Health and Safety Executive
A review of the current state of knowledge on tinnitus in relation to noise exposure and hearing loss
Prepared by the Health and Safety Laboratory for the Health and Safety Executive 2010
RR768 Research Report
Health and Safety Executive
A review of the current state of knowledge on tinnitus in relation to noise exposure and hearing loss
Kerry Poole Health and Safety Laboratory Harpur Hill Buxton Derbyshire SK17 9JN
This report details the results of a search of the published peer-reviewed literature investigating the relationship between tinnitus (ringing or buzzing in the ears), noise exposure at work and noise-induced hearing loss. A total of 12 citation databases (earliest date 1951) were searched which identified 252 publications, of which 34 were found to be relevant to the review. A number of studies have reported the prevalence of tinnitus in populations exposed to noise at work to be between 87.5% and 5.9%. Factors such as the type of participant (eg health surveillance, compensation claimant), the characteristics of the noise exposure and the definition of tinnitus used may contribute to this variability. Furthermore, four studies have shown that the prevalence of tinnitus in workers exposed to noise at work is significantly greater than in workers not exposed to noise. The majority of the published papers support the idea that there is an association between tinnitus and noise-induced hearing loss. The prevalence of tinnitus in those with hearing loss appears to be greater, and the hearing thresholds in those with tinnitus are higher. There is also a suggestion from one 15-year longitudinal study that tinnitus may be an early indicator of risk of the development of noise-induced hearing loss.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the author alone and do not necessarily reflect HSE policy.
HSE Books
© Crown copyright 2010
First published 2010
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the copyright owner.
Applications for reproduction should be made in writing to:Licensing Division, Her Majesty’s Stationery Office,St Clements House, 2-16 Colegate, Norwich NR3 1BQor by e-mail to [email protected]
ACKNOWLEDGEMENTS
I would like to thank the HSE information services team for conducting the search of citation databases and obtaining the references referred to in this report.
ii
CONTENTS
1 INTRODUCTION..........................................................................................1
2 METHODS ...................................................................................................2
3 RESULTS ....................................................................................................3 3.1 Papers reviewed.......................................................................................3 3.2 Question 1 – Is there a relationship between exposure to noise in work and tinnitus?........................................................................................................3 3.3 Question 2 – Is there a relationship between tinnitus and hearing loss in those exposed to noise at work?.......................................................................17
4 MAIN FINDINGS AND DISCUSSION........................................................334.1 Is there a relationship between exposure to noise at work and tinnitus?33 4.2 Is there a relationship between tinnitus and hearing loss in those exposed to noise at work?.................................................................................34
5 REFERENCES ..........................................................................................37
iii
iv
EXECUTIVE SUMMARY
This report details the results of a search of the published peer-reviewed literature investigating
the relationship between tinnitus (ringing or buzzing in the ears), noise exposure at work and
noise-induced hearing loss. A full review of these issues has not previously been published in
the peer-reviewed literature. A total of 12 citation databases (earliest date 1951) were searched
which identified 252 potentially relevant publications. Following an initial sift of the
corresponding abstracts, 34 publications were identified as being written in the English
language and relevant to the objectives of this review. These form the basis of this report.
Objectives
To review the peer-reviewed literature to address two main questions:
1. Is there a relationship between exposure to noise in work and tinnitus?
2. Is there a relationship between tinnitus and hearing loss in those exposed to noise at
work?
Main Findings
1. A number of studies have reported the prevalence of tinnitus in populations exposed to
noise at work (n=23). The prevalence values vary considerably (87.5% to 5.9%) and
factors such as the type of participant (e.g. health surveillance, compensation claimant),
the characteristics of the noise exposure and the definition of tinnitus used may
contribute to this variability.
2. Four studies have shown that the prevalence of tinnitus in workers exposed to noise at
work is significantly greater than in workers not exposed to noise. Eight studies have
also presented evidence to suggest that this is related to the severity (in terms of
exposure level and duration) of noise exposure.
3. The majority of the published papers support the idea that there is an association
between tinnitus and noise-induced hearing loss. The prevalence of tinnitus in those
with hearing loss appears to be greater (shown in 7 out of 9 papers), and the hearing
thresholds in those with tinnitus are higher (shown in 8 out of 9 papers).
4. The majority of the literature (33/34) included in this review are from cross-sectional
studies, which makes it difficult to establish whether there is a causal relationship
between tinnitus and hearing loss. However, one paper based upon the results of a 15-
year longitudinal study has suggested that tinnitus may be an early indicator of risk of
the development of noise-induced hearing loss.
Recommendations
It may be useful to widen the literature search to include general noise exposure, rather than
purely noise exposure at work. This may help to increase the evidence base, particularly in
terms of comparing the prevalence of tinnitus in individuals with and without a history of noise
exposure, and whether there is information clarifying any causal relationship between tinnitus
and hearing loss.
v
vi
1 INTRODUCTION
Tinnitus manifests itself as ringing, buzzing or other sounds in the ear, without an external
sound source. It may be intermittent or continuous. Some victims of tinnitus appear to tolerate
their symptoms with little difficulty. However, others suffer from a wide variety of difficulties
and the effects upon quality of life can be quite significant as it can affect concentration, ability
to sleep and psychological well-being.
The causes of tinnitus are many and varied and include diseases of the ear, exposure to ototoxic
drugs, cardiovascular problems, metabolic disorders and neurological problems [1]
. In addition,
a paper, which included a small review of papers relating to the prevalence of tinnitus in
populations exposed to industrial noise, suggested that noise exposure and noise induced
hearing loss were the most common causes of tinnitus [2]
. However, this was not a full review
of all the literature and it did not review the evidence for a dose-response relationship between
tinnitus and noise exposure, or all the literature investigating the relationship between hearing
loss and tinnitus.
The relationships between tinnitus and exposure to noise at work, and noise-induced hearing
loss are of interest. In particular, if tinnitus were caused by exposure to noise and occurred prior
to noise-induced hearing loss, it may be that reporting of tinnitus may be an early marker of risk
of developing noise-induced hearing loss. If this were found to be the case then asking about
tinnitus as part of a health surveillance programme for noise exposure at work would be of
value in helping to identify those at risk, and helping to persuade individuals to protect
themselves against the risk.
The current review aimed to interrogate the scientific peer-reviewed literature for evidence
related to the relationships between tinnitus, noise exposure and noise-induced hearing loss.
The specific questions addressed by this review are:
1. Is there a relationship between exposure to noise in work and tinnitus?
a. Is the prevalence of tinnitus in those exposed to noise in work greater than those
who aren’t exposed to noise in work?
b. Is there a relationship between tinnitus and the severity of exposure to noise?
2. Is there a relationship between tinnitus and hearing loss in those exposed to noise at
work?
a. Is the prevalence of tinnitus in those with noise-induced hearing loss greater
than that in those with normal hearing?
b. Are the hearing thresholds measured using audiometry greater in those with
tinnitus as compared to those who do not have tinnitus?
c. Is there evidence of an increased risk of tinnitus in those with hearing loss, or
an increased risk of hearing loss in those with tinnitus?
d. What is the temporal relationship between tinnitus and hearing loss?
1
2 METHODS
The information services search team at the Health and Safety Executive conducted the
literature search. The terms used for the search were:
Tinnitus
(Ear or ears) with (ring* or buzz*)
work*, employ*, occupation*
The following twelve databases were searched for relevant information:
CINAHL From 1981
HSELINE From 1974
OSHLINE From 1998
CISDOC From 1987
NIOSHTIC From 1971
RILOSH 1975-2004
MEDLINE From 1951
EMBASE From 1974
Healsafe From 1981
Web of Science From 1990
Ergonomics From 1981
OshUpdate From 1980
The search was restricted to literature in peer-reviewed scientific publications written in the
English language.
Each of the searches was exported in a text file format and then imported into a reference
manager (Endnote� 9). The abstracts were then reviewed by the author to determine which
references were relevant to the current review. Hard-copies of the publications that were
deemed relevant were then obtained and reviewed by the author.
2
3 RESULTS
3.1 PAPERS REVIEWED
The initial literature search identified 252 references that contained the selected keywords (see
Section 2). However, when the author reviewed the references it was found that as different
databases use different reference notation and nomenclature, that there were a substantial
number of duplicates in the reference manager file. In addition, the references were filtered
according to whether they were relevant to address the questions of this review, they were
written in the English language and they were published in peer-reviewed publications.
Following this process, 60 of the original 252 references identified by the search were obtained,
as they were thought to be potentially relevant to the subject of this review. On detailed
examination of the full references, it was found that only 34 of these contained information
relevant to the questions addressed in this review. Therefore, only these references form the
basis of the following report.
Of the 34 papers in the following review, 33 were from cross-sectional studies and only one was
from a 15-year longitudinal study. The populations involved in these studies were
predominantly (31/34 papers) those who were currently exposed, or had been exposed to
occupational noise. Three of the studies involved the use of general populations, with one
focussing on those over 55 years of age. However, all of the general population studies
reported, investigated the risk of occupational noise exposure.
3.2 QUESTION 1 – IS THERE A RELATIONSHIP BETWEEN EXPOSURE TO
NOISE IN WORK AND TINNITUS?
In order to establish whether there is a relationship between exposure to noise in work and
tinnitus, we aimed to review the literature to address the following questions:
1. Is the prevalence of tinnitus in those exposed to noise in work greater than those
who aren’t exposed to noise in work?
2. Is there a relationship between tinnitus and the severity of exposure to noise (i.e.
dose-response)?
3.2.1 Is the prevalence of tinnitus in those exposed to noise in work greater
than those who aren’t exposed to noise in work?
Twenty-three papers were identified that reported information on the prevalence of tinnitus in
populations exposed to noise at work (evidence table one). However, only four of these
publications also contained information related to the prevalence of tinnitus in a non-exposed
population [3-6]
. The prevalence of tinnitus in the non-exposed groups in these studies varied
between 2 and 14.4%, with the highest prevalence being found in a general population study [3]
.
In the same studies, the prevalence of tinnitus in those exposed to noise ranged between 12 and
70.4%. In all four studies the prevalence in the noise-exposed group was statistically
significantly higher than in the non-exposed group.
Nineteen references reported the prevalence of tinnitus in populations exposed to noise at work,
but did not contain prevalence information of those not exposed to noise. The prevalence
3
reported in these studies ranged between 87.5% and 5.9%. Two of the studies, which have
reported a very high prevalence of tinnitus, were conducted in drop-forge workers [6, 7]
. The
noise exposure in this industry is of an impulsive nature, rather than a continuous nature, and
this may suggest that impulsive noise leads to more tinnitus. Some of the other studies with the
highest prevalence of tinnitus have involved the use of claimants for noise-induced hearing loss [8, 9]
and one other has involved the use of individuals reported for noise-induced hearing loss to
the Register of Occupational Diseases in Finland [10]
. Conversely, there have been three very
large studies published, which have reported a relatively low prevalence of tinnitus in workers
exposed to noise [11-13]
. These studies have involved the use of noise-exposed populations of
110,647, 30,000 and 47,388 with reported prevalence of tinnitus of 6.7%, 6.6% and 5.9%
respectively.
Factors such as the definition of tinnitus used, especially whether transient effects after
exposure to noise are included, and the prevalence of noise induced hearing loss may also have
an effect upon the reported prevalence of tinnitus. Evidence table one shows that many studies
did not state the definition used for tinnitus and in many cases it was difficult to ascertain the
prevalence of noise induced hearing loss in the population studied.
4
Ev
iden
ce t
ab
le o
ne:
P
rev
alen
ce o
f ti
nn
itu
s in
po
pu
lati
on
s ex
po
sed
to
no
ise
at w
ork
(in
ord
er o
f p
rev
alen
ce o
f ti
nn
itu
s re
po
rted
in
no
ise
exp
ose
d
po
pu
lati
on
)
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Kam
al,
AA
[7]
19
89
8
8 d
rop
-fo
rge
wo
rker
s
Ex
po
sed
to
im
pac
t n
ois
e r
ang
ing
bet
ween
11
2 t
o 1
39
dB
(A)
at
a
rate
betw
een
20
-50
dro
ps
per
min
ute
No
t g
iven
B
etw
een
30
-60
yea
rs o
f ag
e
91
% (
rep
ort
ed
dif
ficu
lty
hea
rin
g t
he
hu
man
vo
ice)
87
.5%
Bra
y,
A [1
4]
20
04
2
3 d
isc j
ock
eys
So
un
d l
evels
up
to
10
8 d
B(A
)
wer
e re
cord
ed i
n n
igh
tclu
bs,
wit
h
an a
ver
age o
f 9
6 d
B(A
). N
o
info
rmat
ion
on
len
gth
of
exp
osu
re.
No
t g
iven
M
ean
29
(ran
ge
21
-41
)
yea
rs
13
%
74
%
Su
lko
wsk
i, W
[6]
19
99
2
61
dro
p f
org
e
op
erat
ors
Pea
k p
ress
ure
lev
el
for
imp
uls
e
no
ise
of
13
5.1
dB
.
No
t g
iven
M
ean a
ge
31
.2
yea
rs.
No
t g
iven
7
0.4
% i
n n
ois
e
exp
ose
d.
3.5
% i
n a
ge-
mat
ched
con
tro
ls (
chi-
squ
ared
18
5,
p<
0.0
00
1)
Mre
na,
R [1
0]
20
07
C
ases
of
no
ise-
ind
uce
d h
eari
ng
loss
rep
ort
ed i
n t
he
Mea
n d
ura
tio
n o
f ex
po
sure
of
24
.8
(ran
ge
0-5
0)
yea
rs
Tin
nit
us
men
tio
ned
(o
n
bas
is o
f a f
ull
tin
nit
us
qu
esti
on
nai
re)
Mea
n a
ge
51
.6
yea
rs
40
% (
>2
0d
B
hea
rin
g
thre
sho
lds)
.
