retirement appears to be due to the presence of hearing loss rather than vision impairment. A potential explanation for this could be that those who had vision impairment may have already retired (i.e. before BMES-2) or had corrected their impairment (e.g. cataract surgery) so this was not a significant factor influencing workplace participation in our cohort, whereas, hearing loss may be more gradual in nature in most 12
Page 12 of 22
individuals, and is typically diagnosed and managed several years after its onset, often after having led to multiple negative consequences including effects on employment. We observed that participants with any hearing loss at baseline had 82% greater likelihood of retiring by the 10-year follow-up after adjusting for age and sex, however, after
ip t
adjusting for other factors such as visual impairment and self-rated health this association was attenuated. We had very small numbers of people who retired over the 10 years; hence, it
cr
is likely that we had inadequate power to detect a stronger association between hearing loss
us
or DSI and 10-year retirement rates in multivariable analysis. Therefore, other larger cohort studies with adequate follow-up are warranted to assess the longitudinal association between
an
objectively measured sensory loss and retirement rates.
Sensory difficulties are common experiences within older age groups; as a result they
M
are often overlooked or dismissed [13]. However, hearing loss interferes with so many of life’s activities, that it may prove to be a major impediment to society’s need to have people
d
remain longer in the workforce as the proportion of ‘working age’ people in developed
Ac ce pt e
countries like Australia shrinks [28]. Moreover, the annual cost of lost earnings due to workplace separation and early retirement from hearing loss was estimated at $6.7 billion in Australia [29]. Together these findings suggest that moderate to severe hearing loss could be a key factor influencing the decision to retire, hence, correction of hearing impairment, including the provision of assistive devices (e.g., FM systems) and rehabilitative services including hearing aid provision such as auditory and communication training [14,30] should be encouraged by clinicians, as it could improve work opportunities for hearing impaired individuals. This study has several strengths, including its prospective design, representative cohort, and use of objective, standardized audiometric and vision testing.. However, it is also important to discuss some of the study limitations. First, there could have been survival bias 13
Page 13 of 22
in our study, that is, persons who did not attend follow-up visits could have been more likely to have had sensory loss, resulting in an under-reporting event rate of retirement and potential underestimation of the associations. Second, while we adjusted for several confounders we did not collect information on contextual factors e.g. employer or organisation characteristics,
ip t
which could have influenced observed associations. Finally, our study analyzed data collected ~10 years ago, and we cannot discount the possibility that our findings might not be
cr
entirely generalizable to the present day, and would require more contemporary cohorts to
us
validate and confirm our observed associations. Nevertheless, there is a dearth of
epidemiological studies that have assessed the association between concurrent vision loss and
an
hearing loss and retirement status, thus, our study findings is a valuable addition to the existing published literature.
M
In summary, we show that older adults who had moderate to severely impaired hearing and/ or who were using a hearing aid (versus those not using a hearing aid) had greater odds
d
of being retired at baseline. Presence of coexisting visual and hearing impairment at baseline
Ac ce pt e
did not independently influence the decision to retire over the 10 years. Therefore, as the workforce in many developed countries ages the impact of sensory impairment, particularly hearing loss, on workplace separation and early retirement warrants continued attention.
Author contributions:
Conceived and designed the experiments: BG and PM; Performed the experiments: BG and PM; Analyzed and interpreted the data: BG, GL, CM, GB, and PM; Contributed reagents, materials, analysis tools or data: PM and GB; and Wrote the paper: BG. Funding: The Blue Mountains Eye and Hearing Study were supported by the Australian National Health and Medical Research Council (Grant Nos. 974159, 991407, 211069, 262120). The authors also acknowledge the financial support of the NHMRC Partnerships for 14
Page 14 of 22
Better Health (grant no. 571449) and the HEARing CRC, established and supported under the Australian Government's Cooperative Research Centres Program Conflict of interest: All authors have no competing interests to declare.
ip t
References [1] Taylor HR, Keeffe JE, Vu HT, Wang JJ, Rochtchina E, Pezzullo ML et al. Vision loss
cr
in Australia. Med J Aust 2005;182:565-68.
us
[2] Gopinath B, Rochtchina E, Wang JJ, Schneider J, Leeder SR, Mitchell P. Prevalence of age-related hearing loss in older adults: Blue Mountains Study. Arch Intern Med
an
2009;169:415-16.
[3] Schneider JM, Gopinath B, McMahon CM, Leeder SR, Mitchell P, Wang JJ. Dual
M
Sensory Impairment in Older Age. J Aging Health 2011.
[4] Chia EM, Mitchell P, Ojaimi E, Rochtchina E, Wang JJ. Assessment of vision-related
d
quality of life in an older population subsample: the blue mountains eye study.
Ac ce pt e
Ophthalmic Epidemiol 2006;13:371-77.
[5] Tay T, Wang JJ, Kifley A, Lindley R, Newall P, Mitchell P. Sensory and cognitive association in older persons: findings from an older Australian population. Gerontology 2006;52:386-94.
[6] Cugati S, Cumming RG, Smith W, Burlutsky G, Mitchell P, Wang JJ. Visual impairment, age-related macular degeneration, cataract, and long-term mortality: the Blue Mountains Eye Study. Arch Ophthalmol 2007;125:917-24.
[7] Gopinath B, Wang JJ, Schneider J, Burlutsky G, Snowdon J, McMahon CM et al. Depressive symptoms in older adults with hearing impairments: the Blue Mountains Study. J Am Geriatr Soc 2009;57:1306-8.
15
Page 15 of 22
[8] Gopinath B, Schneider J, Hickson L, McMahon CM, Burlutsky G, Leeder SR et al. Hearing handicap, rather than measured hearing impairment, predicts poorer quality of life over 10 years in older adults. Maturitas 2012;72:146-51. [9] Schneider J, Gopinath B, Karpa MJ, McMahon CM, Rochtchina E, Leeder SR et al.
ip t
Hearing loss impacts on the use of community and informal supports. Age Ageing 2010;39:458-64.
cr
[10] Schneider JM, Gopinath B, McMahon CM, Leeder SR, Mitchell P, Wang JJ. Dual
us
sensory impairment in older age. J Aging Health 2011;23:1309-24.
