Clinical Section

Gerontology 1999;45:323–328 Received: June 24, 1998

Accepted: March 10, 1999

Quality of Life Determinants and Hearing

Function in an Elderly Population:
Osservatorio Geriatrico Campano Study Group
Francesco Cacciatore a Claudio Napoli b, e Pasquale Abete a Elio Marciano c
Maria Triassi d Franco Rengo a, f
a Divisionof Geriatrics, Department of Internal Medicine and Cardiovascular Sciences, b Department of
Clinical and Experimental Medicine, c Division of Audiology, d Department of Public Health Sciences, ‘Federico II’
University of Naples, School of Medicine, Naples, Italy; e Department of Medicine, UCSD, La Jolla, Calif., USA;
f IRCCS, Clinical Foundation S. Maugeri, Campoli/Telese (Benevento), Italy

Key Words (MMSE)) was 27.9%, mean depressive symptomatology

Hearing impairment W Aging W Cognitive function W score (evaluated by Geriatric Depression Scale (GDS))
Depression W Disability W Hearing aids was 11.4 B 6.6, while disability assessed by Activity of
Daily Living (ADL) was present in 7.0% of the whole pop-
ulation. A strong relationship was found between both
Abstract decreasing hearing function and MMSE decline, inde-
Background: Hearing impairment (HI) is a very common pendently by the effect of age and education (r = 0.97; p !
condition in elderly people and the epidemiology togeth- 0.01). A positive relationship (r = 0.85; p ! 0.01) between
er with hearing-related problems is still poorly investi- GDS score and hearing function was also found. More-
gated. Moreover, the cognitive status may be impaired in over, at an increased level of hearing loss, a lower ADL
relation to hearing function. Objective: The goal of the score was recorded (r = 0.98; p ! 0.01). Finally, the use of
study was to evaluate: (a) the prevalence of HI in a ran- hearing aids reduced GDS score. In logistic regression
dom sample of elderly people aged 65 and over (n = analysis, gender, age and educational level indicate that
1,750) living in Campania, a region of southern Italy; hearing loss risk increased with age (odds ratio 1.60; 95%
(b) the cross-sectional relationship between hearing confidence interval 1.53–1.71), whereas education plays
function and cognitive status and also depressive symp- a protective role (odds ratio 0.75; 95% confidence inter-
tomatology and disability, and (c) to assess the role of val 0.72–0.80). Conclusion: HI is very prevalent among
hearing aids on depressive symptomatology. Methods: elderly people and is associated with either cognitive
Cross-sectional study on a random sample of elderly impairment and/or depression and reduction of function-
population. Results: The overall participation rate in al status. This study suggests that hearing aids may pro-
the study was 74.8% (n = 1,332, mean age was 74.2 tect against cognitive impairment and disability, improv-
B 6.4 years). The prevalence rate of HI (evaluated by ing quality of life of aged people.
questionnaire) was 27.2%, cognitive impairment preva- Copyright © 1999 S. Karger AG, Basel