67
.7%
5
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Reg
iste
r o
f
Occ
up
ati
on
al
Dis
eas
es i
n F
inla
nd
in 2
00
0.
36
6 i
nd
ivid
uals
wit
h d
isab
ilit
y
cate
go
ry a
nd
con
tact
info
rmat
ion
.
60
% w
ith m
ild
NIH
L (
<2
0 d
B).
Bar
rs, D
M [9
] 1
99
4
24
6 p
atie
nts
seek
ing
com
pen
sati
on
fo
r
no
ise-
ind
uced
hea
rin
g l
oss
Mea
n d
ura
tio
n o
f n
ois
e ex
po
sure
was
28
.1 y
ears
.
No
t g
iven
M
ean a
ge
56
.4
(SD
8.2
) y
ear
s
No
t g
iven
5
8%
Alb
erti
, P
W [8
] 1
98
7
2,4
42
wo
rker
s
clai
min
g
com
pen
sati
on
fo
r
hea
rin
g i
mp
airm
ent
in T
oro
nto
,
Can
ada.
Mea
n e
xp
osu
re t
o i
nte
nse
wo
rkp
lace
no
ise
was
25
.3 y
ear
s.
No
t g
iven
M
ean a
ge
of
56
yea
rs.
10
0%
5
8%
6
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Pra
sher
, D
[15
] 2
00
1
19
8 m
ill
wo
rker
s
exp
ose
d t
o n
ois
e
lev
els
85
-95
dB
(A)
for
a m
ean
of
12
yea
rs.
No
t g
iven
M
ean
(S
D)
for
tho
se w
ith
tin
nit
us
47
(9
)
yea
rs a
nd
39
(9)
yea
rs f
or
tho
se w
ith
ou
t
tin
nit
us.
No
t g
iven
5
2.5
%
Jan
sen
EJM
[16
] 2
00
9
24
1 p
rofe
ssio
nal
mu
sici
ans
No
t re
po
rted
N
ot
giv
en
23
-64
yea
rs
No
t re
po
rted
5
1%
McS
han
e, D
P [1
7]
19
88
3
,46
6 c
laim
ants
fo
r
no
ise
ind
uce
d
hea
rin
g l
oss
Mea
n e
xp
osu
re t
o w
ork
pla
ce
no
ise
was
26
.1 (
SD
11
.8)
yea
rs
Incl
ud
es b
oth
in
term
itte
nt
and
co
nti
nu
ou
s sy
mp
tom
s.
Do
es n
ot
exclu
de s
ym
pto
ms
foll
ow
ing
lo
ud
no
ise.
Av
erag
e a
ge
60
.5 (
ran
ge
16
-81
) y
ears
99
% (
had
hea
rin
g
thre
sho
ld
>2
0d
B)
49
.8%
Kah
ari,
K [1
8]
20
03
1
39
ro
ck/j
azz
mu
sici
ans
Mea
n 1
9 (
ran
ge
5-4
0)
year
s as
mu
sici
an.
No
t g
iven
M
edia
n a
ge i
n
wo
men
35
(2
6-
47
) y
ears
an
d
37
(2
6-5
1)
yea
rs i
n m
en
49
%
48
%
Yli
ko
ski,
ME
[19
] 1
99
4
78
6 F
inn
ish
arm
y
off
icer
s ex
po
sed
to
gu
nfi
re n
ois
e
No
t re
po
rted
D
efin
ed a
s eit
her
occa
sio
nal
or
con
tin
uo
us
Mea
n 3
9.8
(ran
ge
25
-61
)
yea
rs
68
% (
>2
0 d
B)
42
.5%
7
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Lin
dg
ren
, T
[20
] 2
00
9
41
8 m
ale a
nd
42
fem
ale
pil
ots
No
t re
po
rted
E
xp
erie
nce
d t
inn
itu
s fo
r
mo
re t
han
5 m
in d
uri
ng
pas
t
yea
r, r
epo
rted
co
nst
ant
or
sev
ere t
inn
itu
s
Mea
n
age
(SD
) 4
6 (
6.7
)
yea
rs
36
%
40
%
Ho
ng
, O
[21
] 2
00
6
62
3 c
on
stru
cti
on
wo
rker
s
Mea
n (
SD
) y
ears
in
co
nst
ructi
on
18
.14
(1
0.9
3)
yea
rs.
No
t g
iven
M
ean a
ge
(SD
)
was
42
.96
(9.9
8)
yea
rs.
60
% (
>2
5d
B a
t
4, 6
kH
z)
38
%
Ho
ng
, O
S [5
] 1
99
8
25
5 n
ois
e e
xp
ose
d
and
19
5 n
on
-no
ise
exp
ose
d f
ull
-tim
e
airp
ort
wo
rker
s
>8
5 d
B(A
) N
ot
giv
en
Mea
n 3
8.9
yea
rs i
n n
ois
e
exp
ose
d a
nd
37
.4 y
ears
in
no
n-e
xp
ose
d
In n
ois
e g
rou
p:
9.4
% N
IHL
at
low
fre
qu
enci
es
and
49
.4%
at
hig
h f
req
uen
cy.
In n
on
-no
ise
gro
up
: 0
.5%
an
d
6.7
%
resp
ecti
vel
y.
32
.2%
in
no
ise
exp
ose
d
6.7
% i
n c
on
tro
ls
(Ch
i-sq
uar
ed
41
.6, p
<0
.00
01
)
Gri
est,
SE
[22
] 1
99
8
91
male
em
plo
yee
s
exp
ose
d t
o n
ois
e
85
-10
1 d
B(A
) N
ot
giv
en
Ran
ge 1
8-4
1
yea
rs.
52
% (
hea
rin
g
thre
sho
ld �
15
dB
at
4k
Hz)
29
.7%
Beg
ault
, D
[23
] 1
99
8
64
co
mm
ercia
l
airl
ine p
ilo
ts
No
t re
po
rted
Q
ues
tio
n:
Do
yo
u h
ave a
bu
zzin
g,
rin
gin
g o
r
wh
istl
ing
in
on
e o
r b
oth
ear
s
(tin
nit
us)
?
Med
ian
ag
e 5
3
(ran
ge
35
-64
)
yea
rs
Betw
een
27
an
d
53
% (
dep
end
ing
on
ag
e)
29
.5%
8
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Ch
en,
J-D
[24
] 2
00
3
38
4 n
ois
e-ex
po
sed
wo
rker
s in
an
oil
refi
ner
y
Ex
po
sed
to
av
erag
e n
ois
e l
evel
of
73
-89
dB
(A).
M
ean
du
rati
on
of
exp
osu
re w
as
14
(S
D 7
) y
ear
s
Qu
esti
on
: D
o y
ou
ev
er h
ear
rin
gin
g i
n y
ou
r ea
rs?
Mea
n (
SD
) ag
e
39
(5
) y
ears
38
% (
hig
h
freq
uen
cy l
oss
)
28
%
Ph
oo
n, W
[25
] 1
99
3
64
7 n
ois
e-ex
po
sed
wo
rker
s n
oti
fied
as
case
s o
f n
ois
e-
ind
uce
d d
eafn
ess
Mea
n (
SD
) d
ura
tio
n o
f ex
po
sure
was
17
0 (
10
0.7
) m
on
ths
in t
ho
se
wit
h t
inn
itu
s an
d 1
60
(1
06
)
mo
nth
s in
th
ose
wit
ho
ut
tin
nit
us
Ex
per
ien
ced
an
y t
inn
itu
s
wit
hin
th
e p
ast
6 m
on
ths
Mea
n a
ge
39
yea
rs
34
% (
hea
rin
g
thre
sho
ld >
25
dB
)
23
.3%
9
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Co
les,
RR
A [3
] 1
98
1
6,8
04
in
div
idu
als
fro
m t
he g
ener
al
po
pu
lati
on
in
fo
ur
citi
es o
f th
e U
K
32
.3%
rep
ort
ed n
ois
e e
xp
osu
re
Qu
esti
on
: D
o y
ou
get
rin
gin
g o
r b
uzzi
ng
no
ises
in
yo
ur
hea
d o
r ear
s.
Th
e
occ
asi
on
al
wh
istl
ing
or
rin
gin
g i
n t
he e
ars
of
less
than
5 m
inu
tes
du
rati
on
sho
uld
no
t b
e c
ou
nte
d.
Als
o
do
no
t co
un
t th
ose
tim
es
wh
en t
his
hap
pen
s af
ter
ver
y
lou
d s
ou
nd
s, e
.g.
dis
cos,
sho
oti
ng
or
no
ise
at w
ork
.
38
% <
40
yea
rs
32
.8%
40
-60
yea
rs
29
.2%
>6
0
yea
rs
No
t g
iven
In
th
ose
ag
ed
40
-60
, th
e
pre
val
ence
of
tin
nit
us
in t
ho
se
no
t ex
po
sed
to
no
ise
was
13
.2%
com
par
ed t
o
21
.4%
in
th
ose
exp
ose
d.
In
tho
se o
ver
60
the
resp
ect
ive
pro
po
rtio
ns
wer
e 1
8.2
an
d
32
.8%
. O
ver
all
,
the
pre
val
ence
of
tin
nit
us
in
tho
se e
xp
ose
d
to n
ois
e w
as
24
% c
om
par
ed
to 1
4.4
% i
n
tho
se n
ot
exp
ose
d t
o
no
ise
(Ch
i-
squ
ared
69
.7,
p<
0.0
00
1).
10
Fir
st A
uth
or
Yea
r T
yp
e o
f su
bje
cts
No
ise e
xp
osu
re
Ca
se d
efin
itio
n f
or
tin
nit
us
Ag
e P
rev
ale
nce
of
no
ise-
ind
uce
d
hea
rin
g l
oss
Pre
va
len
ce
of
Tin
nit
us
Ch
avali
tsak
ulc
hai
,
P [4
] 1
98
9
1,5
31
wo
rker
s in
a
tex
tile
mil
l. 6
13
wea
ver
s, 1
05
off
ice
wo
rker
s an
d 8
13
do
ing
oth
er
acti
vit
ies.
Av
erag
e n
ois
e l
evels
in t
he
wea
vin
g s
ect
ion
an
d o
ther
secti
on
s w
ere
10
2.3
an
d 8
9.8
dB
(A)
resp
ecti
vel
y
No
t g
iven
M
ean
(S
D)
of
wea
ver
s w
as
29
(6
) y
ears
,
off
ice w
ork
ers
32
(8
) y
ears
and
oth
ers
28
(5)
yea
rs.
No
t g
iven
1
2%
in
no
ise
exp
ose
d
2%
in
co
ntr
ols
(Ch
i-sq
uar
ed
8.5
, p
=0
.00
35
).
Neu
ber
ger
, M
[13
] 1
99
2
11
0,6
47
no
ise
exp
ose
d w
ork
ers
>8
5 d
B f
or
at
leas
t 6
mo
nth
s.
No
t g
iven
A
ged
betw
een
15
-65
yea
rs.
Med
ian
app
rox
imate
ly
10
% (
Ro
eser
’s
spee
ch
imp
air
men
t
ind
ex,
tho
se
wit
h n
o i
nju
ry
or
dis
eas
e)
6.7
%
Ch
un
g, D
Y [1
2]
19
84
3
0,0
00
no
ise
exp
ose
d w
ork
ers
Mea
n o
f 8
5 d
B(A
) o
r m
ore
Q
ues
tio
n:
Do
yo
u h
ave
rin
gin
g i
n y
ou
r ea
rs?
(ex
clu
din
g t
ran
sien
t
tin
nit
us)
No
t re
po
rted
N
ot
giv
en
6.6
%
Bau
er, P
[11
] 1
99
1
47
,38
8 n
ois
e-
exp
ose
d w
ork
ers
Med
ian
lev
el
89
.8 d
B(A
) N
ot
giv
en
Med
ian
ag
e
36
.1 y
ears
No
t g
iven
5
.9%
11
3.2.2 Is there a relationship between tinnitus and the severity of exposure
to noise?
A total of 9 papers were identified which investigated the relationship between severity of noise
exposure and tinnitus (Evidence table two). Four of these publications used actual noise levels
(dB(A)) as the noise exposure metric [26-29]
. Of these four studies, one of them demonstrated a
slight increase in prevalence of tinnitus with increasing noise levels in workers in the cement
industry, but these differences were not statistically significant [29]
. Another study showed that
there was a weak relationship between the LAeq and tinnitus, which was statistically
significant[26]
. A further small study conducted in bar tenders and waitresses showed that
individuals working in the louder clubs experienced more frequent bouts of tinnitus [27]
. These
three studies did not investigate whether the occurrence of noise-induced hearing loss was also
an important factor. One final study which investigated the relationship between tinnitus and
actual noise levels found that the cumulative noise exposure (dB(A)-years) significantly
increased the risk of tinnitus (p=0.03), but only in those who had hearing loss [28]
.
Other studies have attempted to assess the severity of noise exposure by using self-reported
questionnaires. In two of the studies, asking how difficult it was to speak or be heard in the
work environment was used to establish the severity or level of noise exposure [30, 31]
. A study
by Sindhusake has shown that the relative risk of having tinnitus was related to the severity of
work-related noise exposure. In those individuals whose exposure was tolerable the relative risk
was 1.39 and in those where they were unable to hear speech, it was 1.53 [31]
. This study
involved the production of a predictive model for tinnitus and took into account confounding
factors such as age, gender and hearing loss. However, one limitation of this study was that it
was conducted in an older population with an average age of 69.8 years, therefore the relevance
of these findings to the working population is unknown. Another study involving the general
population has shown that the number of years working in a noisy job (defined as one where
you need to shout to be heard) was related to the prevalence ratio for tinnitus [30]
. In those
individuals who had worked in a noisy job for 1-5 years the prevalence ratio was 1.8 and this
increased to 2.6 in those who had worked in such an environment for more than 10 years. This
study went on to suggest that 266,300 men have tinnitus in the UK as a consequence of work-
related noise exposure.