[11] Gopinath B, Schneider J, McMahon CM, Teber E, Leeder SR, Mitchell P. Severity of
an
age-related hearing loss is associated with impaired activities of daily living. Age Ageing 2012;41:195-200.
M
[12] Gopinath B, Schneider J, McMahon CM, Burlutsky G, Leeder SR, Mitchell P. Dual sensory impairment in older adults increases the risk of mortality: a population-based
d
study. PLoS ONE 2013;8:e55054.
Ac ce pt e
[13] Crews JE, Campbell VA. Vision impairment and hearing loss among communitydwelling older Americans: implications for health and functioning. Am J Public Health 2004;94:823-29.
[14] Lee DJ, Gomez-Marin O, Lam BL, Zheng DD, Arheart KL, Christ SL et al. Severity of concurrent visual and hearing impairment and mortality: the 1986-1994 National
Health Interview Survey. J Aging Health 2007;19:382-96.
[15] Lee DJ, Lam BL, Gomez-Marin O, Zheng DD, Caban AJ. Concurrent hearing and visual impairment and morbidity in community-residing adults: the National Health Interview Survey, 1986 to 1996. J Aging Health 2005;17:531-46.
16
Page 16 of 22
[16] Viljanen A, Tormakangas T, Vestergaard S, Andersen-Ranberg K. Dual sensory loss and social participation in older Europeans. European Journal of Ageing 2014;11:15567. [17] Fischer ME, Cruickshanks KJ, Pinto A, Klein BE, Klein R, Dalton DS. Hearing
ip t
impairment and retirement. J Am Acad Audiol 2014;25:164-70.
HUNT study. Eur J Public Health 2013;23:617-22.
cr
[18] Helvik AS, Krokstad S, Tambs K. Hearing loss and risk of early retirement. The
us
[19] Dwyer DS, Mitchell OS. Health problems as determinants of retirement: are self-rated measures endogenous? J Health Econ 1999;18:173-93.
an
[20] Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:357-64.
M
[21] Ventry IM, Weinstein BE. Identification of elderly people with hearing problems. ASHA 1983;??:37-42.
d
[22] Cruickshanks KJ, Wiley TL, Tweed TS, Klein BE, Klein R, Mares Perlman JA et al.
Ac ce pt e
Prevalence of hearing loss in older adults in Beaver Dam, Wisconsin. The
Epidemiology of Hearing Loss Study. Am J Epidemiol 1998;148:879-86.
[23] Kramer SE. Hearing impairment, work, and vocational enablement. Int J Audiol 2008;47 Suppl 2:S124-S130.
[24] Dalton DS, Cruickshanks KJ, Klein BE, Klein R, Wiley TL, Nondahl DM. The impact of hearing loss on quality of life in older adults. Gerontologist 2003;43:661-
68.
[25] Gopinath B, McMahon CM, Burlutsky G, Mitchell P. Hearing and vision impairment and the 5-year incidence of falls in older adults. Age Ageing 2016;45:409-14.
17
Page 17 of 22
[26] Zwerling C, Whitten PS, Sprince NL, Davis CS, Wallace RB, Blanck P et al. Workplace accommodations for people with disabilities: National Health Interview Survey Disability Supplement, 1994-1995. J Occup Environ Med 2003;45:517-25. [27] Wallhagen MI. The stigma of hearing loss. Gerontologist 2010;50:66-75.
ip t
[28] Dillon H. The 2006 Libby Harricks Memorial Oration: Hearing loss: the silent epidemic. Who, why, impact and what can we do about it. 2006. ACT, Deafness
cr
Forum Limited.
us
[29] Listen Hear! The economic impact and cost of hearing loss in Australia. CRC for Cochlear Implant and Hearing Aid Innovation, Vicdeaf, editors. 2006. Australia,
an
Access Economics Pty Limited.
[30] Heine C, Browning CJ. Communication and psychosocial consequences of sensory
Ac ce pt e
d
2002;24:763-73.
M
loss in older adults: overview and rehabilitation directions. Disabil Rehabil
18
Page 18 of 22
Table 1. Baseline characteristics of study participants and non-participants
67.2 (8.9)
Male sex
1160 (48.2)
Low self-rated health
444 (18.5)
Presence of ≥3 comorbidities
260 (11.0)
Employed
650 (27.0)
Retired
1759 (73.0)
cr
Age, yrs
ip t
Participants (n=2409)
us
Characteristics
Type of pension
Invalid
67 (5.0)
Veteran
223 (16.6)
781 (32.4)
Ac ce pt e
Any hearing loss (>25 dB HL)
1 (0.1)
d
Blind
an
940 (69.9)
M
Age
Best corrected visual impairment (<20/40) Presenting visual impairment (<20/40)
250 (10.4) 569 (23.7)
Data are presented as mean (SD) or n (%).