lence (evaluated by the Mini-Mental State Examination

© 1999 S. Karger AG, Basel Francesco Cacciatore, MD, PhD

ABC
198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library

0304–324X/99/0456–0323$17.50/0 Via Merliani, 20

Fax + 41 61 306 12 34 I–80127 Napoli (Italy)
E-Mail karger@karger.ch Accessible online at: Tel. +39 081 7462270, Fax +39 081 7462339
www.karger.com www.karger.com/journals/ger
Downloaded by:
 Introduction planned target of 100 stations. Because of repeated sampling, the two
largest Municipalities had more than two polling-stations (Salerno
had 4 and Naples had 20). Forty Municipalities from a total of 550
Hearing impairment (HI) represents a significant clini-
were involved in the survey. Therefore, just for sampled polling-
cal problem among elderly people. Several studies esti- stations, rolls of voters aged more than 65 were provided by Munici-
mating hearing function in aged people (165 years) have palities. Finally, subjects were chosen randomly within each polling-
shown a prevalence rate ranging from 21 to 72% [1–3]. station, allocation being stratified by sex and age, in order to preserve
Moreover, HI shows a negative correlation with commu- the same age and sex distribution of the target population. Because of
the sampling design, the final planned sample was self-weighting,
nication, social integration and with well-being [4]. The
that is, the observed percentages were a direct expression of popula-
relation between cognitive status and depressive symp- tion percentages.
tomatology in elderly people with HI is not well estab- The study sample consisted of 1,780 subjects, 756 (42.5%) males
lished. Although several studies suggest a relationship and 1,024 (57.5%) females, with a sampling fraction of 0.3%. Of this
between hearing and cognitive impairments and/or de- total, 448 (25.2%) refused to participate in the study resulting in a
final study sample of 1,332 subjects; thus, the overall participation
pression, this has been assessed by various methodologi-
rate in the study was 74.8%. There were no differences in age and sex
cal approaches assessing the cognitive status (various distribution between responders and nonresponders (data not
types of tests with different scores) and hearing function shown). There was a higher percentage of nonresponders in urban
(speech audiometry test, pure tone audiometry, acoustic areas, for example, the overall refused rate in Naples was 30.9%. The
impedance test, etc.) [5–8]. It should be pointed out that subjects were contacted at home or in their institution and examined
by physicians who had been trained to administrate a questionnaire
cognitive impairment is usually evaluated by Mini-Men-
that included cognitive and depression tests. Forty-seven subjects
tal State Examination (MMSE) [9, 10]. Thus, the results (3.5% of the final study sample, n = 1,332) did not complete the ques-
of above studies are not directly comparable. tionnaire, since they refused both the MMSE and Geriatric Depres-
Another issue is that the use of hearing aids plays a sion Scale (GDS).
positive effect on self-perceived hearing handicap in aged A data collection quality control was carried out by the coordinat-
ing center with telephone calls to the subjects. Visits were carried out
people with mild to moderate HI, but no effects are
between October 1991 and April 1992. No age and sex differences
observed on cognitive function and other matters such as were observed. The interview was delayed in 42 cases (3.2%) due to
social activities, satisfactory social relations and well- intercurrent medical problems.
being [4].
The goal of the present study was to investigate the Demographic Variables
Sex, age and marital status were recorded. Education was scored
prevalence of HI and the relation between hearing func-
according to a seven-level scale: illiterate = 1; only able to sign = 2;
tion and cognitive status in a random sample of elderly able to read and write but without formal schooling = 3; 1–5 years of
people resident in Campania, a region of southern Italy. education = 4; 6–8 years of education = 5; 9–13 years of education =
In order to fit each patient’s characteristic, we modulated 6, and postsecondary school diploma = 7.
the voice tone (126 dB in the presence of HI) by which the
Hearing Function
MMSE was administered. We focused our attention also
The function was assessed by considering the ability to hear oth-
on the quality of life assessed by several conditions such as ers talking in a normal voice (! 26 dB) with and without hearing aids,
depressive symptomatology and disability. as described in a previous epidemiological survey [5]. The question
on hearing ability was modified according to the participant’s use of
hearing aids. A score of 1 was given to those who answered positively:
‘Without a hearing aid can you usually hear and understand what a
Methods person says without seeing his face if that person talks to you in a
normal voice in a quiet room?’ For those who reported using a hear-
The ‘Osservatorio Geriatrico Regione Campania’ was a cross-sec- ing aid occasionally, frequently, or practically always, this question
tional study performed in 1992 in Campania, a region of southern was phrased: ‘With a hearing aid can you usually hear and under-
Italy, as previously described [11–13]. To guarantee homogeneous stand what a person says without seeing his face if that person talks to
sampling of the whole territory, and to reduce the costs of the survey, you in a normal voice in a quiet room?’ A score of 2 was assigned to
a stratified multistage sampling design was used. The referendum subjects who were not able to hear and understand, with or without a
electoral roll head as of June 9, 1991, was used as the population hearing aid, people speaking in a normal voice (! 26 dB), while a
source. Municipalities (the lowest level of local government) were the score of 3 was given if they could hear and understand what a person
primary sampling units. They were ordered within Districts (the says without seeing his face if that person used a raised voice in a
intermediate level) according to size, defined as the number of poll- quiet room. A score of 4 was assigned to deaf subjects. We reorgan-
ing-stations. Lists of the Municipalities with the number of poll- ized the answers in a four-level scale (1–4) in which the lowest score
ing-stations were supplied by district authorities. Municipalities indicated no hearing problems and the highest those individuals who
were chosen systematically with repetition, a couple of polling-sta- were deaf. When present, ear-wax was cleared by an otologist before
tions being randomly selected within Municipalities, according to a hearing evaluation.