Three studies have investigated the relationship between tinnitus and duration of noise exposure [6, 17, 32]
. One of these studies demonstrated that the risk of tinnitus in musicians was related to
the number of hours of practice per week [32]
. Another study showed that the prevalence of
tinnitus increased with the number of years of exposure in a drop-forge. However, whether
these changes were statistically significant is difficult to ascertain as there was no statistical
analysis included in this publication [6]
. One further study involving the use of claimants for
noise-induced hearing loss showed that the prevalence of tinnitus in those exposed for 0-10
years was 33.7% and that this rose to 54% in those exposed between 11-20 years. However,
with durations of exposure greater than 20 years, the prevalence remained fairly stable [17]
.
12
Ev
iden
ce t
ab
le t
wo
: R
elat
ion
ship
bet
wee
n t
inn
itu
s an
d s
ever
ity
of
exp
osu
re t
o n
ois
e
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise
exp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ru
bak
, T
[28
] 2
00
8
75
2 w
ork
ers
emp
loy
ed i
n 9
1
wo
rkp
lace
s.
Tin
nit
us
and
no
hea
rin
g h
and
icap
(n=
67
), T
inn
itu
s
and
hea
rin
g
han
dic
ap (
n=
50
)
and
No
tin
nit
us
(n=
63
5).
Av
erag
e
exp
osu
re i
n
ind
ust
rial
trad
es w
as
83
.7
dB
(A)
Mea
n a
ge
(SD
) in
con
tro
ls,
tho
se
wit
h t
inn
itu
s an
d
hea
rin
g h
and
icap
,
and
th
ose
wit
h
tin
nit
us
bu
t n
o
hea
rin
g h
and
icap
was
39
.5 (
9.1
),
46
.9 (
7.8
), 3
6.1
(7.1
) y
ears
resp
ecti
vel
y
Inv
esti
gat
ed t
he
od
ds
rati
o o
f ti
nn
itu
s w
ith
an
d
wit
ho
ut
con
com
itan
t h
ear
ing
han
dic
ap.
Hea
rin
g
han
dic
ap w
as
def
ined
as
an a
ver
age h
ear
ing
thre
sho
ld f
or
eit
her
ear
at
2,3
an
d 4
kH
z ab
ov
e 2
0
dB
. C
ou
ld n
ot
fin
d a
ny
in
creas
e i
n r
isk
of
tin
nit
us
wit
h a
ctu
al
no
ise l
evel,
du
rati
on
of
no
ise e
xp
osu
re
(yea
rs)
or
cum
ula
tiv
e n
ois
e e
xp
osu
re (
dB
(A)-
yea
rs)
in w
ork
ers
wh
o d
id n
ot
hav
e a
hear
ing
han
dic
ap.
In
wo
rker
s w
ho
had
a h
ear
ing
han
dic
ap
ther
e w
as a
n i
ncr
ease
in
od
ds
rati
o f
or
tin
nit
us
wit
h
incr
eas
ing
cu
mu
lati
ve
no
ise
exp
osu
re (
p=
0.0
3).
Ho
wev
er,
alt
ho
ug
h t
her
e w
as a
n i
ncr
ease
in
od
ds
rati
o w
ith
in
creas
ing
no
ise l
evel
an
d d
ura
tio
n o
f
no
ise e
xp
osu
re t
hes
e d
id n
ot
reach
sta
tist
ical
sig
nif
ican
ce.
Th
is s
tud
y s
ug
gest
s th
at t
inn
itu
s is
on
ly r
elat
ed t
o n
ois
e ex
po
sure
if
ther
e i
s als
o h
eari
ng
lo
ss.
Th
e n
um
ber
s in
th
e ti
nn
itu
s
gro
up
s w
ere
small
wh
ich
mea
nt
that
th
e co
nfi
den
ce
inte
rvals
fo
r
the
od
ds
rati
os
wer
e w
ide,
an
d
thu
s o
ther
rela
tio
nsh
ips
did
no
t
reac
h s
ign
ific
ance.
A s
tren
gth
of
this
stu
dy
was
th
at
actu
al n
ois
e l
evel
s w
ere
ob
tain
ed
usi
ng
per
son
al
do
sim
eter
s, a
nd
aud
iom
etr
y m
eas
ure
men
ts w
ere
also
ob
tain
ed.
Hag
ber
g,
M
[32
] 2
00
5
40
7 s
tud
ents
enro
lled
at
the
Sch
oo
l o
f M
usi
c
and
Mu
sic
Ed
uca
tio
n a
t
Go
teb
org
Un
iver
sity
Gro
up
ed l
evel
s
no
t g
iven
Mea
n a
ge
for
men
was
35
(ran
ge
23
-49
)
yea
rs a
nd i
n
wo
men
was
34
(ran
ge
24
-57
)
yea
rs
Qu
esti
on
nai
re b
ase
d s
tud
y.
Fo
un
d t
he
hig
hes
t
inci
den
ce
of
sym
pto
ms
was
for
rep
ort
ed t
inn
itu
s
wit
h a
rat
e o
f 1
0.6
per
10
00
year
s o
f in
stru
men
tal
pra
ctic
e. Im
pai
red
hear
ing
(b
y q
ues
tio
nn
aire
) h
ad
an i
nci
den
ce o
f 6
.5 p
er 1
00
0 i
nst
rum
ent
yea
rs.
Rep
ort
ed a
n i
ncr
eas
ed r
isk
of
tin
nit
us
in t
ho
se
pra
ctic
ing
fo
r m
ore
th
an 2
0 h
ou
rs p
er w
eek
pri
or
to d
evelo
pm
ent
of
the s
ym
pto
ms.
T
he r
ela
tiv
e r
isk
was
1.5
9 (
90
% c
on
fid
ence
inte
rval
of
0.9
4-2
.70
).
Su
gg
ests
th
at
no
ise e
xp
osu
re i
n
mu
sici
ans
can
lead
to t
inn
itu
s an
d
hea
rin
g i
mp
airm
ent.
H
ow
ever
,
hea
rin
g i
mp
airm
ent
was
jud
ged
by
qu
esti
on
nai
re r
ath
er t
han
aud
iom
etr
ic m
eas
ure
men
ts a
nd
no
actu
al n
ois
e le
vel
s ar
e re
po
rted
.
13
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise
exp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Sin
dh
usa
ke,
D [3
1]
20
03
2
,01
5 i
nd
ivid
ual
s
fro
m t
he g
ener
al
po
pu
lati
on
ag
ed
ov
er 5
5 y
ears
of
age
in S
yd
ney
,
Au
stra
lia.
Gro
up
ed l
evel
s
no
t g
iven
Mea
n a
ge
of
69
.8
yea
rs
Wh
en a
ge
and
gen
der
wer
e t
aken
in
to a
cco
un
t,
bo
th t
he d
ura
tio
n o
f n
ois
e e
xp
osu
re i
n y
ear
s an
d
the
sev
erit
y o
f w
ork
rela
ted
no
ise e
xp
osu
re (
self
-
rep
ort
ing
no
ne/q
uie
t, t
ole
rab
le, u
nab
le t
o h
ear
spee
ch)
wer
e si
gn
ific
ant
risk
fac
tors
fo
r ti
nn
itu
s.
Wh
en t
hes
e a
nd
oth
er f
acto
rs (
inclu
din
g h
ear
ing
loss
) w
ere e
nte
red
in
to a
pre
dic
tiv
e m
od
el
(in
wh
ich e
ach
var
iab
le w
as a
dju
sted
fo
r all
th
e
oth
ers)
th
ese
facto
rs w
ere
sig
nif
ican
t an
d
dem
on
stra
ted
a d
ose
-resp
on
se.
Fo
r ex
am
ple
, th
e
rela
tiv
e ri
sk f
or
tho
se w
ho
rep
ort
ed t
hat
th
eir
wo
rk-r
elat
ed n
ois
e e
xp
osu
re w
as t
ole
rab
le w
as
1.3
9 a
nd
fo
r th
ose
wh
o r
epo
rted
th
ey w
ere
un
able
to h
ear
sp
eec
h i
t w
as 1
.53
. O
ver
all
, th
e s
tud
y
rep
ort
s th
at
wo
rk-r
elat
ed n
ois
e e
xp
osu
re m
ay h
ave
cau
sed
13
.6%
of
tin
nit
us
case
s in
th
is c
om
mu
nit
y.
Th
is s
tud
y s
ho
ws
that
no
ise
exp
osu
re l
ead
s to
a g
reate
r ri
sk o
f
tin
nit
us
in a
do
se d
epen
den
t w
ay.
Th
is w
as i
nd
epen
den
t o
f h
ear
ing
loss
.
Th
is s
tud
y w
as p
erfo
rmed
in
an
old
er a
ge
gro
up
fro
m t
he g
ener
al
po
pu
lati
on
. I
n t
his
po
pu
lati
on
wo
rk-r
elat
ed n
ois
e e
xp
osu
re m
ay
hav
e c
ause
d 1
3.6
% o
f ti
nn
itu
s
case
s. H
ow
ever
, it
sh
ou
ld b
e
bo
rne
in m
ind
th
at
aro
un
d 6
5%
of
this
po
pu
lati
on
wer
e n
ot
exp
ose
d
to s
ign
ific
ant
no
ise l
evel
s.
Pal
mer
, K
T
[30
] 2
00
2
12
,90
7 i
nd
ivid
uals
fro
m t
he g
ener
al
po
pu
lati
on
Gro
up
ed l
evel
s
no
t g
iven
Betw
een
35
an
d
64
yea
rs o
f ag
e
Pre
val
ence
rati
o f
or
tin
nit
us
incr
ease
d w
ith
incr
eas
ing
yea
rs w
ork
ing
in
a n
ois
y j
ob
(re
sult
s
adju
sted
fo
r ag
e).
A n
ois
y j
ob
was
def
ined
as
that
wh
ere
they
had
to
sh
ou
t to
be
hear
d.
Fo
r ex
amp
le,
if t
hey
had
wo
rked
fo
r b
etw
een
1 a
nd
5 y
ears
in
a
no
isy
jo
b t
he p
rev
ale
nce
rat
io w
as 1
.8 a
nd
th
is
incr
eas
ed t
o 2
.6 i
n t
ho
se w
ho
had
wo
rked
fo
r m
ore
than
10
yea
rs i
n a
no
isy
jo
b.
Th
e n
um
ber
of
men
in G
reat
Bri
tain
wh
o h
ave
tin
nit
us
att
rib
uta
ble
to
occ
up
ati
on
al
no
ise e
xp
osu
re w
as e
stim
ate
d a
s
26
6,3
00
an
d t
he n
um
ber
of
wo
men
was
84
,00
0.
Th
is s
tud
y d
em
on
stra
ted
a
rela
tio
nsh
ip b
etw
een
th
e
pre
val
ence
of
tin
nit
us
and
len
gth
of
tim
e o
f ex
po
sure
to
a n
ois
y
wo
rk e
nv
iro
nm
ent
in a
gen
eral
po
pu
lati
on
.
Th
is w
as b
ased
up
on
self
-
rep
ort
ing
in
a q
ues
tio
nn
air
e b
ased
stu
dy
, ra
ther
th
an a
ctu
al n
ois
e
mea
sure
men
ts.
14
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise
exp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Su
lko
wsk
i,
W
[6]
19
99
2
61
dro
p f
org
e
op
erat
ors
Pea
k p
ress
ure
lev
el
for
imp
uls
e n
ois
e
of
13
5.1
dB
Mea
n a
ge
31
.2
yea
rs
Th
is s
tud
y a
imed
to
in
ves
tig
ate
the
effe
ct o
f
imp
uls
e n
ois
e o
n t
inn
itu
s an
d h
ear
ing
lo
ss.
As
par
t
of
this
pu
bli
cati
on
, in
form
atio
n o
n t
he p
rev
ale
nce
of
tin
nit
us
and
ex
po
sure
du
rati
on
was
pre
sen
ted
.
Per
cen
tag
e o
f in
div
idu
als
wit
h t
inn
itu
s in
crea
sed
wit
h y
ears
of
exp
osu
re.
Pre
vale
nce
per
cen
tag
es
wit
h y
ears
of
exp
osu
re w
ere
31
.6,
65
.5,
84
.9, 9
5.1
,
92
.5 a
nd
88
.9%
fo
r ex
po
sure
du
rati
on
s o
f 1
-3,
3-5
,
5-1
0, 1
0-1
5, 1
5-2
0, 2
0-2
5 y
ears
.
Th
e p
rev
alen
ce
of
tin
nit
us
was
rela
ted
to
th
e e
xp
osu
re d
ura
tio
n,
bu
t n
o s
tati
stic
al
analy
sis
was
con
tain
ed w
ith
in t
his
pap
er.
Sal
lust
io,
V
[29
] 1
99
8
18
0 w
ork
ers
fro
m
two
cem
ent
ind
ust
ries
Th
ree g
rou
ps
exp
ose
d t
o <
80
dB
(A)
(co
ntr
ols
),
bet
ween
80
-85
dB
(A)
and
tho
se b
etw
een
86
-90
dB
(A)
Mea
n 4
4.9
(ra
ng
e
21
-63
) y
ears
Ov
eral
l, 1
4.3
% o
f w
ork
ers
had
co
nti
nu
ou
s ti
nn
itu
s,
wh
ich
beg
an m
ore
th
an o
ne y
ear
bef
ore
. T
her
e
wer
e sl
igh
t d
iffe
ren
ces
in p
rev
ale
nce
of
tin
nit
us
in
the
thre
e n
ois
e g
rou
ps
(6.3
%,
13
.9%
an
d 1
8.1
%
resp
ecti
vel
y)
bu
t th
ese
wer
e n
ot
stati
stic
all
y
sig
nif
ican
t.