19
Page 19 of 22
Table 2. Cross-sectional association between measured hearing loss, hearing handicap inventory for elderly (HHIE) scores, and hearing aid use and retirement status (n=2409) Prevalence of retirement, OR (95% CI) Age-sex adjusted
Multivariate-adjusted a
No hearing loss (≤25 dB HL), n=1628
1.0 (reference)
1.0 (reference)
Any hearing loss (>25 dB HL), n=781
1.23 (0.90-1.68)
cr
Severity of hearing loss b No hearing loss (≤25 dB HL), n=1628
Moderate to severe hearing loss (>40 dB HL),
Ac ce pt e
Yes, n=253
d
No, n=2150
M
No handicap (HHIE <8), n=1476
Hearing aid use
1.06 (0.75-1.50)
1.04 (0.73-1.47)
an
Severity of hearing handicap
Severe handicap (HHIE ≥26), n=142
1.0 (reference)
2.00 (1.08-3.71)
n=221
1.20 (0.87-1.65)
1.0 (reference)
us
Mild hearing loss (26-40 dB HL), n=545
Moderate handicap (HHIE 8-24), n=723
ip t
Presence of hearing loss b
1.96 (1.05-3.66)
1.0 (reference)
1.0 (reference)
1.17 (0.90-1.53)
1.13 (0.86-1.49)
1.60 (0.91-2.81)
1.52 (0.86-2.67)
1.0 (reference)
1.0 (reference)
1.57 (0.89-2.77)
1.53 (0.86-2.73)
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, best-corrected visual impairment and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
20
Page 20 of 22
cr us
an
Table 3. Cross-sectional association between objectively measured dual sensory impairment and retirement status among Blue Mountains Eye
Prevalence of retirement, OR (95% CI)
Hearing impairment (dB HL)
Age-sex adjusted
Multivariate-adjusted a
No visual impairment
≤25 (No hearing loss), n=1328
1.0 (reference)
1.0 (reference)
>25 (Any hearing loss), n=506
1.21 (0.86-1.71)
1.18 (0.83-1.68)
1.02 (0.71-1.48)
1.00 (0.68-1.45)
2.50 (1.13-5.51)
2.52 (1.13-5.64)
≤25 (No hearing loss), n=297
1.13 (0.78-1.62)
1.09 (0.75-1.58)
>25 (Any hearing loss), n=272
1.43 (0.76-2.67)
1.39 (0.74-2.63)
>25-≤40 (Mild hearing loss), n=173
1.44 (0.65-3.20)
1.43 (0.63-3.22)
1.41 (0.54-3.68)
1.34 (0.52-3.50)
ed
Visual impairment (<20/40)
M
Study participants (n=2403)
>25-≤40 (Mild hearing loss), n=371
pt
>40 (Moderate-severe hearing
Ac
ce
Presenting
loss), n=135
>40 (Moderate-severe hearing loss), n=99
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
21
Page 21 of 22
cr us
an
Table 4. Cross-sectional association between measured dual sensory impairment, and retirement status among Blue Mountains Eye Study
Prevalence of retirement, OR (95% CI)
Hearing impairment (dB HL)
Age-sex adjusted
Multivariate-adjusted a
No visual impairment
≤25 (No hearing loss), n=1524
1.0 (reference)
1.0 (reference)
>25 (Any hearing loss), n=635
1.23 (0.89-1.70)
1.21 (0.87-1.67)
1.05 (0.73-1.49)
1.03 (0.72-1.46)
2.17 (1.12-4.21)
2.15 (1.10-4.21)
≤25 (No hearing loss), n=104
1.31 (0.68-2.51)
1.30 (0.67-2.53)
>25 (Any hearing loss), n=146
1.48 (0.53-4.18)
1.42 (0.50-4.06)
>25-≤40 (Mild hearing loss), n=93
1.66 (0.42-6.49)
1.60 (0.40-6.38)
1.26 (0.26-5.99)
1.19 (0.25-5.80)
ed
Visual impairment (<20/40)
M
participants (n=2409)
>25-≤40 (Mild hearing loss), n=452
pt
>40 (Moderate-severe hearing
Ac
ce
Best corrected
loss), n=183
>40 (Moderate-severe hearing loss), n=53
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
22
Page 22 of 22
S0378-5122(17)30006-3 http://dx.doi.org/doi:10.1016/j.maturitas.2017.03.318 MAT 6796
To appear in:
Maturitas
Received date: Revised date: Accepted date:
2-1-2017 7-3-2017 10-3-2017
Please cite this article as: Gopinath B, Liew G, Burlutsky G, McMahon CM, Mitchell P, Visual and hearing impairment and retirement in older adults: a population-based cohort study, Maturitas (2017), http://dx.doi.org/10.1016/j.maturitas.2017.03.318 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights Older adults with moderate to severe hearing loss had greater odds of being retired
•
Use of a hearing aid was associated with significantly lower mean retirement age
•
Concurrent visual loss and hearing loss was linked to lower mean retirement age
Ac ce pt e
d
M
an
us
cr
ip t
•
1
Page 1 of 22
Visual and hearing impairment and retirement in older adults: a population-based cohort study
ip t
Running title: Sensory impairment and retirement
cr
Bamini Gopinath, PhD1 Gerald Liew, PhD1
us
George Burlutsky, MApplStats1 Catherine M McMahon, PhD2,3
1
an
Paul Mitchell, MD PhD1
Centre for Vision Research, Department of Ophthalmology and Westmead Millennium
M
Institute, University of Sydney, NSW, Australia.
HEARing Co-operative Research Centre, NSW, Australia.
3
Centre for Language Sciences, Linguistics Department, Macquarie University, Sydney,
Ac ce pt e
NSW, Australia.
d
2
Correspondence to: Bamini Gopinath
Centre for Vision Research
Westmead Hospital, Westmead, NSW, 2145, Australia. Telephone: 61-2-86273368 Fax: 61-2-98458345 Email: [email protected]
Abstract: 249; Text: 2495; Tables: 3
2
Page 2 of 22
Abstract Objective: Impaired vision and hearing are common among older adults and has been shown to reduce functional independence and adversely affect quality of life. This cohort study investigated the cross-sectional and temporal association between objectively measured dual
ip t
sensory impairment (DSI) and retirement from employment. Study design: 2409 Blue Mountains Eye Study participants aged 55+ years at baseline were
cr
included for analyses.
us
Main outcome measures: Visual impairment was defined as visual acuity less than 20/40 (better eye), and hearing impairment as average pure-tone air conduction threshold >25 dB
an
HL (500-4000 Hz, better ear). Employment status was questioned at each examination over 10 years.