324 Gerontology 1999;45:323–328 Cacciatore/Napoli/Abete/Marciano/Triassi/

Rengo
198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library
Downloaded by:
 Measure of Cognitive Impairment
The Italian version of the MMSE [9, 10] validated by Measso et
al. [14] was used as a measure of cognitive mental status. Cognitive %
impairment is defined as a score of ! 24 on the MMSE. 90 Score 1
Score 2
Measure of Depression Symptomatology 80 Score 3
Subjects were screened for depression using the GDS [15]. The Score 4
GDS is a questionnaire consisting of 30 true or false items. 70

Disability 60

Disability was measured considering an Activities of Daily Liv-

50
ing (ADLs) index [16], which assessed the following activities (bath-
ing or showering, dressing, eating, getting in and out of bed or rising 40
from a chair and using the toilet). Participants were asked if they had
difficulty performing each of these activities. A disability score was 30
assigned to each response (0 = ‘uses no help to perform the activity’,
1 = ‘uses a device to perform the activity’, 2 = ‘uses assistance of 20
another person to perform the activity’, 3 = ‘does not perform the
activity’). Subjects who cannot perform the function without help 10
were considered disabled.
0
65–69 70–74 75–79 80–84 ³85
Statistical Analyses
Age (years)
The data were analyzed by SPSS 7.0 statistics package [17]. All
results are mean B SD. Differences in age, education, GDS score,
MMSE score, number of ADL subjects unable to perform, number of
diseases and number of drugs among subjects with a hearing impair-
ment (score 1 1) and without hearing impairment (score = 1) were Fig. 1. Prevalence of HI age stratified.
evaluated by one-way ANOVA. To evaluate the role played on hear-
ing impairment by sex, age and education, logistic regression analysis
was performed considering hearing impairment as the dependent
variable. Subjects with a score 1 1 were considered impaired. Results
with respect to age were reorganized into 5-year age cohorts.
Table 1. Respondent characteristics of Osservatorio Geriatrico
Campano, Italy (n = 1,332)
Results
Variables MeanBSD Range %

The final sample of the study consisted of 1,332 sub- Age 74.2B6.4 65–96
jects because in 7 cases information on hearing function Male 42.3 (560)
was not recorded. Mean age was 74.2 B 6.4 years (range Females 57.7 (772)
65–96), and the mean educational level was 3.4 B 1.5. Educational level 3.4B1.5 1–7
Hearing function
27.9% of subjects had a MMSE score !24; GDS had a Score 1 72.8 (970)
mean score of 11.4 B 6.6 (table 1). Moreover, 970 sub- Score 2 21.0 (280)
jects (72.8%) had no self-reported problems with hearing Score 3 3.5 (46)
function (score 1), 280 subjects (21.0%) belong to score 2, Score 4 2.7 (36)
46 subjects (3.5%) to score 3, and 36 subjects (2.7%) to Cognitive function
MMSE ! 24 27.9
score 4 (table 1). Sixty-five subjects (4.9%) were already MMSE 624 72.1
hearing aid users. The prevalence of subjects with a diffi- GDS score 11.4B6.6 0–29
cutly in understanding what a person says without seeing Disability ADL 0–5
his face if that person talks in a normal voice increases No function lost 93.0 (1,239)
progressively with age involving 42.7% of the subjects One or more functions lost 7.0 (93)
aged x85 years (fig. 1). When MMSE values were ad- MMSE = Mini-Mental State Examination; GDS = Geriatric
justed for age and education by multiple regression analy- Depression Scale; ADL = Activity of Daily Living.
sis, MMSE and hearing function were inversely correlated
(score 1 – MMSE = 27.1 B 1.6; score 2 – MMSE = 25.9 B
4.2; score 3 – MMSE = 25.2 B 4.7) (r = 0.97; p ! 0.01)