Dif
fere
nce
s b
etw
een
gro
up
s w
ith
dif
fere
nt
no
ise
exp
osu
res
wer
e
no
n-s
ign
ific
ant.
Gu
nd
erso
n,
E [2
7]
19
97
3
1 e
mp
loy
ees
(bar
ten
der
s, w
aite
rs
and
wait
ress
es)
fro
m 8
liv
e m
usi
c
clu
bs
in N
ew
Yo
rk
Av
erag
e
per
form
ance
sou
nd
lev
els
ran
ged
fro
m
94
.9 t
o 1
06
.7
dB
(A).
Am
bie
nt
sou
nd
lev
els
ran
ged
fro
m 8
3.7
to
97
.1 d
B(A
).
No
t re
po
rted
F
ou
nd
rel
ati
on
ship
betw
een
th
e n
um
ber
of
ind
ivid
uals
ex
per
ien
cin
g t
inn
itu
s af
ter
wo
rk a
nd
the
no
ise l
evels
in t
he
clu
bs,
wit
h i
nd
ivid
uals
wo
rkin
g i
n t
he
lou
der
clu
bs
exp
erie
ncin
g t
he
gre
ater
fre
qu
ency
of
tin
nit
us
(p=
0.0
1).
Sm
all
stu
dy
bu
t d
oes
seem
to
sug
ges
t th
at
the p
rev
ale
nce
and
freq
uen
cy o
f sy
mp
tom
s o
f ti
nn
itu
s
foll
ow
ing
wo
rk i
s re
late
d t
o t
he
no
ise
lev
el.
15
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise
exp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
McS
han
e, D
P [1
7]
19
88
3
,46
6 c
laim
ants
for
no
ise
ind
uce
d
hea
rin
g l
oss
No
in
form
ati
on
reg
ard
ing
th
e
actu
al n
ois
e
lev
els
, o
ther
than
du
rati
on
of
exp
osu
re,
is
rep
ort
ed
Av
erag
e a
ge 6
0.5
(ran
ge
16
-81
)
yea
rs
Inv
esti
gat
ed t
he
rela
tio
nsh
ip b
etw
een
th
e d
ura
tio
n
of
occ
up
ati
on
al
no
ise e
xp
osu
re (
yea
rs)
and
th
e
pre
val
ence
of
tin
nit
us.
F
ou
nd
th
at
the
pre
val
ence
of
tin
nit
us
was
33
.7%
in
th
ose
ex
po
sed
fo
r
bet
ween
0 a
nd
10
year
s.
Th
is t
hen
ro
se t
o 5
4%
in
tho
se e
xp
ose
d f
or
11
-20
year
s, a
nd
th
en s
tay
ed
rela
tiv
ely
sta
ble
.
Su
gg
ests
th
at
afte
r 1
0 y
ears
exp
osu
re t
he
pre
val
ence
of
tin
nit
us
is i
nd
epen
den
t o
f ti
me
of
exp
osu
re.
Van
Dij
k,
FJH
[26
] 1
98
7
53
9 m
ale
wo
rker
s
in s
even
in
du
stri
es
Dis
trib
uti
on
of
per
son
al n
ois
e
exp
osu
re l
evels
was
13
%
<8
6d
B(A
),
28
% 8
6-9
0
dB
(A),
33
%
91
-95
dB
(A),
and
25
% >
95
dB
(A)
21
% <
25
year
s,
31
% 2
5-3
4 y
ears
,
33
% 3
5-4
9 y
ears
,
9%
40
-54
yea
rs,
6%
55
-65
yea
rs
Rep
ort
a w
eak
rela
tio
nsh
ip b
etw
een
no
ise
exp
osu
re (
LA
eq)
and
tin
nit
us
(p<
0.0
5).
Au
tho
rs s
ug
ges
t th
at
the
rela
tio
nsh
ip b
etw
een
tin
nit
us
and
hea
rin
g l
oss
is
stro
ng
er t
han
th
at
bet
ween
tin
nit
us
and
no
ise l
evel.
Ind
ivid
ual
data
fo
r th
e
rela
tio
nsh
ip b
etw
een
tin
nit
us
and
no
ise l
evel
are
no
t p
rese
nte
d i
n
the
pap
er.
16
3.3 QUESTION 2 – IS THERE A RELATIONSHIP BETWEEN TINNITUS ANDHEARING LOSS IN THOSE EXPOSED TO NOISE AT WORK?
In order to establish whether there is truly a relationship between tinnitus and hearing loss we
aimed to review the literature to address the following questions:
1. Is the prevalence of tinnitus in those with noise-induced hearing loss greater than
than in those with normal hearing?
2. Are the hearing thresholds measured using audiometry greater in those with tinnitus
as compared to those who do not have tinnitus?
3. Is there evidence of an increased risk of tinnitus in those with hearing loss, or an
increased risk of hearing loss in those with tinnitus?
4. What is the temporal relationship between tinnitus and hearing loss?
3.3.1 Is the prevalence of tinnitus in those with noise-induced hearing loss greater than in those with normal hearing?
Evidence table three details the papers reviewed in this area. A total of 12 papers were
identified in this area and they tended to fall into three main areas:
1. Those comparing the prevalence of tinnitus in those with or without noise-induced
hearing loss
2. Those investigating a relationship between the severity of hearing loss and the
prevalence of tinnitus
3. Those reporting the prevalence of tinnitus and noise-induced hearing loss in the
same population
Two studies have compared the prevalence of tinnitus in populations with or without noise-
induced hearing loss and have demonstrated an increase in prevalence of tinnitus in those with
hearing loss [22, 30]
. One study, which was a 15-year longitudinal study involving noise-exposed
workers, found that the prevalence of tinnitus in workers with �10 dB hearing thresholds was
16% compared to 42% in workers with �15 dB hearing thresholds (p=0.005) [22]
. Another study
by Palmer, which used self-reported hearing difficulty as their measure of hearing loss, showed
that the age standardised prevalence of persistent tinnitus was 16.1% in those who reported
severe difficulties in hearing, as compared to 5% in those with slight or no difficulties in hearing [30]
.
Other studies have investigated the relationship between the severity of hearing loss and the
prevalence of tinnitus [8, 12, 17, 19, 25, 26, 33]
. One particularly large study involved 30,000 workers
involved in a hearing conservation programme (similar to health surveillance) and found a clear
non-linear exponential relationship between the prevalence of tinnitus and extent of hearing loss
at all frequencies [12]
. Another study suggested a linear relationship between the prevalence of
tinnitus and the severity of hearing loss as defined by their audiometric categorisation. However
this paper did not contain any statistical analysis, so the validity of these findings is difficult to
17
ascertain [33]
. Other studies have also found an increase in the prevalence of tinnitus with
increasing hearing loss [19, 25, 26]
. However, two studies have failed to establish a relationship
between these two factors [8, 17]
. Both of these studies had quite large sample sizes, but only
consisted of compensation claimants. Whether this has contributed to these findings is unclear.
Three studies reported both the prevalence of tinnitus and noise-induced hearing loss in the
populations studied [14, 21, 23]
. Two of these studies reported a greater prevalence of hearing loss
compared to tinnitus [21, 23]
whereas the other reported a greater prevalence of tinnitus [14]
. In
none of the studies were the prevalence of noise-induced hearing loss and tinnitus the same.
These three studies were conducted in different populations (construction workers, disc jockeys
and airline pilots), who would be likely to have different noise exposure characteristics.
18
Ev
iden
ce t
ab
le t
hre
e – P
rev
alen
ce o
f ti
nn
itu
s w
ith
an
d w
ith
ou
t h
eari
ng
lo
ss
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ho
ng
, O
[21
] 2
00
6
62
3 c
on
stru
cti
on
wo
rker
s
Mea
n (
SD
)
18
.14
(1
0.9
3)
yea
rs i
n
con
stru
cti
on
.
Mea
n (
SD
)
42
.96
(9
.98
)
yea
rs.
38
% o
f co
nst
ructi
on
wo
rker
s re
po
rted
rin
gin
g o
r b
uzzi
ng
in
th
e ea
r an
d 6
2%
ind
icat
ed h
avin
g p
rob
lem
s in
un
der
stan
din
g w
hat
peo
ple
say
in
lo
ud
no
ise.
O
ver
60
% o
f w
ork
ers
sho
wed
hea
rin
g l
oss
(h
eari
ng
th
resh
old
�2
5d
B)
at
4
and
6k
Hz.
Th
is s
tud
y g
ives
an
ov
eral
l p
rev
alen
ce
of
tin
nit
us
bu
t d
oes
no
t re
late
th
is t
o
hea
rin
g l
oss
.
Bra
y,
A [1
4]
20
04
2
3 d
isc j
ock
eys.
S
ou
nd
lev
els
up
to 1
08
dB
(A)
wer
e re
cord
ed i
n
nig
htc
lub
s, w
ith
an a
ver
age o
f 9
6
dB
(A).
N
o
info
rmat
ion
on
len
gth
of
exp
osu
re.
Mea
n 2
9
(ran
ge
21
-41
)
yea
rs
Th
ree d
isc
jock
eys
(13
%)
sho
wed
ev
iden
ce
of
no
ise i
nd
uced
hear
ing
lo
ss.
74
%
rep
ort
ed t
inn
itu
s, w
ith
12
% e
xp
erie
ncin
g
this
all
of
the
tim
e.
Sm
all
stu
dy
. P
rese
nts
so
me d
ata
to
sho
w a
hig
h p
rev
ale
nce
of
tin
nit
us
in
dis
c j
ock
eys
exp
ose
d t
o n
ois
e in
nig
htc
lub
s. T
he
pre
val
ence
of
no
ise
ind
uce
d h
eari
ng
lo
ss w
as
mu
ch l
ow
er.
19
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Pal
mer
, K
T [3
0]
20
02
1
2,9
07
in
div
idu
als
fro
m t
he g
ener
al
po
pu
lati
on
Gro
up
ed l
evel
s
no
t g
iven
Betw
een
35
and
64
yea
rs
of
age
Qu
esti
on
nai
re b
ase
d s
tud
y w
ith
no
hea
rin
g
or
no
ise
meas
ure
men
ts.
Au
tho
rs r
epo
rt
that
th
e se
lf r
epo
rts
of
tin
nit
us
and
hear
ing
dif
ficu
ltie
s w
ere
stro
ng
ly a
sso
cia
ted
. I
n
men
, th
e a
ge
stan
dar
dis
ed p
rev
ale
nce
of
per
sist
ent
tin
nit
us
was
16
.1%
in
th
ose
wh
o
rep
ort
ed s
ever
e d
iffi
cult
ies
in h
ear
ing
,
com
par
ed w
ith
5%
in
th
ose
wit
h s
lig
ht
or
no
dif
ficu
ltie
s o
f h
ear
ing
. T
his
dif
fere
nce
was
gre
ater
in
wo
men
wit
h 3
3.1
% a
nd
2.6
% r
esp
ect
ively
.
Th
is s
tud
y w
as a
gen
eral
po
pu
lati
on
qu
esti
on
nai
re b
ased
stu
dy
an
d s
ho
wed
that
th
e p
rev
alen
ce
of
tin
nit
us
was
rela
ted
to
self
-rep
ort
ed d
iffi
cult
ies
in
hea
rin
g.
Th
is s
tud
y r
elie
d u
po
n s
elf-
rep
ort
ing
of
hea
rin
g d
iffi
cult
ies,
rath
er t
han
mea
sure
men
t o
f h
eari
ng
th
resh
old
s
usi
ng
au
dio
metr
y.
Beg
ault
, D
R [2
3]
19
98
6
4 c
om
mer
cia
l
airl
ine p
ilo
ts
No
t re
po
rted
M
edia
n a
ge 5
3
(ran
ge
35
-64
)
yea
rs
29
.5%
of
pil
ots
rep
ort
ed o
ccas
ion
al o
r
freq
uen
t ti
nn
itu
s. T
he p
rop
ort
ion
rep
ort
ing
per
man
ent
hea
rin
g l
oss
(co
nfi
rmed
by
th
eir
do
cto
r) w
as
betw
een
27
an
d 5
3%
,
dep
end
ing
up
on
th
eir
age.
Th
e p
rev
alen
ce
of
hear
ing l
oss
in
th
is
stu
dy
was
gen
eral
ly g
reate
r th
an t
he
pre
val
ence
of
tin
nit
us.
T
his
stu
dy
did
no
t p
rese
nt
dat
a sh
ow
ing
th
e
pre
val
ence
of
tin
nit
us
wit
h o
r w
ith
ou
t
hea
rin
g l
oss
.
20
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Gri
est,
SE
[22
] 1
99
8
91
male
em
plo
yee
s
wo
rkin
g i
n n
ois
y
env
iro
nm
ent
85
-10
1d
B(A
)
for
a m
inim
um
of
12
yea
rs
Ag
e ra
ng
e o
f
ind
ivid
uals
at
beg
inn
ing
of
the
stu
dy
was
18
-41
yea
rs.
Th
is s
tud
y w
as a
15
-year
lo
ng
itu
din
al
stu
dy
.
Lo
ok
ed a
t th
e m
axim
um
th
resh
old
sh
ift
in
aud
iom
etr
ic t
est
ing
at
40
00
Hz o
ver
th
e 1
5-
yea
r p
erio
d.
Sp
lit
into
tw
o g
rou
ps
acco
rdin
g t
o m
axim
um
sh
ift:
�1
0d
B o
r
�1
5d
B a
nd
co
mp
ared
to
pre
vale
nce
of
rep
ort
ed t
inn
itu
s.