M
Results: At baseline, 650 (27.0%) were employed and 1759 (73.0%) were retired. Crosssectional analysis showed that participants with moderate to severe hearing loss compared to
d
those with normal hearing had greater odds of being retired, multivariable-adjusted OR 1.96
Ac ce pt e
(95% CI 1.05-3.66). Participants with concurrent presenting visual impairment and moderate to severe hearing loss versus participants with no sensory loss had significantly lower mean retirement age, 57.1 versus 58.7 years (multivariable-adjusted p-value=0.04). Participants with any hearing loss at baseline had significantly higher odds of being retired by the 10-year follow-up, age-sex adjusted OR 1.82 (95% CI 1.00-3.30), this became marginally nonsignificant after adjusting for all other covariates, multivariable-adjusted OR 1.74 (95% CI 0.95-3.17). No associations were observed between DSI and incidence of retirement. Conclusions: Sensory impairment in older adults was independently associated with the decision to retire.
3
Page 3 of 22
Keywords: Blue Mountains Eye Study; dual sensory impairment; employment; hearing loss;
Ac ce pt e
d
M
an
us
cr
ip t
and vision loss.
4
Page 4 of 22
1. Introduction Impaired vision and hearing are common among older adults [1,2] and can occur separately or concurrently [3]. Visual loss adversely affects functional independence, mental health and cognition, reduces quality of life, and increases mortality risk [4-6]. Age-related
ip t
hearing loss is more frequent and is associated with a greater risk of depression, and impairs quality of life and the ability to conduct activities of daily living [2,7-12]. Moreover,
cr
previous literature shows that the presence of more than one sensory impairment increases
us
morbidity relative to single sensory impairments [13-15].
Hearing and vision difficulties hinder access to environmental information and may
an
also become an obstacle to communication [16]. Given that communication in the workplace is important, the presence of one or more sensory impairments could impact on an
M
individual’s ability to perform their work duties and these experiences could influence older workers' labour force participation [17]. Recently, the HUNT study found that that the degree
d
of objectively measured hearing loss and self-perceived hearing disability was associated
Ac ce pt e
with early retirement [18]. Conversely, the Epidemiology of Hearing Loss Study [17] showed that hearing impairment was not associated with the 15-year retirement rate. The Health and Retirement Study found that self-report of poor hearing or vision was linked to an earlier (~1year) than expected retirement age [19]. However, studies on the combined effect of vision and hearing loss on the decision to retire are currently lacking. In this cohort of adults aged 55 years and over, we aimed to: 1) Establish whether objectively measured hearing loss, self-perceived hearing handicap and/ or hearing aid use is independently associated with the prevalence and 10-year incidence of retirement from employment; and 2) Determine if dual sensory impairment (DSI) at baseline is independently associated with retirement 10 years later.
5
Page 5 of 22
2. Methods 2.1 Study sample The Blue Mountains Eye Study (BMES) is a population-based cohort study of common eye diseases and other health outcomes in a suburban Australian population located west of
ip t
Sydney. Study methods and procedures have been described elsewhere [20]. Baseline examinations of 3654 residents aged >49 years were conducted during 1992-4 (BMES-1,
cr
82.4% participation rate). Of the baseline participants, 2335 (75.1% of survivors) returned for
us
5-year follow-up examinations during 1997-9 (BMES-2), and 1952 participants (53.4% of the original cohort, or 76.6% of survivors) returned for 10-year follow-up examinations during
an
2002-4 (BMES-3). Hearing was measured at BMES-2 and BMES-3 only; i.e. in BMES-2, 2956 participants had audiometric testing performed. In this report, baseline is referring to
M
BMES-2. The study was approved by the Human Research Ethics Committee of the University of Sydney and was conducted adhering to the tenets of the Declaration of
Ac ce pt e
d
Helsinki. Signed informed consent was obtained from all participants at each examination.
2.2 Audiological assessment
Pure-tone audiometry was performed by audiologists in sound-treated booths, using TDH-39 earphones and Madsen OB822 audiometers (Madsen Electronics, Denmark). Bilateral hearing impairment was determined as the pure-tone average of audiometric hearing thresholds at 500,1000, 2000, and 4000 Hz (PTA0.5-4kHz) in the better ear, defining any hearing loss as PTA0.5-4kHz > 25 dB HL; mild hearing loss as PTA0.5-4kHz > 25-40 dB HL; and moderate to severe hearing loss as PTA0.5-4kHz >40 dB HL. An audiologist asked questions about history of any self-perceived hearing problem, and if a hearing aid had been provided. Specifically, participants were asked: ‘Do you or have you ever worn a hearing aid?’ (Yes/No/Don’t know). The Hearing Handicap Inventory for 6
Page 6 of 22
the Elderly - Shortened version (HHIE-S), developed by Ventry and Weinstein [21], was also administered. The HHIE-S includes 10 questions and a response of ‘yes’ is given 4 points; ‘sometimes’ is given 2 points, and ‘no’ is given 0 points. Scores ≥8 were taken to indicate the
ip t
presence of a hearing handicap [22].
2.3 Assessment of visual impairment
cr
Monocular distance logMAR (logarithm of the minimum angle of resolution) visual
us
acuity was measured with forced-choice procedures using a retroilluminated chart with automatic calibration to 85 cd/m2 (Vectorvision CSV-100TM; Vectorvision Inc, Dayton,
an
Ohio) according to the Early Treatment Diabetic Retinopathy Study protocol [20]. This was conducted with habitual correction (presenting visual acuity, with current eyeglasses, if worn)
M
and after subjective refraction (best-corrected visual acuity). For each eye, visual acuity was recorded as the number of letters read correctly from 0 to 70. For the present study, any
d
visual impairment was defined as presenting or best-corrected visual acuity of the better eye
Ac ce pt e
less than 39 letters (<20/40). DSI was defined as concurrent visual (either presenting or bestcorrected) and hearing impairment, as determined using the above definitions.