Hearing Impairment in the Elderly Gerontology 1999;45:323–328 325

198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library
Downloaded by:
 Table 2. Distribution of variables according to hearing function
dichotomized considering subjects with no hearing problems versus
30 y = 27.9 – 0.9x, r = 0.97, p < 0.01 the others

28 Variables Score 1 Score 2–4 p value

(meanBSD) (meanBSD)
26
MMSE

Age 73.1B5.7 77.2B7.0 0.001

24 Educational level 3.5B1.4 2.9B1.4 0.001
GDS score 10.6B6.5 13.6B6.4 0.001
MMSE score 26.0B4.0 24.0B5.7 0.001
22
ADL score 0.1B0.6 0.5B1.3 0.001
20 a
Score 1 Score 2 Score 3
MMSE = Mini-Mental State Examination; GDS = Geriatric
Depression Scale; ADL = Activity of Daily Living.

30
y = 9.4 + 1.7x, r = 0.85, p < 0.01
25

20 Table 3. Logistic regression analysis considering the role of sex, age

and educational-level on HI
GDS

15

10 Variables Odds ratio 95% CI p value

5 Sex (female/male) 0.79 0.70–0.91 0.09

Age (65–96) 1.60 1.53–1.71 0.001
0 b
Educational level 0.75 0.72–0.80 0.001
Score 1 Score 2 Score 3

1.0
y = –3.3 + 0.2x, r = 0.98, p < 0.01
0.8
score 3 – ADL = 0.7 B 1.6) (r = 0.98; p ! 0.01) (fig. 2c).
0.6 Table 2 shows the age, education, depression symptom-
ADL

atology and ADL score differences according to hearing

0.4
function in subjects with score 11 or without reported
0.2
hearing problems (score = 1). Subjects with reported HI
have a greater mean age, a lower MMSE score, a higher
0.0 c GDS score and a lower ADL score. In table 3 are shown
Score 1 Score 2 Score 3 the results from a logistic regression model to evaluate the
Hearing function
role of gender, age and educational level on hearing
impairment (score 11). The probability of having an HI
increased with age (odds ratio for 5 years = 1.60; confi-
Fig. 2. Cognitive status evaluated by MMSE (a), depressive symp- dence interval 1.53–1.71) while decreased with increasing
tomatology evaluated by GDS (b) and disability evaluated by ADL educational level (odds ratio = 0.75; confidence interval
(c) stratified by hearing function.
0.70–0.91). In subjects with hearing problems we observe
that hearing aid users have a low GDS score (9.7 vs. 13.3,
F = 7.26; p = 0.007).

(fig. 2a). A positive relationship between hearing function

and GDS score was also found (score 1 – GDS = 10.6 B Discussion
6.5; score 2 – GDS = 13.7 B 6.5; score 3 – GDS = 14.1 B
6.4) (r = 0.85; p ! 0.01) (fig. 2b). An increased level of We report that HI is a common condition affecting
hearing impairment corresponds to a lower ADL score elderly people with a prevalence rate of 27.2%; an age-
(score 1 – ADL = 0.1 B 0.6; score 2 – ADL = 0.5 B 1.3; related increase is observed with a peak prevalence in