Pre
val
ence
of
tin
nit
us
in �
10
dB
(n
=4
4)
and
�1
5d
B (
n=
47
) g
rou
ps
was
16
an
d 4
2%
resp
ecti
vel
y (
p=
0.0
05
).
Th
is s
tud
y d
em
on
stra
ted
th
at t
he
pre
val
ence
of
tin
nit
us
is g
reate
r in
tho
se w
ith
hear
ing
lo
ss (�
15
dB
at
4k
Hz)
. T
he s
tud
y o
nly
lo
ok
ed a
t
4k
Hz.
Th
e s
tren
gth
of
this
stu
dy
is
that
it i
s a
15
-yea
r lo
ng
itu
din
al
stu
dy
of
ind
ivid
uals
wit
hin
a h
eari
ng
con
serv
ati
on
pro
gra
mm
e.
Yli
ko
ski,
ME
[19
] 1
99
4
78
6 F
inn
ish
arm
y
off
icer
s ex
po
sed
to
gu
nfi
re n
ois
e
No
t re
po
rted
M
ean
39
.8
(ran
ge
25
-61
)
yea
rs
42
.5%
of
sub
ject
s re
po
rted
th
at t
hey
exp
erie
nce
d t
inn
itu
s ei
ther
occ
asio
nal
ly o
r
con
tin
uo
usl
y.
Th
e o
ccu
rren
ce o
f ti
nn
itu
s
was
rel
ated
to
th
e le
vel
of
hea
rin
g l
oss
. F
or
exam
ple
, th
ose
rep
ort
ing
co
nti
nu
ou
s
tin
nit
us
was
2.2
% i
n t
ho
se w
ith
no
rmal
hea
rin
g,
2.7
% w
ith
sli
gh
t to
mo
der
ate
hea
rin
g l
oss
(h
eari
ng
th
resh
old
>2
0 d
B a
nd
bel
ow
64
dB
at
som
e o
f fr
equ
encie
s fr
om
3-8
kH
z an
d i
n l
ow
-fre
qu
ency
ran
ge
<2
0
dB
), 1
9.8
% w
ith
sev
ere h
eari
ng
lo
ss
(hea
rin
g t
hre
sho
ld �
65
dB
at
som
e o
f
freq
uen
cies
fro
m 3
-8k
Hz
and
in
lo
w-
freq
uen
cy r
ang
e <
20
dB
) an
d 2
5.5
% w
ith
dis
abli
ng
hear
ing
lo
ss (
hea
rin
g t
hre
sho
ld o
f
>2
0 d
B a
t lo
w f
req
uen
cies
).
Th
ese
data
wer
e n
ot
anal
yse
d s
tati
stic
all
y.
Stu
dy
seem
s to
sh
ow
th
at
tin
nit
us
is
rela
ted
to
hea
rin
g l
oss
in
th
ose
ex
po
sed
to g
un
fire
no
ise.
Sta
tist
ical
sig
nif
ican
ce o
f re
sult
s
un
clear
.
21
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ph
oo
n, W
H [2
5]
19
93
6
47
no
ise e
xp
ose
d
wo
rker
s n
oti
fied
as
case
s o
f n
ois
e
ind
uce
d d
eafn
ess
Mea
n (
SD
)
du
rati
on
of
exp
osu
re w
as
17
0 (
10
0.7
)
mo
nth
s in
th
ose
wit
h t
inn
itu
s an
d
16
0 (
10
6)
mo
nth
s in
th
ose
wit
ho
ut
tin
nit
us
Mea
n a
ge
39
yea
rs
Sp
lit
into
gro
up
s b
ased
up
on
av
erag
e
hea
rin
g l
oss
lev
el
(AH
L)
at
1,
2 a
nd
3k
Hz.
‘Ear
ly’
case
s h
ad A
HL
of �
25
dB
,
‘in
term
edia
te c
ase
s’ h
ad >
25
dB
bu
t <
50
dB
, ‘l
ate
’ cas
es
had
AH
L o
f �
50
dB
.
Th
ere w
as a
n i
ncr
ease
in
pre
val
ence
of
tin
nit
us
bet
wee
n t
he
‘ear
ly’
and
‘in
term
edia
te’
case
s (2
0%
ver
sus
30
%),
and
th
e ‘
ear
ly’
and
‘la
te’
case
s (2
0%
ver
sus
27
.3%
) (p
=0
.02
). 3
0%
of
suff
erer
s
rep
ort
ed t
hat
this
in
terf
ered
wit
h d
aily
acti
vit
ies
e.g
. te
lep
ho
ne c
on
ver
sati
on
.
Th
is i
s a f
air
ly l
arg
e s
tud
y i
nv
olv
ing
a
ran
ge
of
hear
ing
lo
ss i
n a
fai
rly
yo
un
g
po
pu
lati
on
.
It s
eem
s to
su
gg
est
that
tin
nit
us
is
rela
ted
to
hea
rin
g l
oss
. H
ow
ever
, it
did
no
t co
nta
in a
gro
up
th
at
did
no
t h
ave
hea
rin
g l
oss
.
McS
han
e, D
P [1
7]
19
88
3
,46
6 c
laim
ants
fo
r
no
ise
ind
uce
d
hea
rin
g l
oss
Mea
n e
xp
osu
re
to w
ork
pla
ce
no
ise
was
26
.1
(SD
11
.8)
year
s
Av
erag
e a
ge
60
.5 (
ran
ge
16
-81
) y
ears
Lo
ok
ed a
t th
e p
rev
ale
nce
of
tin
nit
us
wit
h
10
dB
in
crem
ents
of
hea
rin
g l
oss
at
4 k
Hz.
Th
ere w
as n
o r
ela
tio
nsh
ip b
etw
een
th
e
pre
val
ence
of
tin
nit
us
and
ex
ten
t o
f h
eari
ng
loss
in
th
e ri
gh
t ea
r.
Th
is s
tud
y d
oes
no
t su
pp
ort
th
e i
dea
that
th
ere i
s a l
ink
betw
een
tin
nit
us
and
exte
nt
of
hea
rin
g l
oss
.
Th
is s
tud
y d
id n
ot
inclu
de a
co
ntr
ol
po
pu
lati
on
wit
ho
ut
hea
rin
g l
oss
an
d
pu
rely
co
nsi
sted
of
tho
se c
laim
ing
fo
r
hea
rin
g l
oss
.
Alb
erti
, P
W [8
] 1
98
7
2,4
42
wo
rker
s
clai
min
g
com
pen
sati
on
fo
r
hea
rin
g i
mp
airm
ent
Mea
n e
xp
osu
re
was
25
.3 y
ears
Mea
n a
ge
of
56
yea
rs
Th
e d
ata
was
sp
lit
into
3 c
ateg
ori
es
acco
rdin
g t
o t
he a
ver
age h
ear
ing l
oss
(4
0
dB
or
less
, 4
1-6
9 d
B,
and
70
dB
or
mo
re).
Th
e p
rev
alen
ce
of
tin
nit
us
in t
hes
e th
ree
cate
go
ries
of
hea
rin
g l
oss
fo
r th
e ‘i
mp
uls
e
no
ise’
gro
up
was
67
, 6
3 a
nd
70
%
resp
ecti
vel
y.
Fo
r th
e ‘s
tead
y-s
tate
’ n
ois
e
gro
up
th
ey w
ere
48
, 4
7 a
nd
57
%
resp
ecti
vel
y.
Th
is s
tud
y w
as m
ain
ly d
escr
ipti
ve a
nd
had
no
sta
tist
ical
analy
sis
inclu
ded
.
Th
e au
tho
rs c
on
clu
de
that
im
pu
lse
no
ise
lead
s to
a g
reat
er p
rev
alen
ce
of
tin
nit
us,
bu
t th
at
tin
nit
us
was
no
t
rela
ted
to
th
e e
xte
nt
of
hea
rin
g l
oss
.
Van
Dij
k,
FJH
[26
] 1
98
7
53
9 m
ale
wo
rker
s in
D
istr
ibu
tio
n o
f 2
1%
< 2
5
Hea
rin
g l
oss
def
ined
as
the l
arg
est
loss
at
Th
is s
tud
y s
ug
gest
s th
at t
her
e i
s a l
ink
22
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
sev
en i
nd
ust
ries
. p
erso
nal
no
ise
exp
osu
re l
evels
was
13
%
<8
6d
B(A
), 2
8%
86
-90
dB
(A),
33
% 9
1-9
5
dB
(A),
an
d 2
5%
> 9
5 d
B(A
)
yea
rs,
31
% 2
5-
34
yea
rs,
33
%
35
-49
yea
rs,
9%
40
-54
yea
rs,
6%
55
-
65
yea
rs
3, 4
, o
r 6
kH
z f
or
the l
eft
or
rig
ht
ear
.
Fo
un
d t
hat
th
e p
rev
alen
ce o
f ti
nn
itu
s w
as
rela
ted
to
th
e e
xte
nt
of
hea
rin
g l
oss
.
Wo
rker
s w
ith a
hea
rin
g l
oss
of �
15
dB
had
a 7
% g
rou
p p
rev
ale
nce
of
tin
nit
us
and
tho
se w
ith
40
dB
lo
ss h
ad a
pre
vale
nce
of
22
% (
p<
0.0
5).
bet
ween
hea
rin
g l
oss
an
d o
ccu
rren
ce
of
tin
nit
us.
Miy
akit
a,
T [3
3]
19
86
3
62
mal
e w
ork
ers
exp
ose
d t
o
occ
up
ati
on
al
imp
uls
ive n
ois
e
Mea
n d
ura
tio
n
was
> 3
0 y
ears
Mea
n 5
1.3
(ran
ge
48
-53
)
yea
rs
Th
is s
tud
y l
oo
ked
at
the c
lass
ific
atio
n o
f
aud
iom
etr
ic m
eas
ure
men
ts,
bu
t as
par
t o
f
this
th
ey a
lso
pre
sen
ted
dat
a re
late
d t
o t
he
pre
val
ence
of
tin
nit
us
and
au
dio
metr
y.
Th
ey f
ou
nd
th
at a
s th
e se
ver
ity
of
hea
rin
g
loss
(ac
cord
ing
to
cat
ego
risa
tio
n)
incr
ease
d
so d
id t
he
pre
vale
nce
of
tin
nit
us.
Th
is
app
eare
d t
o i
ncr
ease
in a
lin
ear
way
bu
t n
o
stati
stic
al a
nal
ysi
s w
as p
rese
nte
d,
and
it
is
no
t cle
ar w
heth
er t
he d
iffe
ren
ces
in
pre
val
ence
wit
h s
ever
ity
are
sta
tist
ical
ly
sig
nif
ican
t.
Stu
dy
su
gg
ests
th
at
ther
e i
s a
rela
tio
nsh
ip b
etw
een
th
e p
rev
ale
nce
of
tin
nit
us
and
sev
erit
y o
f h
eari
ng
lo
ss i
n
no
ise
exp
ose
d w
ork
ers.
23
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ch
un
g, D
Y [1
2]
19
84
3
0,0
00
wo
rker
s
exp
ose
d t
o n
ois
e,
inv
olv
ed i
n a
hear
ing
con
serv
ati
on
pro
gra
mm
e
Mea
n o
f 8
5
dB
(A)
or
mo
re
No
t re
po
rted
In
ves
tig
ated
th
e re
lati
on
ship
bet
wee
n t
he
pre
val
ence
of
tin
nit
us
and
lev
el
of
hea
rin
g
loss
at
dif
fere
nt
freq
uen
cie
s. F
ou
nd
a c
lear
no
n-l
inea
r ex
po
nen
tial
rela
tio
nsh
ip
bet
ween
pre
val
ence
of
tin
nit
us
and
ex
ten
t
of
hea
rin
g l
oss
at
all
freq
uen
cie
s (0
.5k
Hz
to
8k
Hz)
. T
her
e w
as
no
cle
ar i
nd
epen
den
t
effe
ct
of
age u
po
n t
he p
rev
ale
nce
of
tin
nit
us.
Th
is l
arg
e st
ud
y f
ocu
ssed
on
th
ose
wit
hin
hear
ing
co
nse
rvati
on
pro
gra
mm
es (
cf h
ealt
h s
urv
eill
ance)
and
th
eref
ore
in
vo
lved
a r
ang
e o
f
hea
rin
g l
evel
s.
Stu
dy
cle
arly
dem
on
stra
ted
a
rela
tio
nsh
ip b
etw
een
an
in
creas
e i
n
pre
val
ence
of
tin
nit
us
and
ex
ten
t o
f
hea
rin
g l
oss
.
Sh
ow
ed t
hat
an
y a
sso
cia
tio
n b
etw
een
age
and
tin
nit
us
was
a c
on
seq
uen
ce o
f
the
rela
tio
nsh
ip b
etw
een
ag
e a
nd
hea
rin
g l
oss
.
24
3.3.2 Are the hearing thresholds measured using audiometry greater in
those with tinnitus as compared to those who do not have tinnitus?
A total of 9 papers (evidence table four) were identified from the review, which compared the
hearing thresholds measured in populations with or without tinnitus [3, 7, 11, 13, 15, 24, 25, 34, 35]
. Eight
of these publications showed that the group hearing thresholds in populations with tinnitus were
higher than the thresholds measured in those without tinnitus [3, 7, 11, 13, 15, 24, 25, 35]
. These studies
have involved sample sizes ranging between 31 and 110,647 individuals. The smallest of these
studies found that there was an increase in the hearing thresholds measured in those with
tinnitus, but that this wasn’t statistically significant, possibly because of the small sample size [35]
. One further study, did not contain any statistical analysis, although it was conducted with a
large sample size (6,804) [3]
. This study reported that the mean hearing threshold in those
without tinnitus was 15dB as compared to 25dB in those with tinnitus. This difference was
even greater in those with self-reported hearing impairment, with hearing thresholds in those
without and with tinnitus of 33dB and 53dB respectively. The remaining 6 studies that have
shown a positive relationship between the hearing thresholds and tinnitus have shown
statistically significant findings and have involved reasonable sample sizes [7, 11, 13, 15, 24, 25]
. The
study published by Phoon has also shown that the differences in hearing threshold between
workers with or without tinnitus remain even when confounding factors such as gender, age,
race and noise duration are taken into account [25]
. Two studies have involved statistical
analysis to investigate which factors are the most important in determining hearing loss, as
measured by audiometry [11, 13]
. Both of these studies showed that tinnitus was an important risk
factor for the development of hearing loss [11, 13]
.
Only one study was found that did not support the evidence that hearing thresholds are higher in
those with tinnitus [34]
. In this study the total area under the curve of the audiogram was
calculated as the total hearing loss. This was a small study focusing on compensation claimants
with moderate to severe hearing loss. The overall hearing loss in the group with tinnitus was
significantly lower than in the group who did not have tinnitus. The authors acknowledge that
these findings contrast with the rest of the literature, but suggest that this could be explained by
differing criteria use to define whether tinnitus is present.
25
Ev
iden
ce t
ab
le f
ou
r – H
eari
ng
th
resh
old
s fr
om
au
dio
met
ric m
eas
ure
men
t w
ith
or
wit
ho
ut
tin
nit
us
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Mu
luk
, N
[35
] 2
00
8
Wo
rker
s in
a s
teel
fact
ory
ex
po
sed
to
occ
up
ati
on
al
no
ise.
16
male
wo
rker
s w
ith
tin
nit
us
and
15
ag
e-
mat
ched
wo
rker
s
wit
ho
ut
tin
nit
us.
No
ise l
evels
var
ied
bet
wee
n
73
an
d 1
10
dB
.
Mea
n a
ge
for
tho
se w
ith
tin
nit
us
was
42
.2 (
ran
ge
33
-53
) y
ears
.
Fo
r th
e g
rou
p
wit
ho
ut
tin
nit
us
it w
as
42
(ra
ng
e 3
4-
53
) y
ears
Hea
rin
g t
hre
sho
lds
wer
e g
reate
r in
th
ose
wit
h t
inn
itu
s at
all
freq
uen
cie
s.
Th
ese
dif
fere
nces
wer
e n
ot
stat
isti
call
y
sig
nif
ican
t.
Wh
ilst
th
ere
was
a s
ug
gest
ion
th
at
tho
se w
ith
tin
nit
us
had
gre
ate
r h
ear
ing
thre
sho
lds,
th
ese
dif
fere
nces
wer
e n
ot
stati
stic
ally
sig
nif
ican
t.
Th
is m
ay b
e
bec
ause
of
the s
mal
l n
um
ber
s in
vo
lved
in t
he
stu
dy
.
Ko
nig
, O
[34
] 2
00
6
71
co
mp
ensa
tio
n
clai
man
ts w
ith
mo
der
ate
to s
ever
e
wo
rk-r
elat
ed n
ois
e-
ind
uce
d h
eari
ng
lo
ss
in G
erm
any
.
30
wit
ho
ut
tin
nit
us,
24
wit
h t
on
e-li
ke
tin
nit
us
and
17
wit
h
no
ise-
lik
e t
inn
itu
s.
No
t re
po
rted
M
ean a
ge
56
(ran
ge
38
-69
)
yea
rs
Calc
ula
ted o
ver
all
hea
rin
g l
oss
as
the
area
un
der
th
e c
urv
e o
f th
e au
dio
gra
m.
Fo
un
d
that
th
e o
ver
all
hea
rin
g l
oss
in
th
ose
wit
h
tin
nit
us
was
sig
nif
ican
tly
lo
wer
(p
=0
.00
06
)
than
in
th
ose
wh
o d
id n
ot
hav
e t
inn
itu
s.
Wh
en t
he
aud
iog
ram
s w
ere
com
par
ed i
t
was
fo
un
d t
hat
au
dio
gra
ms
wer
e s
imil
ar
bet
ween
th
ose
wit
h o
r w
ith
ou
t ti
nn
itu
s at
hig
h a
nd
lo
w f
req
uen
cies
, b
ut
dif
fere
d o
ver
the m
id-r
ang
e. T
he
gro
up
s w
ith
ou
t ti
nn
itu
s
sho
wed
th
e h
igh
est
hear
ing
th
resh
old
s.
Th
is s
tud
y i
s sm
all
co
mp
ared
to
so
me
of
the
oth
ers
bu
t d
oes
rep
ort
stati
stic
ally
sig
nif
ican
t fi
nd
ing
s.
Th
e au
tho
rs a
ckn
ow
led
ge
that
their
resu
lts
are
dif
fere
nt
to o
ther
stu
die
s,
and
su
gg
est
that
th
is m
ay b
e th
e r
esu
lt
of
dif
fere
nt
cho
ices
of
sub
jects
an
d
dif
fere
nt
incl
usi
on
cri
teri
a re
late
d t
o
tin
nit
us.
In
par
ticu
lar,
th
is s
tud
y o
nly
incl
ud
ed t
ho
se w
ith
ch
ron
ic t
inn
itu
s
rath
er t
han
th
ose
wit
h i
nte
rmit
ten
t
pro
ble
ms.
26
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ch
en,
J-D
[24
] 2
00
3
38
4 n
ois
e-ex
po
sed
wo
rker
s in
an
oil
refi
ner
y
Ex
po
sed
to
aver
age n
ois
e
lev
el
of
73
-89
dB
(A)
Mea
n d
ura
tio
n
of
exp
osu
re w
as
14
(S
D 7
) y
ear
s
Mea
n a
ge
39
(SD
5)
yea
rs
Th
ose
rep
ort
ing
tin
nit
us
had
sig
nif
ican
tly
gre
ater
(p
<0
.05
) h
eari
ng
th
resh
old
s in
bo
th
the
low
an
d h
igh
fre
qu
ency
ran
ges
. I
n t
he
low
fre
qu
ency
ran
ge t
he
mean
(S
D)
hea
rin
g t
hre
sho
ld w
as
18
(11
) d
B i
n t
ho
se
rep
ort
ing
tin
nit
us
and
15
(8)
dB
in
th
ose
wit
ho
ut
tin
nit
us.
In
th
e h
igh
fre
qu
ency
ran
ge
the
corr
esp
on
din
g v
alu
es w
ere
35
(21
) d
B a
nd
24
(14
) d
B.
Th
is s
tud
y s
ug
gest
s th
at t
ho
se w
ith
tin
nit
us
foll
ow
ing
no
ise e
xp
osu
re h
ave
gre
ater
hear
ing
th
resh
old
s
Pra
sher
, D
[15
] 2
00
1
19
8 m
ill
wo
rker
s
exp
ose
d t
o n
ois
e
85
-95
dB
(A)
for
a m
ean
of
12
yea
rs
Mea
n a
ge
(SD
)
for
the
gro
up
wit
h t
inn
itu
s
47
(9
) y
ears
and
39
(9
)
yea
rs f
or
the
gro
up
wit
ho
ut
tin
nit
us
Pu
re t
on
e av
erag
e h
ear
ing t
hre
sho
ld a
cro
ss
all
freq
uen
cies
(1
25
Hz, 2
50
Hz, 5
00
Hz,
1k
Hz,
2k
Hz, 4
kH
z, 6
kH
z, 8
kH
z) w
as
mea
sure
d.
Als
o m
easu
red
tra
nsi
ent
evo
ked
oto
aco
ust
ic e
mis
sio
ns.
T
ho
se w
ith
tin
nit
us
had
a s
ign
ific
antl
y r
aise
d h
ear
ing
th
resh
old
com
par
ed t
o t
ho
se w
ith
ou
t ti
nn
itu
s. O
n
aver
age t
he
hea
rin
g t
hre
sho
ld w
as a
rou
nd
15
dB
hig
her
in
th
e t
inn
itu
s g
rou
p.
Th
ey
also
had
sig
nif
ican
tly
red
uced
tra
nsi
ent
oto
aco
ust
ic e
mis
sio
ns
of
an a
ver
age v
alu
e
of
2.6
dB
.
Th
is s
ug
ges
ts t
hat
ind
ivid
uals
wit
h
tin
nit
us
hav
e h
igh
er h
eari
ng
th
resh
old
s
than
th
ose
wit
ho
ut
tin
nit
us.
27
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ph
oo
n, W
H [2
5]
19
93
6
47
no
ise e
xp
ose
d
wo
rker
s n
oti
fied
as
case
s o
f n
ois
e
ind
uce
d d
eafn
ess
Mea
n (
SD
)
du
rati
on
of
exp
osu
re w
as
17
0 (
10
0.7
)
mo
nth
s in
th
ose
wit
h t
inn
itu
s
and
16
0 (
10
6)
mo
nth
s in
th
ose
wit
ho
ut
tin
nit
us
Mea
n a
ge
39
yea
rs
Hea
rin
g t
hre
sho
lds
at h
igh
an
d l
ow
freq
uen
cies
wer
e s
ign
ific
antl
y h
igh
er i
n
tho
se w
ith
tin
nit
us.
T
hes
e d
iffe
ren
ces
rem
ain
ed e
ven
wh
en a
dju
sted
fo
r se
x,
age,
race
an
d n
ois
e d
ura
tio
n.
Th
is i
s a f
air
ly l
arg
e s
tud
y i
nv
olv
ing
a
ran
ge
of
hear
ing
lo
ss i
n a
fai
rly
yo
un
g
po
pu
lati
on
.
It d
oes
seem
to
su
gg
est
that
tin
nit
us
is
rela
ted
to
hea
rin
g l
oss
, h
ow
ever
it
did
no
t co
nta
in a
gro
up
th
at
did
no
t h
ave
hea
rin
g l
oss
.
Neu
ber
ger
, M
[13
] 1
99
2
11
0,6
47
no
ise
exp
ose
d w
ork
ers
.
>8
5 d
B
Ag
ed b
etw
een
15
-65
yea
rs
Fo
un
d t
hat
in
div
idu
als
wit
h e
ar d
iseas
e,
hea
d i
nju
ry o
r ti
nn
itu
s h
ad h
igh
er h
ear
ing
thre
sho
lds
on
au
dio
metr
y.
Tin
nit
us
app
eare
d t
o b
e t
he m
ost
im
po
rtan
t
sym
pto
m i
n p
red
icti
ng
th
e in
creas
ed
hea
rin
g t
hre
sho
lds
in i
nd
ivid
uals
ex
po
sed
to n
ois
e l
evel
s o
f 8
5-1
05
dB
an
d t
ho
se
exp
ose
d b
etw
een
10
5-1
25
dB
.
Th
is s
tud
y s
ug
gest
s th
at t
inn
itu
s is
an
imp
ort
ant
facto
r in
no
ise i
nd
uce
d
hea
rin
g l
oss
.
Th
e p
op
ula
tio
n i
n t
his
stu
dy
was
no
t
pre
-sel
ecte
d t
o i
nclu
de
on
ly t
ho
se w
ith
hea
rin
g l
oss
. It
in
clu
ded
all
of
tho
se
exp
ose
d t
o n
ois
e o
ver
85
dB
fo
r a
min
imu
m o
f 6
mo
nth
s.
28
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Bau
er, P
[11
] 1
99
1
47
,38
8 n
ois
e-ex
po
sed
wo
rker
s
Med
ian
89
.8
dB
(A).
M
edia
n
len
gth
of
exp
osu
re w
as
11
.8 (
ran
ge
0.5
-
45
) y
ears
Med
ian
ag
e
36
.1 y
ears
Un
der
too
k w
eig
hte
d r
egre
ssio
n a
naly
sis
to
inv
esti
gate
ris
k f
act
ors
fo
r h
eari
ng
lo
ss a
t
dif
fere
nt
freq
uen
cies
0.5
to
6 k
Hz.
Big
ges
t fa
cto
r re
late
d t
o h
ear
ing
th
resh
old
s
was
ag
e. T
inn
itu
s w
as r
elat
ed t
o h
ear
ing
loss
acr
oss
th
e fr
equ
enci
es b
ut
was
fai
rly
con
stan
t ac
ross
th
e fr
equ
enci
es.
Rep
ort
ed t
hat
mo
st i
mp
ort
ant
freq
uen
cy
rela
ted
to
no
ise e
mis
sio
n i
s 4
kH
z
Th
is w
as a
ver
y l
arg
e s
tud
y a
nd
con
tain
ed s
om
e f
air
ly d
etai
led
mat
hem
atic
s.
Th
e st
ud
y s
ho
ws
that
tin
nit
us
is r
ela
ted
to n
ois
e i
nd
uce
d h
eari
ng
lo
ss.
Kam
al,
AA
[7]
19
89
8
8 d
rop
-fo
rge
wo
rker
s
Ex
po
sed
to
imp
act
no
ise
ran
gin
g b
etw
een
11
2 t
o 1
39
dB
(A)
at
a ra
te
bet
ween
20
-50
dro
ps
per
min
ute
Betw
een
30
-60
yea
rs o
f ag
e
Fo
un
d t
hat
wo
rker
s re
po
rtin
g t
inn
itu
s h
ad
the
hig
hes
t h
ear
ing
th
resh
old
s at
all
freq
uen
cies
test
ed (
p<
0.0
5).
T
he
hig
hes
t
hea
rin
g t
hre
sho
ld w
as
mea
sure
d a
t 6
kH
z.
Rep
ort
ed ‘
dif
ficu
ltie
s in
hear
ing
th
e h
um
an
vo
ice’
was
sig
nif
ican
tly a
sso
ciat
ed w
ith
rep
ort
ing
of
tin
nit
us
(p<
0.0
01
).
Th
is s
tud
y s
ho
ws
that
th
ere i
s a
rela
tio
nsh
ip b
etw
een
tin
nit
us
and
hea
rin
g l
oss
fo
llo
win
g e
xp
osu
re to
imp
uls
e n
ois
e.
It i
s al
so a
pp
aren
t th
at
this
was
hav
ing
an i
mp
act
on
qu
alit
y o
f li
fe a
s ti
nn
itu
s
was
ass
ocia
ted
wit
h d
iffi
cult
y i
n
hea
rin
g t
he h
um
an v
oic
e.
29
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Co
les,
RR
A [3
] 1
98
1
6,8
04
in
div
idu
als
fro
m t
he g
ener
al
po
pu
lati
on
in
fo
ur
citi
es o
f th
e U
K
32
.3%
rep
ort
ed
no
ise
exp
osu
re
38
% <
40
yea
rs
32
.8%
40
-60
yea
rs
29
.2%
>6
0
yea
rs
Fo
un
d t
hat
th
e m
ean
hear
ing
th
resh
old
was
gre
ater
wh
en t
inn
itu
s w
as p
rese
nt.
In
th
e
gro
up
wit
h n
orm
al h
eari
ng
an
d n
o t
inn
itu
s,
the
mean
hear
ing
th
resh
old
was
15
dB
com
par
ed t
o 2
5 d
B i
n t
he g
rou
p w
ith
no
rmal
hea
rin
g a
nd
tin
nit
us.
Th
e g
rou
p
wit
h s
elf
-rep
ort
ed h
ear
ing
im
pair
men
t an
d
no
tin
nit
us
had
a m
ean
th
resh
old
of
33
dB
wit
h t
ho
se w
ith
hea
rin
g i
mp
air
men
t an
d
tin
nit
us
hav
ing
th
e g
reate
st t
hre
sho
ld o
f 5
3
dB
. I
t is
un
cle
ar w
heth
er t
hese
fin
din
gs
are
stati
stic
ally
sig
nif
ican
t as
th
ere a
re n
o
stati
stic
s p
rese
nte
d i
n t
his
pap
er.
Th
is s
tud
y w
as p
ub
lish
ed a
s a p
ilo
t
stu
dy
fo
r a l
arg
er m
ain
stu
dy
.
Th
is s
tud
y s
ho
ws
that
th
ose
wh
o h
ave
tin
nit
us
hav
e g
reate
r h
eari
ng
thre
sho
lds
and
th
us
mo
re h
ear
ing
lo
ss
than
th
ose
wit
ho
ut
tin
nit
us.
30
3.3.3 Is there evidence of an increased risk of tinnitus in those with hearing
loss, or an increased risk of hearing loss in those with tinnitus?
Only three papers (Evidence table five) were identified which reported odds ratios or relative
risks linking hearing loss and tinnitus [24, 31, 36]
. Two of these studies reported the risks of
tinnitus in those with hearing loss [31, 36]
. Sindhusake [31]
reported a relative risk of 1.12 (95%
confidence interval of 1.06-1.18) of tinnitus in those with hearing loss. This was statistically
significant but only represents a small increase in risk. Another study reported that the odds
ratio of tinnitus occurring in those with deafness was 1.11 (95% confidence interval 0.5-2.44)
and not statistically significant [36]
. However, a further study investigating the risk of
developing hearing loss in noise-exposed workers reported that the presence of tinnitus
increased the risk of developing hearing loss by 2-3 times [24]
.
3.3.4 What is the temporal relationship between tinnitus and hearing loss?
Only one study was identified which had sought to establish the temporal relationship between
tinnitus and hearing loss [22]
. This was a 15-year longitudinal study which involved workers in
environments with average noise levels ranging between 85 and 101 dB(A) over the period of
the study. Of the 91 workers involved in the study 27 reported tinnitus. Twenty of the
individuals who reported tinnitus also had noise-induced hearing loss (as defined by a hearing
threshold � 15dB at 4kHz) and 90% of these had reported tinnitus prior to this change in hearing
threshold. Although this study contains relatively small numbers of individuals reporting
tinnitus the association between time of reporting of tinnitus and the time when hearing loss
occurred was statistically significant (p=0.03). On average tinnitus occurred 5.8 years prior to
the maximum hearing loss, with a range of 1-16 years.
31
Ev
iden
ce t
ab
le f
ive –
Stu
die
s p
rese
nti
ng
ris
k f
act
ors
or
od
ds
rati
os
rela
ted
to
hear
ing
lo
ss a
nd
tin
nit
us
Au
tho
r
Yea
r
Ty
pe o
f su
bje
cts
No
ise e
xp
osu
re
Ag
e
Fin
din
gs
Co
mm
ents
Ch
en,
J-D
[24
] 2
00
3
38
4 n
ois
e-ex
po
sed
wo
rker
s in
an
oil
refi
ner
y
Ex
po
sed
to
aver
age n
ois
e
lev
el
of
73
-89
dB
(A).
M
ean
du
rati
on
of
exp
osu
re w
as
14
(SD
7)
yea
rs
Mea
n a
ge
39
(SD
5)
yea
rs
Th
e o
dd
s ra
tio
fo
r d
evel
op
ing
hea
rin
g l
oss
in t
he
low
fre
qu
ency
ran
ge
in t
ho
se w
ith
tin
nit
us
was
2.6
(9
5%
co
nfi
den
ce
inte
rval
1.3
-5.5
) an
d i
n t
he
hig
h f
req
uen
cy r
ang
e
was
3.0
(9
5%
CI
1.9
-4.9
).
Th
is s
tud
y s
ug
gest
s th
at t
he
risk
of
dev
elo
pin
g h
eari
ng
lo
ss i
f ti
nn
itu
s is
pre
sen
t is
2-3
tim
es g
reate
r th
an i
f
tin
nit
us
is n
ot
pre
sen
t.
Ho
wev
er,
oth
er f
acto
rs s
uch
as
self
-
rep
ort
ing
of
hea
rin
g l
oss
gav
e m
uch
hig
her
od
ds
rati
os.
Sin
dh
usa
ke, D
[31
] 2
00
3
2,0
15
in
div
idu
als
fro
m t
he g
ener
al
po
pu
lati
on
in
Sy
dn
ey, A
ust
rali
a
Gro
up
ed l
evel
s
no
t g
iven
Mea
n a
ge
of
69
.8 y
ears
Wh
en a
ge
and
gen
der
wer
e t
aken
in
to
acco
un
t h
ear
ing
lo
ss (
10
dB
) at
0.5
, 1
, 2
or
4 k
Hz
was
fo
un
d t
o b
e a
sig
nif
ican
t ri
sk
fact
or
for
tin
nit
us
(rela
tiv
e ri
sk o
f 1
.12
–
95
% c
on
fid
ence
inte
rval
1.0
6-1
.18
).
Pro
du
ced
a p
red
icti
ve m
od
el
for
tin
nit
us,
wh
ich
in
clu
ded
deg
ree o
f h
eari
ng
lo
ss.
Ho
wev
er,
hea
rin
g l
oss
on
ly h
ad a
mo
des
t
asso
cia
tio
n w
ith t
inn
itu
s –
11
% i
ncr
ease
d
lik
eli
ho
od
of
rep
ort
ing
tin
nit
us
wit
h e
ach
10
dB
in
creas
e in
hear
ing
th
resh
old
.
Stu
dy
sh
ow
s th
at
in t
his
old
er a
ge g
rou
p
hea
rin
g l
oss
is
a p
red
icti
ve f
act
or
for
rep
ort
ing
of
tin
nit
us.
H
ow
ever
, th
e ri
sk
was
qu
ite
small
an
d o
ther
fac
tors
(e.g
.
no
ise
exp
osu
re)
wer
e m
ore
im
po
rtan
t.
McB
rid
e, D
[36
] 1
99
3
18
9 w
ork
ers
in a
min
ing
co
llie
ry
Mea
n d
ura
tio
n o
f
exp
osu
re 9
.8
yea
rs
No
t re
po
rted
D
eafn
ess
was
det
erm
ined
by
au
dio
met
ry a
s
ou
tlin
ed b
y t
he
HS
E d
ocu
men
t
‘Au
dio
met
ry i
n I
nd
ust
ry’.
T
he o
dd
s ra
tio
of
tin
nit
us
occ
urr
ing
in t
ho
se w
ith
dea
fnes
s
was
calc
ula
ted
as
1.1
1 (
95
% c
on
fid
ence
inte
rval
of
0.5
-2.4
4).
Th
is s
tud
y d
id n
ot
sho
w a
n o
bv
iou
s
incr
eas
ed r
isk
of
tin
nit
us
in t
ho
se w
ith
hea
rin
g l
oss
.
32
4
4.1
MAIN FINDINGS AND DISCUSSION
The main issues addressed by this review are:
1. Is there a relationship between exposure to noise at work and tinnitus?
2. Is there a relationship between tinnitus and hearing loss in those exposed to noise at
work?
In order to address these questions an extensive search of 12 different databases, from as early
as 1951, was conducted to identify publications relevant to this review. However, even
following this extensive search only 34 publications were identified that were written in the
English language and were relevant to the questions posed by this review. Therefore, there does
not appear to be a vast body of knowledge available investigating these issues. If the scope of
the search had been widened to include noise in general, rather than specifically occupational
noise exposure, then it is possible that more publications would have been identified. However,
it was felt that the focus of this review should be on occupational noise rather than noise in
general. The 34 publications identified have been reviewed in detail and form the basis of the
main findings in this report.
IS THERE A RELATIONSHIP BETWEEN EXPOSURE TO NOISE AT
WORK AND TINNITUS?
In order to address this question the information in the published literature could be split into
references that provided information regarding the prevalence of tinnitus in populations exposed
to noise at work, and references investigating the relationship between tinnitus and the severity
of noise exposure. Ideally to answer the question of whether occupational noise exposure leads
to tinnitus one would also need information on a control group, who are not exposed to
occupational noise, to act as a comparison. However, only four studies were identified that also
contained information regarding the prevalence of tinnitus in a non-noise exposed population [3-
6] . The relative prevalence reported for tinnitus in noise-exposed populations compared to
controls were 70.4% vs 3.5%, 32.2% vs 6.7%, 24% vs 14.4% and 12% vs 2% respectively.
Therefore, these four studies would suggest that the prevalence of tinnitus in workers exposed to
noise at work is higher than in control non-noise exposed populations.
Overall, in all of the studies that reported some prevalence of tinnitus in populations exposed to
occupational noise the prevalence was widely variable (between 87.5% and 5.9%). Factors such
as the type of participant (e.g. compensation claimants vs health surveillance), the type of noise
an individual is exposed to (e.g. impulsive vs continuous) and the definition of tinnitus used (i.e.
whether transient tinnitus is excluded) may all contribute to the prevalence measured in any
given population. In addition, if noise-induced hearing loss is related to tinnitus, then the level
of noise-induced hearing loss within the studied population will in turn affect the measured
prevalence of tinnitus. In general, it does appear that the prevalence of noise-induced hearing
loss (where reported) was higher in those studies with the highest prevalence of tinnitus.
Interestingly, three studies that had particularly large sample sizes of noise-exposed workers
(110,647, 30,000 and 47,388 respectively) had the lowest reported prevalence of tinnitus of
6.7%, 6.6% and 5.9% respectively. Unfortunately, two of the studies did not report the
prevalence of noise-induced hearing loss in the studied population and one reported that around
10% had some impairment.
33
4.2
There have been a few studies that have reported the prevalence of tinnitus within the general
population [37-40]
. These studies suggest a range of prevalence between 14.2% and 33%. In the
literature detailed in evidence table one 11 of the studies reported a prevalence of tinnitus
greater than 33% and 10 below.
Some studies have aimed to look at the relationship between tinnitus and the severity of
exposure to noise, rather than simply whether they have been exposed to noise at work or not.
These studies have either quantified noise exposure using actual noise measurements [26-29]
or
estimated exposure based upon the total number of years working in a noisy environment [6, 17]
,
and difficulty in hearing or speaking in that environment [30, 31]
. There was only a small amount
of literature published in this area (9 papers) but the majority of these studies show some
relationship between severity of exposure and tinnitus [6, 17, 26, 27, 29-32]
. Rubak [28]
published a
study involving 752 workers that involved the use of actual cumulative noise exposure derived
from personal dosimeters. They found that tinnitus was only related to noise exposure if there
was also hearing loss present (defined from audiometry). However, the numbers with tinnitus
were small and the confidence intervals wide, which may have affected the outcome. Another
study involving 180 workers from the cement industry found that the prevalence of tinnitus had
a tendency to increase with the actual noise level but these differences were not statistically
significant [29]
. Two large studies conducted with the general population have shown
relationships between working in a noisy job and tinnitus [30, 31]
. Sindhusake et al conducted a
study in 2,015 individuals over the age of 55 years in Sydney [31]
. They constructed a predictive
model for tinnitus taking account of age, gender and noise-induced hearing loss, and found that
the relative risk of tinnitus increased with both the number of years of noise exposure and the
severity in terms of the self-reported tolerability of the noise level. Whilst this study was
conducted in an older population, a study carried out in the UK general population found similar
findings in a population aged between 35 and 64 years [30]
.
Overall, the literature related to the relationship between exposure to noise at work and the
prevalence of tinnitus suggests that the prevalence of tinnitus is greater in workers exposed to
noise and that it is greater in those workers exposed for greater durations or to greater levels.
However, the relative contribution of confounding factors, including the definition of tinnitus
used and the prevalence of noise-induced hearing loss, are difficult to untangle from the
published literature.
IS THERE A RELATIONSHIP BETWEEN TINNITUS AND HEARING
LOSS IN THOSE EXPOSED TO NOISE AT WORK?
The literature identified in this area tended to address this question in two different ways:
1. It compared the prevalence of tinnitus in those with or without noise-induced hearing
loss, or different severities of hearing loss and;
2. It compared the hearing thresholds in those with or without self-reported tinnitus.
Overall, nine papers were identified which addressed the first of these issues. Of these papers
the majority (n=7) appeared to show a positive relationship between the prevalence of tinnitus
and hearing loss [12, 19, 22, 25, 26, 30, 33]
. However, whilst two of these studies do appear to report
data to support this association they did not contain any statistical analysis, so the statistical
validity of their findings is unclear [19, 33]
. The remaining studies contain populations ranging
between 91 and 30,000 workers. The largest of these studies was conducted within a hearing
conservation programme (similar to health surveillance) and involved workers exposed to noise
at a mean level of 85dB(A) or more[12]
. This study reported that there was a non-linear
34
exponential relationship between the prevalence of tinnitus and the extent of hearing loss at
frequencies between 0.5 and 8kHz.
Two studies did not report an association between the prevalence of tinnitus and severity of
hearing loss [8, 17]
. These were large studies that contained workers with significant noise
exposure, and thus it is unclear why these studies did not find a positive relationship. However,
one possible consideration is that these were the only studies that involved claimants for noise-
induced hearing loss, and the impact of this on the self-reporting of tinnitus is unclear.
Nine papers were identified which investigated the impact of tinnitus on the actual hearing
thresholds as measured by audiometry [3, 7, 11, 13, 15, 24, 25, 34, 35]
. Six of these publications
demonstrated a significant relationship between hearing thresholds and tinnitus [3, 7, 11, 13, 15, 24, 25]
.
Some of these studies have conducted group comparisons of the hearing thresholds in
individuals with or without tinnitus, and demonstrated that the hearing thresholds in those with
tinnitus are significantly higher [3, 7, 15, 24, 25]
. Other studies have investigated tinnitus as a factor
predicting hearing thresholds in regression analysis, and found that it is an important risk factor [11, 13]
tinnitus
. The studies investigating the relationship between tinnitus and hearing thresholds found
that the measured hearing thresholds at all frequencies (0.5 to 8kHz) are increased in those with [7, 11, 15, 24, 25]
. These differences appear to remain even when confounding factors such as
sex, age, race and noise duration are taken into account [25]
. These studies consisted of a range
of sample sizes from relatively small (n=88) [7]
to very large (n=110,647) [13]
, and have involved
noise-exposed workers, cases of noise induced deafness and the general population. Three of
the published studies did not report an association between hearing thresholds and tinnitus. One
of these studies reported that the hearing thresholds were higher in those with tinnitus.
However, this was not statistically significant [35]
. This was a very small study involving 16
workers with tinnitus and 15 without tinnitus, and thus may not have had the power to detect a
significant effect. Another study reported elevated values in tinnitus, but did not contain any
statistical analysis [3]
. The third study involved 71 compensation claimants with noise-induced
hearing loss and found that the overall hearing loss, as assessed by measuring the area under the
curve of the audiogram, in those with tinnitus was significantly lower than in those who did not
have tinnitus [34]
. The authors acknowledge that these results are contrary to much of the other
published literature and possibly factors such as type of subjects and the definition of tinnitus
used may have contributed to this difference.
Other studies have tried to establish the risk factors for the development of hearing loss or the
occurrence of tinnitus [24, 31, 36]
. One study reported that the odds ratio for developing hearing
loss in those with tinnitus was between 2-3 times greater than if tinnitus was not present [24]
,
whereas another study suggested that there was no increased risk of tinnitus occurring in those
with deafness [36]
. However, it does appear that reporting of hearing loss is a modest predictive
factor for the occurrence of tinnitus [31]
.
Overall, the weight of the evidence that is available in the literature supports the hypothesis that
there is a link between hearing loss and tinnitus. However, it is far from clear whether hearing
loss and tinnitus occur as independent effects of noise exposure, or whether one is causally
related to the other. The majority of the studies published are cross-sectional in nature and thus
do not help to give information on the temporal relationship between these two health outcomes.
However, there is one longitudinal study published that helps to provide some information on
this issue[22]
. This study was a 15-year longitudinal follow-up study that involved 91 noise-
exposed workers. It was found that 90% of workers with both tinnitus and hearing loss had
tinnitus on average 5.8 years before hearing loss took place. Whilst this does not necessarily
show that tinnitus is causally related to hearing loss, it does suggest that where they both occur,
tinnitus does tend to occur earlier and may be an earlier indicator of potential problems.
However, the fact that the prevalence of tinnitus appears to be related not only to whether
35
hearing loss occurs, but also its severity, it would suggest that there may be some causal link.
To fully understand this more research is required in this area.
36
5 REFERENCES
1. LaMarte, F. and R. Tyler, Noise-induced tinnitus. AAOHN journal, 1987. 35(9): p. 403-
406.
2. Axelsson, A. and D. Prasher, Tinnitus induced by occupational and leisure noise. Noise
& health, 2000. 8: p. 47-54.
3. Epidemiology of tinnitus, Medical Research Council's Institute of Hearing Research.
Ciba Foundation symposium, 1981. 85: p. 16-34.
4. Chavalitsakulchai, P., et al., Noise Exposure and Permanent Hearing Loss of Textile
Workers in Thailand. Industrial Health, 1989. 27(4): p. 165-173.
5. Hong, O., S.P. Chen, and K. Conrad, Noise Induced Hearing Loss among Male Airport
Workers in Korea. AAOHN journal, 1998. 46(2): p. 67-75.
6. Sulkowski, W., et al., Tinnitus and impulse noise-induced hearing loss in drop-forge
operators. International Journal of Occupational Medicine and Environmental Health,
1999. 12(2): p. 177-182.
7. Kamal, A.A., R. Mikael, and R. Faris, Follow-Up of Hearing Thresholds among Forge
Hammering Workers. American Journal of Industrial Medicine, 1989. 16(6): p. 645-
658.
8. Alberti, P., Tinnitus in Occupational Hearing Loss: Nosological Aspects. Journal of
Otolaryngology, 1987. 16(1): p. 34-35.
9. Barrs, D.M., et al., WORK-RELATED, NOISE-INDUCED HEARING-LOSS -
EVALUATION INCLUDING EVOKED-POTENTIAL AUDIOMETRY. Otolaryngology-
Head and Neck Surgery, 1994. 110(2): p. 177-184.
10. Mrena, R., et al., Occupational noise-induced hearing loss reports and tinnitus in
Finland. Acta Oto-Laryngologica, 2007. 127(7): p. 729-735.
11. Bauer, P., et al., RISK-FACTORS FOR HEARING-LOSS AT DIFFERENT
FREQUENCIES IN A POPULATION OF 47388 NOISE-EXPOSED WORKERS.
Journal of the Acoustical Society of America, 1991. 90(6): p. 3086-3098.
12. Chung, D.Y., R.P. Gannon, and K. Mason, FACTORS AFFECTING THE
PREVALENCE OF TINNITUS. Audiology, 1984. 23(5): p. 441-452.
13. Neuberger, M., et al., Hearing Loss from Industrial Noise, Head Injury and Ear
Disease. A Multivariate Analysis on Audiometric Examinations of 110647 Workers.
Audiology, 1992. 31(1): p. 45-57.
14. Bray, A., et al., Noise induced hearing loss in dance music disc jockeys and an
examination of sound levels in nightclubs. Journal of Laryngology and Otology, 2004.
118(2): p. 123-128.
15. Prasher, D. and B. Ceranic, Objective Evidence for Tinnitus from Spontaneous Emission
Variability. Noise and Health, 2001. 3(12): p. 61-73.
37
16. Jansen, E.J., et al., Noise induced hearing loss and other hearing complaints among
musicians of symphony orchestras. Int Arch Occup Environ Health, 2009. 82(2): p.
153-64.
17. McShane, D., M. Hyde, and P. Alberti, Tinnitus Prevalence in Industrial Hearing Loss
Compensation Claimants. Clinical Otolaryngology, 1988. 13(5): p. 323-330.
18. Kahari, K., et al., Assessment of hearing and hearing disorders in rock/jazz musicians.
International Journal of Audiology, 2003. 42(5): p. 279-288.
19. Ylikoski, M.E. and J.S. Ylikoski, HEARING-LOSS AND HANDICAP OF
PROFESSIONAL SOLDIERS EXPOSED TO GUNFIRE NOISE. Scandinavian Journal
of Work Environment & Health, 1994. 20(2): p. 93-100.
20. Lindgren, T., et al., Tinnitus among airline pilots: prevalence and effects of age, flight
experience, and other noise. Aviat Space Environ Med, 2009. 80(2): p. 112-6.
21. Hong, O., Hearing loss among operating engineers in American construction industry.
International archives of occupational and environmental health, 2006. 78(7): p. 565-
574.
22. Griest, S. and P. Bishop, Tinnitus as an Early Indicator of Permanent Hearing Loss : a
Fifteen Year Longitudinal Study of Noise Exposed Workers. AAOHN journal, 1998.
46(7): p. 325-329.
23. Begault, D.R., et al., Survey of commercial airline pilots' hearing loss. Perceptual and
Motor Skills, 1998. 86(1): p. 258-258.
24. Chen, J.D. and J.Y. Tsai, Hearing loss among workers at an oil refinery in Taiwan.
Archives of Environmental Health, 2003. 58(1): p. 55-58.
25. Phoon, W.H., H.S. Lee, and S.E. Chia, TINNITUS IN NOISE-EXPOSED WORKERS.
Occupational Medicine-Oxford, 1993. 43(1): p. 35-38.
26. Dijk, F.V. and A. Souman, Non-Auditory Effects of Noise in Industry : 6, a Final Field
Study in Industry. International archives of occupational and environmental health,
1987. 59(2): p. 133-145.
27. Gunderson, E., J. Moline, and R. Catalano, Risks of Developing Noise-Induced Hearing
Loss in Employees of Urban Music Clubs. American Journal of Industrial Medicine,
1997. 31(1): p. 75-79.
28. Rubak, T., et al., The risk of tinnitus following occupational noise exposure in workers
with hearing loss or normal hearing. International Journal of Audiology, 2008. 47(3):
p. 109-114.
29. Sallustio, V., et al., Auditory dysfunction in occupational noise exposed workers. Scand
Audiol Suppl, 1998. 48: p. 95-110.
30. Palmer, K.T., et al., Occupational exposure to noise and the attributable burden of
hearing difficulties in Great Britain. Occupational and Environmental Medicine, 2002.
59(9): p. 634-639.
31. Sindhusake, D., et al., Risk factors for tinnitus in a population of older adults: The Blue
Mountains Hearing Study. Ear and Hearing, 2003. 24(6): p. 501-507.
38
32. Hagberg, M., G. Thiringer, and L. Brandstroem, Incidence of tinnitus, impaired hearing
and musculoskeletal disorders among students enroled in academic music education - A
retrospective cohort study. International archives of occupational and environmental
health, 2005. 78(7): p. 575-583.
33. Miyakita, T. and H. Miura, A Tentative Proposal For Classification Of Audiograms In
Noise-Induced Deafness: Relationship Between Audiogram And Subjective Complaints
In Noise-Exposed Workers. International archives of occupational and environmental
health, 1986. 57(3): p. 173-183.
34. Koenig, O., et al., Course of hearing loss and occurrence of tinnitus. Hearing Research,
2006. 07: p. 221.
35. Muluk, N.B. and O. Oguzturk, Occupational noise-induced tinnitus: does it affect
workers' quality of life? J Otolaryngol Head Neck Surg, 2008. 37(1): p. 65-71.
36. McBride, D., Hearing Conservation in the Mining Industry. Evaluation of a Risk Factor
Questionnaire. Occupational medicine, 1993. 43(4): p. 185-192.
37. Axelsson, A. and A. Ringdahl, Tinnitus--a study of its prevalence and characteristics.
Br J Audiol, 1989. 23(1): p. 53-62.
38. Hannaford, P.C., et al., The prevalence of ear, nose and throat problems in the
community: Results from a national cross-sectional postal survey in Scotland. Family
Practice, 2005. 22(3): p. 227-233.
39. Heller, A.J., Classification and epidemiology of tinnitus. Otolaryngol Clin North Am,
2003. 36(2): p. 239-48.
40. Sindhusake, D., et al., Factors predicting severity of tinnitus: a population-based
assessment. J Am Acad Audiol, 2004. 15(4): p. 269-80.
39
40
41
Published by the Health and Safety Executive 01/10
Health and Safety Executive
A review of the current state of knowledge on tinnitus in relation to noise exposure and hearing loss This report details the results of a search of the published peer-reviewed literature investigating the relationship between tinnitus (ringing or buzzing in the ears), noise exposure at work and noise-induced hearing loss. A total of 12 citation databases (earliest date 1951) were searched which identified 252 publications, of which 34 were found to be relevant to the review. A number of studies have reported the prevalence of tinnitus in populations exposed to noise at work to be between 87.5% and 5.9%. Factors such as the type of participant (eg health surveillance, compensation claimant), the characteristics of the noise exposure and the definition of tinnitus used may contribute to this variability. Furthermore, four studies have shown that the prevalence of tinnitus in workers exposed to noise at work is significantly greater than in workers not exposed to noise. The majority of the published papers support the idea that there is an association between tinnitus and noise-induced hearing loss. The prevalence of tinnitus in those with hearing loss appears to be greater, and the hearing thresholds in those with tinnitus are higher. There is also a suggestion from one 15-year longitudinal study that tinnitus may be an early indicator of risk of the development of noise-induced hearing loss.
This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.
RR768
www.hse.gov.uk