2.4 Assessment of retirement status
Employment status was determined from the question: ‘Are you retired or still employed?’ For incidence analysis, we included those participants who reported being employed at baseline, and if any of these participants retired at the follow-up examinations then they were considered as an incident (new) case of retirement. If they reported being ‘retired’ then they were asked ‘If retired, how old were you when you retired?’ They were asked to provide their retirement age in years. Participants were also asked whether they receive a pension and if they answered ‘yes’ they were asked what type of pension did they 7
Page 7 of 22
receive with the following options: age pension, veteran pension, invalid pension, or blind pension.
2.5 Assessment of covariates
ip t
A face-to-face interview with trained interviewers was conducted, and comprehensive data including information about medical history, demographic factors, socio-economic
cr
characteristics, and lifestyle behaviours were obtained from all participants. The medical
us
history included cardiovascular or other systemic disease, and medications used. A past history of angina, diabetes, myocardial infarction, and stroke was determined by responses to
an
a question: “Has a doctor advised you that you have any of the following conditions?” Selfrated health was assessed by asking: “For somebody your age, would you say your health is
2.6 Statistical analysis
d
M
excellent, very good, good, fair, or poor?” Low self-rated health was defined as fair or poor.
Ac ce pt e
SAS statistical software performed all analysis (v9.2, SAS Institute, NC), including ttests, χ2-tests and linear regression. The cross-sectional association between hearing related measures and DSI with retirement status was examined using logistic regression models to estimate odds ratios (OR) and 95% confidence intervals (CI). Multiple regression models were first adjusted for age (entered as a continuous variable) and sex, and then further adjusted for confounders that were found to be significantly associated with retirement status i.e. self-rated health, presence of ≥3 chronic conditions (e.g. hypertension, diabetes, angina, and stroke). Analysis of covariance was used to calculate differences in mean retirement age between those with and without sensory impairment(s), after adjusting for the aforementioned covariates. For longitudinal analysis of baseline sensory impairment and 10-year incidence of retirement we used Cox discrete-time modelling. 8
Page 8 of 22
3. Results 3.1 Hearing-related measures and retirement status Of the 2956 participants who had hearing data at baseline (i.e. at BMES-2), 547 did not
ip t
have information on visual acuity and/or retirement status and thus, were excluded, leaving 2409 participants for cross-sectional analysis. Table 1 shows the study characteristics of
cr
participants included in cross-sectional analysis. At baseline, 650 (27.0%) were employed
us
and 1759 (73.0%) were retired. The majority were receiving the age pension. Just under a third had any hearing impairment while over 1 in 5 participants had presenting visual
an
impairment (Table 1). Cross-sectional data showed that among those with any hearing impairment, 697 (89.2%) were retired compared to 1062 (65.2%) with no hearing loss who
M
were retired (p <0.0001). The highest retirement frequency was observed among participants with moderate to severe hearing loss, 93.6% (n=221). In multiple logistic regression analysis
d
(Table 2), participants with moderate to severe hearing loss versus those with normal hearing
Ac ce pt e
had 96% higher likelihood of being retired. Non-significant associations were observed between self-perceived hearing handicap and hearing aid use and the likelihood of retirement (Table 2).
Analysis of baseline data showed that those with moderate to severe hearing loss had a significantly lower adjusted mean retirement age of 57.2 years compared to a mean retirement age of 58.6 years among those with no hearing loss (multivariable-adjusted pvalue = 0.009). We found no significant differences in adjusted mean retirement ages between those with and without self-perceived hearing handicap (data not shown). However, persons who used a hearing aid versus those who did not use an aid had a significantly lower adjusted mean retirement age, 57.3 versus 58.7 years, multivariable-adjusted p=0.004.
9
Page 9 of 22
Longitudinal or incidence analysis (using Cox discrete-time modelling) included all participants who reported being employed at baseline and who also had complete sensory information i.e. 650. Of the 650 who were employed at baseline, 455 were followed-up 10 years later. Participants who were followed up versus those lost-to follow up, had similar
ip t
proportions of hearing loss (17.3% versus 18.2%, p=0.90), presenting visual impairment (11.1% versus 18.2%, p=0.23) and best corrected visual impairment (2.9% versus 3.0%,
cr
p=0.98) at baseline. Of those who were followed up, 190 participants (41.8%) retired during
us
the 10-year follow-up period. After adjusting for age and sex, participants with any hearing loss had significantly higher odds of being retired by the 10-year follow-up, OR 1.82 (95%
an
CI 1.00-3.30), but this became marginally non-significant after adjusting for all other covariates, multivariable-adjusted OR 1.74 (95% CI 0.95-3.17). No other associations were
Ac ce pt e
3.2 DSI and retirement status
d
retirement.
M
observed between severity of hearing loss, hearing handicap or hearing aid use and incident
Participants with both presenting visual impairment and any level of hearing loss had the highest retirement rate i.e. 94.5% (n=257) compared to those with no presenting visual impairment and normal hearing function – 62.5% (n=830). Participants with no presenting visual impairment but who had moderate to severe hearing loss had greater odds of reporting retirement compared to those with no visual and hearing impairment (Table 3), multivariableadjusted OR 2.52 (95% CI 1.13-5.64). Similarly, participants with no best-corrected visual impairment but with moderate to severe hearing loss had increased likelihood of reporting retirement compared to those with no sensory loss (Table 4), multivariable-adjusted OR 2.15 (95% CI 1.10-4.21). No other significant associations were observed between best-corrected visual impairment and/or hearing loss and retirement (Table 4). 10
Page 10 of 22
Analysis of covariance (ANCOVA) showed that participants who had no presenting visual impairment but who had moderate to severe hearing loss had significantly earlier mean retirement age compared to those with no sensory loss, 57.1 versus 58.7 years (multivariableadjusted p value=0.02). Similarly, those with concurrent presenting visual impairment and
ip t
moderate to severe hearing loss versus participants with no hearing or vision loss had significantly earlier mean retirement age, 57.1 versus 58.7 years (multivariable-adjusted p
cr
value=0.04). No significant associations were observed between best-corrected visual
us
impairment and/or hearing loss and adjusted mean retirement age (data not shown). Additionally, no significant associations were observed between DSI and 10-year incident
M
4. Discussion
an
retirement (data not shown).