326 Gerontology 1999;45:323–328 Cacciatore/Napoli/Abete/Marciano/Triassi/

Rengo
198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library
Downloaded by:
very old subjects (685 years). MMSE score progressively Many conditions can interfere with hearing impair-
decreased with HI whereas GDS and disabilities in- ment in determining depressive symptomatology. The
creased with severity of HI. Low educational level and maintenance of social support and social leisure activity
aging are independently predictive of hearing impairment can help the aged subject, especially people who live
in multivariate analysis. Finally, the use of hearing aids alone, to preserve functional status. We found the highest
reflects a lower depressive symptomatology score on level of disability in performing activities of daily living in
GDS. hearing-impaired subjects. Therefore, a greater impor-
Epidemiological studies indicate that prevalence of HI tance must be given to the supply of hearing aids for the
in the elderly varies from 21 to 72% [1–3]. The difference elderly considering the role played in depressive symp-
is due to the instruments used to assess the hearing func- tomatology [24]. Our data strengthen the growing body of
tion and to the population studied [18]. Cognitive func- literature confirming the association between hearing
tion and depressive symptomatology are associated with function, cognitive impairment, depressive symptomatol-
HI. Several cross-sectional and longitudinal studies have ogy and disability in performing activities of daily living
demonstrated a predictive role of hearing function on cog- and quality of life in elderly people [25, 26]. In our opin-
nition. With respect to the study design, conflicting evi- ion, hearing function must be preserved if possible in
dence has been given on the relationship between hearing order to prevent cognitive impairment, mood problems
and cognitive impairment. Cutler and Grams [19] dem- and disability.
onstrate a positive correlation between these two condi- In conclusion, HI is very common among elderly peo-
tions in a cross-sectional survey conducted on 14,783 sub- ple and is associated with either cognitive impairment
jects, also after age, sex, education, health, and functional and/or depression and reduction of functional status. This
status adjustment; in this survey both cognitive and hear- study suggests that hearing aids may protect against cogni-
ing function were self-reported. Gennis et al. [20] demon- tive impairment and disability, improving quality of life
strated on 259 elderly healthy subjects enrolled in a longi- of aged people.
tudinal survey a small correlation between hearing acuity
and memory scores at baseline observation, the age and
gender adjustment made no longer significant the statisti- Acknowledgements
cal association. No correlation between hearing function
The study was supported by a grant from Regione Campania to
at entry and cognitive status at 5 years’ follow-up was
the Osservatorio Geriatrico Regione Campania. The authors thank
found as described by Jerger et al. [21]. Hearing aids pro- Drs. Claudio Calabrese and Pietro Ferrara for their skillful collabora-
vide sustained benefit in elderly individuals with hearing tion in the data handling.
problems in respect of cognitive function and depressive
symptomatology. GDS score in subjects with hearing
problems stratified by hearing aid use is statistically dif-
ferent in the two groups, indicating a lower score for hear-
ing aid users. The finding that people with hearing aids
have lower depression symptomatology reflects that the
unalleviated impairment may feed back into the disorder,
accentuating the problems that contribute to depression.
In the present study we found that a limited number of
people with hearing impairment are treated with hearing
aids. The responsibility is probably due both to doctors
and to the elders who tend to underestimate the degree of
their hearing impairment and the importance of hearing
aids [22]. In fact, neurosensorial presbyacusia is consid-
ered a physiological condition especially in very old peo-
ple [23] who do not need to be treated. Moreover, the pro-
tective role of education, that we have found, reinforced
the concept that probably subjects with a higher educa-
tional level have a greater attention to preventive care and
fewer problems in acceptance of acoustic aids.

Hearing Impairment in the Elderly Gerontology 1999;45:323–328 327

198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library
Downloaded by:
 References