The present study provides new insight into the societal impacts of sensory
d
impairments among older adults. We show that older adults with moderate to severe hearing
Ac ce pt e
impairment compared to those with normal hearing had a greater likelihood of being retired at baseline examination. Further, those who used a hearing aid reported a significantly lower adjusted mean retirement age (~1-year). Older adults who had concurrent presenting visual impairment and moderate to severe hearing loss compared to those with no impairment had lower adjusted mean retirement age. Hearing loss at baseline showed a marginally nonsignificant association with the decision to retire by the 10-year follow-up. Study participants with moderate to severe hearing loss (>40 dB HL) versus those with normal hearing not only had a higher likelihood of reporting retirement, but also retired around one to two years earlier. These data concur with other studies which have shown that the degree of acquired hearing impairment is associated with earlier retirement from the labour force [18,23]. Exchange of information which constitutes an important aspect of the 11
Page 11 of 22
work environment, can be seriously impaired in individuals with severe hearing impairment [24], thus, influencing a person’s decision to retire early. Further, we show that hearing aid use was associated with earlier retirement age. Hearing aid use is a marker of more severely impaired hearing, hearing handicap (or high communication needs), or potentially a marker
ip t
of frailty and aging [25], thus, explaining its association with earlier retirement. Our findings are not unexpected given that hearing is an important component of the
cr
face-to-face communicative and interactive processes that occur in the workplace setting
us
[17]. Despite this, there is evidence that workers with severe hearing impairment, defined as having difficulty hearing normal conversations even while wearing a hearing aid, were
an
approximately half as likely to report receiving employer-initiated accommodations (e.g. telephone amplifiers, FM or infrared sound systems) compared to workers with other
M
disabilities [17,26]. One potential reason could be that ‘denial’ or non-acceptance of hearing loss, and the stigma associated with impaired hearing [27], could lead to reluctance to
Ac ce pt e
relevant accommodations [17].
d
disclose a hearing loss and without this disclosure, the employer may not be able to provide
In our study, we found that older adults who had moderate to severe hearing loss and no visual impairment (either best-corrected or presenting) had approximately 2-fold higher odds of reporting retirement at baseline. In regards to retirement age, those with moderate to severe hearing loss reported an earlier retirement age regardless of whether or not they had accompanying presenting visual impairment. Therefore, these findings suggest that the crosssectional association with early retirement appears to be due to the presence of hearing loss rather than vision impairment. A potential explanation for this could be that those who had vision impairment may have already retired (i.e. before BMES-2) or had corrected their impairment (e.g. cataract surgery) so this was not a significant factor influencing workplace participation in our cohort, whereas, hearing loss may be more gradual in nature in most 12
Page 12 of 22
individuals, and is typically diagnosed and managed several years after its onset, often after having led to multiple negative consequences including effects on employment. We observed that participants with any hearing loss at baseline had 82% greater likelihood of retiring by the 10-year follow-up after adjusting for age and sex, however, after
ip t
adjusting for other factors such as visual impairment and self-rated health this association was attenuated. We had very small numbers of people who retired over the 10 years; hence, it
cr
is likely that we had inadequate power to detect a stronger association between hearing loss
us
or DSI and 10-year retirement rates in multivariable analysis. Therefore, other larger cohort studies with adequate follow-up are warranted to assess the longitudinal association between
an
objectively measured sensory loss and retirement rates.
Sensory difficulties are common experiences within older age groups; as a result they
M
are often overlooked or dismissed [13]. However, hearing loss interferes with so many of life’s activities, that it may prove to be a major impediment to society’s need to have people
d
remain longer in the workforce as the proportion of ‘working age’ people in developed
Ac ce pt e
countries like Australia shrinks [28]. Moreover, the annual cost of lost earnings due to workplace separation and early retirement from hearing loss was estimated at $6.7 billion in Australia [29]. Together these findings suggest that moderate to severe hearing loss could be a key factor influencing the decision to retire, hence, correction of hearing impairment, including the provision of assistive devices (e.g., FM systems) and rehabilitative services including hearing aid provision such as auditory and communication training [14,30] should be encouraged by clinicians, as it could improve work opportunities for hearing impaired individuals. This study has several strengths, including its prospective design, representative cohort, and use of objective, standardized audiometric and vision testing.. However, it is also important to discuss some of the study limitations. First, there could have been survival bias 13
Page 13 of 22
in our study, that is, persons who did not attend follow-up visits could have been more likely to have had sensory loss, resulting in an under-reporting event rate of retirement and potential underestimation of the associations. Second, while we adjusted for several confounders we did not collect information on contextual factors e.g. employer or organisation characteristics,
ip t
which could have influenced observed associations. Finally, our study analyzed data collected ~10 years ago, and we cannot discount the possibility that our findings might not be
cr
entirely generalizable to the present day, and would require more contemporary cohorts to
us
validate and confirm our observed associations. Nevertheless, there is a dearth of
epidemiological studies that have assessed the association between concurrent vision loss and
an
hearing loss and retirement status, thus, our study findings is a valuable addition to the existing published literature.
M
In summary, we show that older adults who had moderate to severely impaired hearing and/ or who were using a hearing aid (versus those not using a hearing aid) had greater odds
d
of being retired at baseline. Presence of coexisting visual and hearing impairment at baseline
Ac ce pt e
did not independently influence the decision to retire over the 10 years. Therefore, as the workforce in many developed countries ages the impact of sensory impairment, particularly hearing loss, on workplace separation and early retirement warrants continued attention.
Author contributions:
Conceived and designed the experiments: BG and PM; Performed the experiments: BG and PM; Analyzed and interpreted the data: BG, GL, CM, GB, and PM; Contributed reagents, materials, analysis tools or data: PM and GB; and Wrote the paper: BG. Funding: The Blue Mountains Eye and Hearing Study were supported by the Australian National Health and Medical Research Council (Grant Nos. 974159, 991407, 211069, 262120). The authors also acknowledge the financial support of the NHMRC Partnerships for 14
Page 14 of 22
Better Health (grant no. 571449) and the HEARing CRC, established and supported under the Australian Government's Cooperative Research Centres Program Conflict of interest: All authors have no competing interests to declare.
ip t
References [1] Taylor HR, Keeffe JE, Vu HT, Wang JJ, Rochtchina E, Pezzullo ML et al. Vision loss
cr
in Australia. Med J Aust 2005;182:565-68.
us
[2] Gopinath B, Rochtchina E, Wang JJ, Schneider J, Leeder SR, Mitchell P. Prevalence of age-related hearing loss in older adults: Blue Mountains Study. Arch Intern Med
an
2009;169:415-16.
[3] Schneider JM, Gopinath B, McMahon CM, Leeder SR, Mitchell P, Wang JJ. Dual
M
Sensory Impairment in Older Age. J Aging Health 2011.
[4] Chia EM, Mitchell P, Ojaimi E, Rochtchina E, Wang JJ. Assessment of vision-related
d
quality of life in an older population subsample: the blue mountains eye study.
Ac ce pt e
Ophthalmic Epidemiol 2006;13:371-77.
[5] Tay T, Wang JJ, Kifley A, Lindley R, Newall P, Mitchell P. Sensory and cognitive association in older persons: findings from an older Australian population. Gerontology 2006;52:386-94.
[6] Cugati S, Cumming RG, Smith W, Burlutsky G, Mitchell P, Wang JJ. Visual impairment, age-related macular degeneration, cataract, and long-term mortality: the Blue Mountains Eye Study. Arch Ophthalmol 2007;125:917-24.
[7] Gopinath B, Wang JJ, Schneider J, Burlutsky G, Snowdon J, McMahon CM et al. Depressive symptoms in older adults with hearing impairments: the Blue Mountains Study. J Am Geriatr Soc 2009;57:1306-8.
15
Page 15 of 22
[8] Gopinath B, Schneider J, Hickson L, McMahon CM, Burlutsky G, Leeder SR et al. Hearing handicap, rather than measured hearing impairment, predicts poorer quality of life over 10 years in older adults. Maturitas 2012;72:146-51. [9] Schneider J, Gopinath B, Karpa MJ, McMahon CM, Rochtchina E, Leeder SR et al.
ip t
Hearing loss impacts on the use of community and informal supports. Age Ageing 2010;39:458-64.
cr
[10] Schneider JM, Gopinath B, McMahon CM, Leeder SR, Mitchell P, Wang JJ. Dual
us
sensory impairment in older age. J Aging Health 2011;23:1309-24.
[11] Gopinath B, Schneider J, McMahon CM, Teber E, Leeder SR, Mitchell P. Severity of
an
age-related hearing loss is associated with impaired activities of daily living. Age Ageing 2012;41:195-200.
M
[12] Gopinath B, Schneider J, McMahon CM, Burlutsky G, Leeder SR, Mitchell P. Dual sensory impairment in older adults increases the risk of mortality: a population-based
d
study. PLoS ONE 2013;8:e55054.
Ac ce pt e
[13] Crews JE, Campbell VA. Vision impairment and hearing loss among communitydwelling older Americans: implications for health and functioning. Am J Public Health 2004;94:823-29.
[14] Lee DJ, Gomez-Marin O, Lam BL, Zheng DD, Arheart KL, Christ SL et al. Severity of concurrent visual and hearing impairment and mortality: the 1986-1994 National
Health Interview Survey. J Aging Health 2007;19:382-96.
[15] Lee DJ, Lam BL, Gomez-Marin O, Zheng DD, Caban AJ. Concurrent hearing and visual impairment and morbidity in community-residing adults: the National Health Interview Survey, 1986 to 1996. J Aging Health 2005;17:531-46.
16
Page 16 of 22
[16] Viljanen A, Tormakangas T, Vestergaard S, Andersen-Ranberg K. Dual sensory loss and social participation in older Europeans. European Journal of Ageing 2014;11:15567. [17] Fischer ME, Cruickshanks KJ, Pinto A, Klein BE, Klein R, Dalton DS. Hearing
ip t
impairment and retirement. J Am Acad Audiol 2014;25:164-70.
HUNT study. Eur J Public Health 2013;23:617-22.
cr
[18] Helvik AS, Krokstad S, Tambs K. Hearing loss and risk of early retirement. The
us
[19] Dwyer DS, Mitchell OS. Health problems as determinants of retirement: are self-rated measures endogenous? J Health Econ 1999;18:173-93.
an
[20] Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:357-64.
M
[21] Ventry IM, Weinstein BE. Identification of elderly people with hearing problems. ASHA 1983;??:37-42.
d
[22] Cruickshanks KJ, Wiley TL, Tweed TS, Klein BE, Klein R, Mares Perlman JA et al.
Ac ce pt e
Prevalence of hearing loss in older adults in Beaver Dam, Wisconsin. The
Epidemiology of Hearing Loss Study. Am J Epidemiol 1998;148:879-86.
[23] Kramer SE. Hearing impairment, work, and vocational enablement. Int J Audiol 2008;47 Suppl 2:S124-S130.
[24] Dalton DS, Cruickshanks KJ, Klein BE, Klein R, Wiley TL, Nondahl DM. The impact of hearing loss on quality of life in older adults. Gerontologist 2003;43:661-
68.
[25] Gopinath B, McMahon CM, Burlutsky G, Mitchell P. Hearing and vision impairment and the 5-year incidence of falls in older adults. Age Ageing 2016;45:409-14.
17
Page 17 of 22
[26] Zwerling C, Whitten PS, Sprince NL, Davis CS, Wallace RB, Blanck P et al. Workplace accommodations for people with disabilities: National Health Interview Survey Disability Supplement, 1994-1995. J Occup Environ Med 2003;45:517-25. [27] Wallhagen MI. The stigma of hearing loss. Gerontologist 2010;50:66-75.
ip t
[28] Dillon H. The 2006 Libby Harricks Memorial Oration: Hearing loss: the silent epidemic. Who, why, impact and what can we do about it. 2006. ACT, Deafness
cr
Forum Limited.
us
[29] Listen Hear! The economic impact and cost of hearing loss in Australia. CRC for Cochlear Implant and Hearing Aid Innovation, Vicdeaf, editors. 2006. Australia,
an
Access Economics Pty Limited.
[30] Heine C, Browning CJ. Communication and psychosocial consequences of sensory
Ac ce pt e
d
2002;24:763-73.
M
loss in older adults: overview and rehabilitation directions. Disabil Rehabil
18
Page 18 of 22
Table 1. Baseline characteristics of study participants and non-participants
67.2 (8.9)
Male sex
1160 (48.2)
Low self-rated health
444 (18.5)
Presence of ≥3 comorbidities
260 (11.0)
Employed
650 (27.0)
Retired
1759 (73.0)
cr
Age, yrs
ip t
Participants (n=2409)
us
Characteristics
Type of pension
Invalid
67 (5.0)
Veteran
223 (16.6)
781 (32.4)
Ac ce pt e
Any hearing loss (>25 dB HL)
1 (0.1)
d
Blind
an
940 (69.9)
M
Age
Best corrected visual impairment (<20/40) Presenting visual impairment (<20/40)
250 (10.4) 569 (23.7)
Data are presented as mean (SD) or n (%).
19
Page 19 of 22
Table 2. Cross-sectional association between measured hearing loss, hearing handicap inventory for elderly (HHIE) scores, and hearing aid use and retirement status (n=2409) Prevalence of retirement, OR (95% CI) Age-sex adjusted
Multivariate-adjusted a
No hearing loss (≤25 dB HL), n=1628
1.0 (reference)
1.0 (reference)
Any hearing loss (>25 dB HL), n=781
1.23 (0.90-1.68)
cr
Severity of hearing loss b No hearing loss (≤25 dB HL), n=1628
Moderate to severe hearing loss (>40 dB HL),
Ac ce pt e
Yes, n=253
d
No, n=2150
M
No handicap (HHIE <8), n=1476
Hearing aid use
1.06 (0.75-1.50)
1.04 (0.73-1.47)
an
Severity of hearing handicap
Severe handicap (HHIE ≥26), n=142
1.0 (reference)
2.00 (1.08-3.71)
n=221
1.20 (0.87-1.65)
1.0 (reference)
us
Mild hearing loss (26-40 dB HL), n=545
Moderate handicap (HHIE 8-24), n=723
ip t
Presence of hearing loss b
1.96 (1.05-3.66)
1.0 (reference)
1.0 (reference)
1.17 (0.90-1.53)
1.13 (0.86-1.49)
1.60 (0.91-2.81)
1.52 (0.86-2.67)
1.0 (reference)
1.0 (reference)
1.57 (0.89-2.77)
1.53 (0.86-2.73)
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, best-corrected visual impairment and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
20
Page 20 of 22
cr us
an
Table 3. Cross-sectional association between objectively measured dual sensory impairment and retirement status among Blue Mountains Eye
Prevalence of retirement, OR (95% CI)
Hearing impairment (dB HL)
Age-sex adjusted
Multivariate-adjusted a
No visual impairment
≤25 (No hearing loss), n=1328
1.0 (reference)
1.0 (reference)
>25 (Any hearing loss), n=506
1.21 (0.86-1.71)
1.18 (0.83-1.68)
1.02 (0.71-1.48)
1.00 (0.68-1.45)
2.50 (1.13-5.51)
2.52 (1.13-5.64)
≤25 (No hearing loss), n=297
1.13 (0.78-1.62)
1.09 (0.75-1.58)
>25 (Any hearing loss), n=272
1.43 (0.76-2.67)
1.39 (0.74-2.63)
>25-≤40 (Mild hearing loss), n=173
1.44 (0.65-3.20)
1.43 (0.63-3.22)
1.41 (0.54-3.68)
1.34 (0.52-3.50)
ed
Visual impairment (<20/40)
M
Study participants (n=2403)
>25-≤40 (Mild hearing loss), n=371
pt
>40 (Moderate-severe hearing
Ac
ce
Presenting
loss), n=135
>40 (Moderate-severe hearing loss), n=99
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
21
Page 21 of 22
cr us
an
Table 4. Cross-sectional association between measured dual sensory impairment, and retirement status among Blue Mountains Eye Study
Prevalence of retirement, OR (95% CI)
Hearing impairment (dB HL)
Age-sex adjusted
Multivariate-adjusted a
No visual impairment
≤25 (No hearing loss), n=1524
1.0 (reference)
1.0 (reference)
>25 (Any hearing loss), n=635
1.23 (0.89-1.70)
1.21 (0.87-1.67)
1.05 (0.73-1.49)
1.03 (0.72-1.46)
2.17 (1.12-4.21)
2.15 (1.10-4.21)
≤25 (No hearing loss), n=104
1.31 (0.68-2.51)
1.30 (0.67-2.53)
>25 (Any hearing loss), n=146
1.48 (0.53-4.18)
1.42 (0.50-4.06)
>25-≤40 (Mild hearing loss), n=93
1.66 (0.42-6.49)
1.60 (0.40-6.38)
1.26 (0.26-5.99)
1.19 (0.25-5.80)
ed
Visual impairment (<20/40)
M
participants (n=2409)
>25-≤40 (Mild hearing loss), n=452
pt
>40 (Moderate-severe hearing
Ac
ce
Best corrected
loss), n=183
>40 (Moderate-severe hearing loss), n=53
OR – odds ratio; CI – confidence interval. a Further adjusting for self-rated general health, and presence of 3 or more comorbidities. b Separate logistic regression models were run when examining the association between presence of hearing loss (any versus no hearing loss) and severity of hearing loss (mild or moderate to severe hearing loss versus no hearing loss).
22
Page 22 of 22