1 Moore AA, Siu AL: Screening for common 10 Tombaugh TN, McIntyre NJ: The Mini-Men- 17 SPPS/PC+ User’s Guide. Chicago, SPSS Inc,
problems in ambulatory elderly: Clinical con- tal State Examination: A comprehensive re- 1997.
firmation of a screening instrument. Am J Med view. J Am Geriatr Soc 1992;40:922–935. 18 Stumer J, Hickson L, Worrall L: Hearing im-
1996;100:438–443. 11 Cacciatore F, Abete P, Ferrara N, Paolisso G, pairment, disability and handicap in elderly
2 Feller BA: Prevalence of selected impairments, Amato L, Canonico S, Maggi S, Varricchio M, people living in residential care and in the com-
US, 1977 Vital and Health Statistics Depart- Rengo F: The role of blood pressure in cogni- munity. Disabil Rehabil 1996;18:76–82.
ment of Health and Human Services, National tive impairment in an elderly population. J 19 Cutler SJ, Grams AE: Correlates of self-re-
Center for Health Statistics, DHHS Publ Hypertens 1997;15:135–142. ported everyday memory problems. J Gerontol
No (PHS) 81-1562, Series 10, No 134. Wash- 12 Cacciatore F, Gallo C, Ferrara N, Abete P, 1988;43:582–590.
ington, US Government Printing Office, 1981. Paolisso G, Canonico S, Signoriello G, Ter- 20 Gennis V, Garry PJ, Haaland KY, Yeo RA,
3 Gates GA, Cooper JC, Kannel WB, Miller NJ: racciano C, Napoli C, Varricchio M, Rengo F: Goodwin JS: Hearing and cognition in the
Hearing in the elderly: The Framingham co- Morbidity patterns in aged population in elderly. New findings and a review of the litera-
hort 1983–85. Ear Hear 1990;11:247–256. Southern Italy: A survey sampling. Arch Ge- ture. Arch Intern Med 1991;151:2259–2264.
4 Tesch-Romer C: Psychological effects of hear- rontol Geriatr 1998;26:201–213. 21 Jerger J, Jerger S, Pirozzolo FJ: Correlational
ing aid use in older adults. J Gerontol B Psy- 13 Cacciatore F, Abete P, Ferrara N, Calabrese C, analysis of speech audiometric scores, hearing
chol Sci 1997;52:127–138. Napoli C, Maggi S, Varricchio M, Rengo F: loss, age, and cognitive abilities in the elderly.
5 Berkowitz AO, Hochberg I: Self-assessment of Congestive heart failure and cognitive impair- Ear Hear 1991;12:103–109.
hearing handicap in the aged. Arch Otolaryngol ment in an older population. Osservatorio Ge- 22 Weinstein BE: Age-related hearing loss: How to
1971;93:25–28. riatrico Campano Study Group. J Am Geriatr screen for it, and when to intervene. Geriatrics
6 Ives DG, Bonino P, Traven ND, Kuller LH: Soc 1998;46:1343–1348. 1994;49:40–45.
Characteristics and comorbidities of rural old- 14 Measso G, Cavarzeran F, Zappalà G, Leibo- 23 Quaranta A, Assennato G, Sallustio V: Epide-
er adult with hearing impairment. J Am Ger- witz BD, Crook TR, Pirzzolo FJ: The Mini- miology of hearing problems among adults in
iatr Soc 1995;43:803–806. Mental State Examination: Normative study Italy. Scand Audiol Suppl 1996;42:9–13.
7 Van Rooij JCGM, Plomp R: Auditive and cog- on an Italian random sample. Dev Neuropsy- 24 Quaranta A, Salonna I, Longo G: Subclinical
nitive factors in speech perception by elderly chol 1993;9:77–85. changes of auditory function in the aged. Acta
listeners. Acta Otolaryngol 1990;suppl 476: 15 Brink TL, Yesavage JA, Lum O: Screening tests Otolaryngol Suppl (Stockh) 1990;476:91–95.
177–181. for geriatric depression. Clin Gerontol 1982;1: 25 Apollonio I, Carabellese C, Frattola L, Trabuc-
8 Maggi S, Minicuci N, Martini A, Langlois J, 37–43. chi M: Effects of sensory aids on the quality of
Siviero P, Pavan M, Enzi G: Prevalence rates 16 Katz S, Ford AB, Moskowitz RW, Jackson BA, life and mortality of elderly people: A multivar-
of hearing impairment and comorbid condi- Jaffe MW: Studies of illness in the aged. The iate analysis. Age Ageing 1996;25:89–96.
tions in older people: The Veneto Study. J Am index of ADL; a standardized measure of bio- 26 Verbrugge LM, Patrick DL: Seven chronic con-
Geriatr Soc 1998;46:1069–1074. logical and psychological functions. JAMA ditions: Their impact on US adults’ activity
9 Folstein MF, Folstein SE, McHugh PR: ‘Mini- 1963;185:914–919. levels and use of medical services. Am J Public
mental state’: A practical method for grading Health 1995;85:173–182.
the cognitive state of patients for the clinician.
J Psychiatr Res 1975;12:189–198.

328 Gerontology 1999;45:323–328 Cacciatore/Napoli/Abete/Marciano/Triassi/

Rengo
198.143.32.65 - 2/25/2016 7:12:07 AM
Rutgers University Alexander Library
Downloaded by: