AMRL-TR-73-91
EPA-550/9-73-001-B
PREDICTION OF NIPTS DUE TO
CONTINUOUS NOISE EXPOSURE
DANIEL L. JOHKSON, MAJOR, USAF
AEROSPACE MEDICAL RESEARCH LABORATORY
JULY 1973
JOINT EPA/USAF STUDY
Approved for public release; distribution unlimited.
AEROSPACE MEDICAL RESEARCH LABORATORY
AEROSPACE MEDICAL DIVISION
AIR FORCE SYSTEMS COMMAND
WRIGHT-PATTERSON AIR FORCE BASE, OHIO
-------
NOTICES
When US Government drawings, specifications, or other data are used for any purpose other than
a definitely related Government procurement operation, the Government thereby incurs no respon-
sibility nor any obligation whatsoever, and the fact that the Government may have formulated,
furnished, or in any way supplied the said drawings, specifications, or other data, is not to be
regarded by implication or otherwise, as in any manner licensing the holder or any other person or
corporation, or conveying any rights or permission to manufacture, use, or sell any patented in-
vention that may in any way be related thereto.
Organizations and individuals receiving announcements or reports via the Aerospace Medical Re-
search Laboratory automatic mailing lists should submit the addressograph plate stamp on the
report envelope or refer to the code number when corresponding about change of address or can-
cellation.
Do not return this copy. Retain or destroy.
Please do not request copies of this report from Aerospace Medical Research Laboratory. Additional
copies may be purchased from:
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22151
This report has been reviewed and cleared for open publication and/or public re-
lease by the appropriate Office of Information (01) in accordance with APR 190-17
and DODD 5230.0. There is no objection to unlimited distribution of this report
to the public at large, or by DDC to the National Technical Information Service
(NTIS).
This technical report has been reviewed and is approved for publication.
FOR THE COMMANDER
-r
Osfi/
HENNINOsfi/ von GIERKE, Dr. Ing. ALVIN F. MEYER, JR.'
Director, Biodynamics and Bionics Division Deputy Assistant Administrator
Aerospace Medical Research Laboratory for Noise Control Program
United States Environmental
Protection Agency
AIR FORCE/56780/14 August 1973 — 2500
-------
DOCUMENT CONTROL DATA - R & D
(Security cl*»iitlcmtion of f/fft>, body of abmtrmct and Indexing annotation must b» antarad when tfta overall report 13 ctufiHted)
I. ORIGINATING ACTIVITY (Cotpotflt author)
Aerospace Medical Research Laboratory, Aerospace
Medical Division, Air Force Systems Command,
Wright-Patterson Air Force Base, Ohio 45433
2«. REPORT SECURITY CLASSIFICATION
UNCLASSIFIED
26. GROUP
N/A
3. REPORT TITLE
PREDICTION OF NIPTS DUE TO CONTINUOUS NOISE EXPOSURE
4.. DESCRIPTIVE NOTES (Typ* ol report and Inclusive detee)
8. AUTHOR(S) (Flral nmme, middle Inltlml. laet name)
Daniel L. Johnson, Major, USAF
«. REPORT DATE
M. CONTRACT OR
b. PROJECT NO.
c. Task No.
* Work Unit
July 1973
GRANT NO.
7231
723103
16
^m. TOTAL NO. OF PAGES 7b. NO. OF REFS
66 16
»a. ORIGINATOR'S REPORT NUMBER(S)
AMRL-TR-73-91
»t>. OTHER REPORT NO (Any other munbere that may be aaatgrtad
thl* report) •
EPA-550/9-73-001-B
10. DISTRIBUTION STATEMENT
Approved for public release; distribution unlimited
II. SUPPLEMENTARY NOTES
12. SPONSORING MILITARY ACTIVITY
Aerospace Medical Research Laboratory,
Aerospace Medical Div, Air Force Systems
Command, Wright-Patterson AFB, OH 45433
II. ABSTRACT
In support of the main document, "A Basis for Limiting Noise Exposure for Hearing
Conservation," this report compares the relationship of noise exposure to Noise
Induced Permanent Threshold Shift (NIPTS) as predicted by the currently available
works of Passchier-Vermeer, Robinson, Baughn and Kryter, and the yet unpublished
work of the National Institute of Occupational Safety and Health. The works of
Passchier-Vermeer, Robinson, and Baughn are selected since these are the only works
that completely predict the relationship between NIPTS and noise exposure for
various audiometric frequencies, sound pressure levels and population percentiles.
The predictions of these three methodologies are averaged in order to provide one
single relationship between continuous noise exposure and NIPTS. This relationship
is presented in various ways so that the effect of noise exposure on hearing can be
viewed in more than one way. Discussion concerning the type of frequency weighting,
the equal energy rule, and long duration exposures is also provided.
DD
FORM
i NOV ee
1473
Security Classification
-------
PREFACE
The Biodynamics and Bionics Division of the Aerospace Medical Re-
search Laboratory was given the responsibility under an Interagency Agree-
ment with the Environmental Protection Agency, to develop a document which
would serve as a basis for limiting noise for purposes of hearing conserva-
tion. The preparation of this document was accomplished by the University
of Dayton Research Institute (UDRI) under Contract F33615-72-C-1402.
The Aerospace Medical Research Laboratory efforts in support of this pro-
ject were included under Project 7231-03-16, "Auditory Responses to Acous-
tical Energy Experienced in Air Force Activities. "
In order to resolve certain issues that developed during preparation of
the primary document, the material of this supporting document was develop-
ed. This document does not cover all facets of the relations between hear-
ing and noise exposure, and should be used only in conjunction with the
primary document "A Basis for Limiting Noise Exposure for Hearing Con-
servation" (AMRL-TR-73-90) (EPA-550/9-73-001-A).
Acknowledgement is made of the assistance provided by Dr. H. E. von
Gierke, Dr. C. W. Nixon and Capt. David Krantz of the Biodynamics and
Bionics Division.
ill
-------
TABLE OF CONTENTS
Page
I INTRODUCTION 1
II RELATION OF NOISE TO HEARING LOSS 1
A. Relation of Noise to Hearing Loss for Constant SPL
for 8 Hour Working Day 1
1. Exposure Situation of Data Base 1
2. Selection of Data Base 2
3. Other Methods 10
4. Simplication of Data 10
5. Details of Selected Methodologies 10
6. Manipulation of Data 19
7. Considerations 32
8. Risk of Noise Relative to Hearing Level Exceeding
a Predetermined Level or Fence 39
9. Percent of the Population with more than 5 dB
NIPTS at 400.0 Hz Versus 8 Hour Noise Exposure
Level 44
10. Selection of Limit for 8 Hour Day 44
11. Criticism of Kryter's Method 44
12, D-Versus A -Weighting of Frequency 50
13. Duration of the Exposure 50
14. Estimation of the Accuracy in Relating NIPTS to Noise
Noise Exposure 55
B. Requirement for "Quiet" 58
in SUMMARY 58
IV CONCLUSIONS 60
REFERENCES 61
iv
-------
PREDICTION OF NIPTS DUE TO CONTINUOUS NOISE EXPOSURE
I. INTRODUCTION
This report was written to support certain parts of the criteria document,
"A Basis for Limiting Noise Exposure for Hearing Conservation". Specifically,
several different predictive methods are presented that estimate the effects
of noise on hearing. The predictive results will then be manipulated until
they are reduced to a format that allows a basis for administratively proposing
a specific noise limit.
This report relies on the main document (AMRL-TR-73-90) for defini-
tion of terms, arguments concerning impulsive noise, relationships between
Temporary Threshold Shifts (TTS) and Noise Induced Permanent Threshold
Shift (NIPTS), etc.
Method of Attack. With respect to NIPTS, the duration, spectrum and
intensity of the noise exposure, the sensitivity of the individual, and the life-
time noise exposure history of the individual are all important parameters.
With this many parameters, it is predictable that there are varied opinions
as to how NIPTS will develop in a group of people exposed to noise. If one
adds to the problem various interpretations of what constitutes a significant
hearing loss, then it is not surprising that a resulting jumble of noise limit-
ing criteria will develop. The intent of this supplement is not to be inter-
pret what constitutes a significant hearing loss until such interpretations are
required in order to suggest a recommended limit. Therefore, major em-
phasis will be placed on the relationship of NIPTS to noise for various popu-
lation percentiles.
H. RELATION OF NOISE TO HEARING LOSS
A. Relation of Noise to Hearing Loss for Constant SPL for 8 Hour
Working Day
1. Exposure Situation of Data Base. This situation is the basis
of much of the human data with respect to actual hearing loss. Therefore it
is this situation that by necessity anchors any criterion which will relate
hearing loss to noise. Once this point is selected, exposure duration is then
handled such that shorter or longer exposures are expected to be as noxious
as the 8 hour exposure. The 8 hour permissible exposure point,' therefore,
must be set with great care. Since this 10 the heart of the report, a consid-
erable amount of detail will be presented that will hopefully allow selection
of permissible noise exposure for an 8 hour day.
-------
2. Selection of Data Base. Various researchers have made an
attempt to develop a predictive relationship between noise exposure in the
8 hour working day and the resulting hearing losses. The relationships were
investigated and either accepted or rejected based on whether or not they
(a) allowed calculation of NIPTS at various percentile points and (b) consider-
ed at least speech frequencies (. 5, 1 and 2 kHz) and the audiometric frequency
of 4 kHz. The methods of Passchier-Vermeer, Robinson and Baughn satisfy
these restrictions.
Passchier-Vermeer1 s method is attractive in that it correlates
the data of many different reports. Inclusion of her method thus provides a
rather broad data base (see Table 1 for a summary of her sources). A weak-
ness of her method is that for much of her data base only the 25, median,
and 75 percentile levels of the population were provided.
Robinson's method provides one mathematical relationship (the
hyperbolic tangent) which is adjusted for the audiometric frequencies con-
sidered and the percentile levels used. The method's strength is that it allows
calculation of predicted NIPTS for a wide variety of conditions. A criticism of
the method might be that it uses only one careful study of an otologically
screened population of British subjects. Such a population may not be typical
of average US population. It is also difficult to visualize how the hyperbolic
tangent could be a best approximation to NIPTS for all frequencies and condi-
tions. Nevertheless, Robinson's methodology is well conceived and provides
an additional .data base.
Baughn's data provides superior insight into how NIPTS develops
at various percentile points, not just the median* It has also been used as
the basis for the ISO standard. Its weakness, as typical with many industrial
studies, is that some residual TTS will have been measured since an occasion
only 20 minutes recovery was allowed before audiometric testing was performed.
Lack of recovery would tend to make the predicted NIPTS too high. A second
problem is that the control (or non-noise exposed group) must be considered
to have been exposed to 78 dBA or less. Therefore from Baughn's data
alone, it would be impossible to show that the 78 dBA exposure was not in
itself causing a significant NIPTS.
In summary, all three methods have both strengths and weak-
nesses and it would be hard to say which of the three methods (Robinson's,
Passchier-Vermeer1 s or Baughn) gives the best estimates of the true situ-
ation. Therefore, the predicted NIPTS values were tabulated for each method
and compared. The results, as seen in Table 2, speak for themselves. In
general, there are not large (greater than 10 dB) differences between the
three methods. Most differences are less than 5 dB. For this reason, all
three methods were used to derive predicted values of NIPTS. The final
prediction is the average of the NIPTS of each method; and, as a consequence,
should give a final result that is not unduly influenced by the weakness of any
single method.
-------
TABLE 1
Work Included In Passchier-Vermeer's (1968) Analysis
W. Burns, R. Hinchcliffe, T.S. Littler,
An exploratory study of hearing loss and noise exposure in textile
workers.
The Ann. of Occ. Hyg. 1_ (1964) 323-333.
R. Gallo, A. Glorig,
P.T.S. changes produced by noise exposure and aging
Am. Ind. Hyg. Ass. Journal 2,5 (1964) 237-245.
The relations of hearing loss to noise exposure
A Report by subcommittee Z 24-X-2 (1954) 34.
N. E. Rosenwinkel, U. C. Stewart,
The relationship of Hearing Loss to Steady-State Noise Exposure
Am. Ind. Hyg. Ass. Quart. J£, (1957) 227-230.
J. Nixon, A. Glorig,
Noise Induced P. T. S. at 2000 and 4000 Hz.
J.A. S.A. 33(1961) 904-913.
W. Taylor, J. Pearson, A. Mair, W. Burns,
Study on noise and hearing in Jute weaving
J.A. S.A. 37 (1964) 113-120.
B. Kylin,
T. T. S. and auditory trauma following exposure to steady-state noise
Acta Oto-Laryng. Suppl. 152 (I960).
F. v. Laar,
Results of audiometric research at some hundreds of persons, working
in different Dutch factories
Publication: A. G. /S.A. C 23 of N. I. P. G. - TNO.
A. Spoor,
Presbyacusis values in relation to noise-induced hearing loss
Int. Aud. 6 (1967) 48-57.
C. W. Kosten and G. J. van Os,
Community reaction criteria for external noises
The Control of Noise, NPL-Symposion no. 12, P. 373-382, HMSO
1962.
-------
TABLE 2a
Predicted N1PTS for 75 dBA
^
H
5
A N
u -
^ •— 4
V
O, in
03 ^
n
.•«
i
1 .
a< fy-
w ""1
<
90
75
50
25
10
90
75
50
25
10
90
75
M 50
N
25
10
90
75
MI 50
91 25
10
90
75
Ml 50
VD| 25
10
90
75
00
50
25
10
er-Vermeer
'A
u
m
CO
it)
|li
0.0
0.0
0. 0
0.0
0.0
2. 5
1. 5
.5
0.0
0..0
0.0
0.0
0,0
0.0
0.0
10.0
6.0
2.0
Q.O
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
c
o
to
• r*
ja
o
K
10 Year
. 8
. 5
.3
. 2
. 1
1.5
.9
.6
.3
.2
1.6
1.0
.6
.4
. 3
3.6
2.3
1.4
.8
.5
2.5
1.6
.9
.5
.3
_
-
-
_
j3
U)
ft
a
ra
—
••
-
-
—
-
_
_
-
_
_
.
-
-
_
-
-
-
-
_
-
-
• ^
-
—
„
•,
_
_
V
0)
o
t>
Passchier-'
20
0.0
0.0
0.0
0.0
0.0
2.5
1.5
. 5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
6.0
2.0
0.0
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
Robinson
Year
1. 1
. 7
.4
.3
.1
2.1
1.4
.8
.5
.2
2.0
1.3
.8
.5
. 3
5.2
3.4
2.0
1.2
.7
3.6
2.3
1.4
.8
.5
—
_
_
_
_
fl
A
cd
CQ
_
_
-
-
-
—
_
-
.
-
„
-
-
_
-
—
H
-
-
•
.
„
-
.
-
_
_
_
—
to
V
i
Passchier-Ve:
40
0.0
0.0
0.0
0.0
0.0
2.5
1.5
. 5
0.0
0.0
Q.O
0.0
0.0
o.o
o.o
10.0
6.0
2.0
0.0
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
Robinson
Year
1.6
1.0
.6
.4
.2
3.0
2.0
1.2
.8
.4
3.0
1.9
1.1
.7
.5
7.5
4.9
3.0
1.8
1.1
5.2
3.4
2.0
1.2
.7
—
_
_
-
_
bo
1
^
.
-
_
-
„
-
_
_
-
^
-
.
.
-
^
_
-
.
•
.
-
.
-
.
_
-
-
_
4
-------
TABLE 2b
Predicted NIPTS for 80 dBA
^^
ffi
u .
cu m
m
•4
A tu
y •*
a, M
CO -^
in
90
75
50
25
10
90
75
50
25
10
90
75
M 50
(VJ
25
10
90
75
Wl 50
'1 25
10
90
75
&
50
*° 25
10
90
75
col 5°
°° 25
10
chier-Vermeer
W
C
I s 4 3 §
« .5 M a .S
» -° 2 » ja
at o n rt O
PH tf fl p^ gj
20 Year _ 40 Year _
0.0 2.2 0.0 0.0 3.2 0.0
0.0 1.4 0.0 0.0 2.1 0.0
0. 0 .8 0. 0 0. 0 1.2 0. 0
0.0 .5 0.0 0.0 .7 0.0
0.0 .3 0.0 0.0 .4 0.0
3.5 4.0 1.0 3.5 5.6 .8
2. 5 2.6 1.0 2. 5 3. 7 .7
1.5 1.6 .9 1.5 2.3 .5
.2 .9 .8 .2 1.4 .4
0.0 .6 .7 0.0 1.0 .3
0.0 3. 8 - 0.0 5. 5
0.0 2.4 - 0.0 3.6
0.0 1.5 - 0.0 2. 2
0.0 .9 - 0.0 1.3
0.0 .5 - 0.0 .8
13.8 9.3 4.1 13.8 12.9 .9
9.9 6.2 3.9 9.4 8. 7 1.5
6.0 3.8 3.7 5. 5 5. 5 2.2
1.0 2.3 3.1 .5 3.4 2.8
0.0 1.4 2.5 0.0 2.1 3.1
3.8 6.6 - 3. 8 9. 3
3.6 4.3 - 3.6 6.2
3.4 2.6 - 3.4 3. 8
2. 1 1.6 - 2. 1 2. 3
.8 1.0 - ,8 .1.4
.6 - .6
.4 .4
.2 - .2 -
.2 - .2
.2 .2
-------
u
CO .
in
TABLE 2c
Predicted N1PTS for 85 dBA
f4 fe U
« » «
4) 0) 0
U » »
•rj O _, *d O rt *r} O
o 2 ,2 u 2 •* o S
«0 .5 M 09 .5 J? W .H
00 J3 P TO , J2 ••: * J3
-------
n
>
H
8
•8 3
o -:
in
TABLE 2d
Predicted NIPTS for 90 dBA
t* M t*
0) « 4>
V V V
4) " 4) 4>
^ ^ ^
I I I
o) d v d v d
s g d 2 g « s i d
8 .S •§> S .3 f 85-4
10 Year20 Year40
90 2.4 4.2 5.5 3.2 5.1 6.9 4.5 8.6 7.3
75 1.6 2.4 3.8 2.4 3.1 4.9 3.8 5.4 5.5
50 .8 1.5 2.6 1.6 2.0 3.3 3.1 3.5 3.2
25 .6 1.0 2.0 1.4 1.1 2.6 2.8 2.1 3.0
1° .5 .8 1.8 1.2 .8 2.2 2.5 1.4 2. 5
90 7.3 7.8 11.6 8.3 9.8 9.8 9.5 13.8 6.6
75 6.4 5. 1 8. 5 7.0 6.0 8. 5 8. 2 9. 8 6. 1
50 5.1 3.3 6.3 5.7 4.5 7.1 6.9 6.7 5.4
25 3. 7 2. 1 4. 4 4. 3 2. 8 5. 7 5. 5 4. 3 5.9
10 2,3 1.5 3.3 2.9 1.9 4.7 4.1 2.7 6.2
90 6.8 8.8 - 9. 2 12.2 - 13.4 16.4
75 4.6 5.8 - 7.0 8.3 - 11.2 11.5
50 2.4 3.6 - 4.8 5.2 - 9.0 7.4
25 1.6 2.1 - 4.0 3.1 - 8.2 4.6
10 .8 1.4 - 3.2 2.0 - 7.4 2.9
90 23.6 18.8 30.1 23.6 24.0 18.7 23.6 29.5 46
75 20.8 13.4 22i7 20.8 17.8 19.2 21.3 22.9 7*8
50 18.0 8.7 17.4 18.0 12.1 18.6 18.5 16.3 lo!4
25 13.2 5.5 11.5 13.2 7.8 14.9 13.7 10.9 14.8
10 8.4 3.5 7.7 8.4 5.1 12.4 8.4 7.3 17.'4
90 18.3 14.2 - 18.3 18.8 - 18. 3 24.0
75 15.6 9.8 - 15.6 13.4 - 15.6 17.8
Ml 50 12.9 6.2 - 12.9 8.7 - 12.9 12.0
1 25 6.. 7 3.8 - 6. 7 5.5 - 6.7 7. 8
10 ..5 2.4 - .5 3. 5 .5 5.1
90 8.9 - - 8.9 - 8.9
75 6.7 - - 6.7 - 6.5
•gl 50 4. 5 - - - 4. 5 - - 4.5
25 4. 5 - 4. 5 - 4.5
10 4. 5 - - 4. 5 - 4. 5
-------
Iff
s
u
O -*
0>
ft
w
fc^
5
OT
TABLE 2e
Predicted NIPTS for 95 dBA
ft>
0) 4) Q)
i * t
* h ti
2 g 0 c v «
2gg 22S 2 g
S -5 -S 8 .« f 85
m J3 J3 SO JO 3 WpQ
rtOrt rt ° .5 cSO
& _ « _ W & « « PH tf
10 Year 20 Year 40 Year
90 5.6 8.1 9.7 7.9 12.6 11.7 11.6 15.1 12.9
75 4.5 5.1 6.8 6.8 8.4 8.2 10.5 10.2 9.1
50 3.4 3.3 4.6 5.7 5.6 5.5 9.4 6.9 6.1
25 2. 3 2. 1 3. 7 4.6 3. 5 4. 5 8. 2 4. 6 5.0
10 1.2 1.4 3.0 3.5 2.3 3.7 7.0 3.0 4.1
90 12.1 13.0 17.6 13.7 17.7 14.1 15.3 10.6 11.1
75 10.8 9.1 13.0 12.5 12.9 12.0 13.6 14.3 9.3
50 9.5 6. 2 9.4 11.3 9.1 9.9 11.9 10.0 7.7
" 7-8 4-° 6.3 9.6 6.0 7.9 9.2 6.8 8.6
10 ^ 1 2. 7 4. 7 7.9 3. 7 6.6 7.5 4. 4 9.0
90 12.4 14.9 - 18.2 19.6 - 27.6 24.9
75 9.1 10.3 - 14.9 14.1 - 24.3 18.6
Wl 50 5. 8 6.6 - H.6 9.2 - 21.0 12.7
' 25 2.6 4.0 - 8.4 5.8 - 17.8 8.3
10 0.0 2.6 - 4.2 3.8 r 13.6
5.4
90 31.4 27.7 41.2 31.4 33.1 21.3 31.4 38.1 5.8
75 29.7 21.2 31.7 29.7 26.6 23.6 29.7 32.0 9.8
50 28.0 14.8 23.7 28.0 19.5 23.1 28.0 24.7 12.7
25 24.5 9.8 14.1 24.5 13.4 18.1 24.5 17.8 19.4
10 21.0 6.5 9.8 21.0 9.1 15.5 21.0 12.6 23.9
90 25. 7 22. 2 - 25. 7 27. 7 - 25. 7 33. 1
75 22.1 16.3 - 22.1 21.2 - 23.1 27.6
!g| 50 18.5 10.9 - 18.5 14.8 - 19.5 19.5
25 n-4 6.9 - 11.4 9.8 - 12.4 13.4
J0 4.3 4.5 - 4.3 6.5 - 4. 3 9. 1
90 15.1 - - 15,1 - - 15.5
75 12.1 . 12.1 - - 12.5
50 9.1 . . 9>1 . _ 9<5
25 9.1 _ 9.! . . 9>5
10 9.1 - 9.1 - - 9.5
-------
TABLE 2f
Predicted NIPTS for 90 dBA
"5"
*%
u -
0) .
£•«>
w .
to
V
43 "2
o "P
0,™
to "^
irv
90
75
50
25
10
90
75
50
25
10
90
75
«| 50
N| 25
10
90
75
MI 50
^1 25
10
90
75
£
50
° 25
10
90
75
S
GO
50
25
10
schier-Vermeer
u
«s
2.4
1.6
.8
.6
.5
7.3
6.4
5.1
3.7
2.3
6.8
4.6
2.4
1.6
.8
23.6
20.8
18.0
13.2
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.7
4.5
4.5
4.5
§
09
1
O
10 Year
4.2
2.4
1.5
1.0
.8
7.8
5.1
3.3
2. 1
1.5
8.8
5.8
3.6
2.1
1.4
18.8
13.4
8.7
5.5
3.5
14.2
9.8
6.2
3.8
2.4
—
-
_
-
-
rt
"rt
P K
O
1.0
4.3
4.3
3.3
1.3
_
-
-
^
—
-1.0
4.0
4.0
3.0
0
5.0
11.0
9.0
3.0
1.0
-8.0
2.0
3.0
4.0
7.0
—
-
_
-
-
JH
4>
g
Paaschier-Ven
20
3.2
2.4
1.6
1.4
1.2
8.3
7.0
5.7
4.3
2.9
9.2
7.0
4.8
4.0
3.2
23.6
20.8
18.0
13.2
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.7
4.5
4.5
4.5
Robinson
Year
5. 1
3. 1
2.0
1.1
.8
9.8
6.0
4.5
2. 8
1.9
L2. 2
8.3
5.2
3. 1
2.0
24.0
17.8
12.1
7.8
5.1
18.8
13.4
8.7
5.5
3.5
-
-
*•
-
a
15
P Jj
DC v
°o
10.6
8.0
5.0
2.0
1.0
mm
.
_
_
-
14.0
16.0
9.0
5.0
2.0
26.0
24.0
20.0
15.0
13.0
10.0
19.0
18.0
12.0
6.0
-
-
-
.
V
V
Passchier-Ve
40
4.5
3.8
3.1
2.8
2.5
9.5
8.2
6.9
5.5
4. 1
13.4
11.2
9.0
8.2
7.4
23.6
21. 3
18.5
13.7
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.5
4.5
4.5
4.5
Robinson
Year
8.6
5.4
3.5
2. 1
1.4
13.8
9.8
6.7
4.3
2.7
16.4
11.5
7.4
4.6
2.9
29.5
22.9
16.3
10.9
7.3
24.0
17.8
12.0
7.8
5.1
_
-
-
-
-
NIOSH Data
34 Year
13.3
13.0
6.3
2.6
1.3
mm
• _
-
-
-
20.0
27.0
12.0
4.0
5.0
-10.0
14.0
20.0
7.0
5.0
-3.0
12.0
22.0
9.0
10.0
_
-
-
-
-
-------
3. Other Methods. The National Institute of Occupational Health
and Safety (NIOSH) also presented data which have not been smoothed. Table
2f has some of these same data incorporated for comparison. This data base
was not used because (1) it only predicts NIPTS for 90 dBA, (2) the sample
size was very small (22 workers for some of the age groups), and (3) gome
type of smoothing of the data would be required in order to make it a pre-
dictive method. The data is presented in Table 2f in order to show (1) that
raw data requires treatment (such as provided by Robinson, Passchier-
Vermeer or Baughn) before it is useful, and (2) the NIOSH data is not out of
line with the predictive methods used in this report. There is, however,
one method in the literature which differs greatly with other methodologies.
This is Kryter's latest work published in the Journal of the Acoustical Society
of America, 1973.
Figure 1 shows a plot of predicted NIPTS values for each of the
three selected methods as well as Kryter's predicted values. Of all the studies
compared, only Kryter does not seem to be in general agreement with the
three methods selected. Therefore, a special discussion of his method i>
included. At this point, however, attention will focus only on the methods of
Passchier-Vermeer, Robinson, and Baughn.
4. Simplification of Data. Now that three different methods have
been selected, the question remains as to how to use the data. The data are
simplified to three curves (representing different philosophies of what and
whose hearing should be protected) for three audiometric frequencies. Two
curves are the expected NIPTS (maximum and a 10 year exposure point) of
of the sensitive ears on the 90 percentile points with respect to SPL. The
other curve is the average NIPTS expected during 40 years of exposure as
averaged over all the population percentiles. This third curve is approximated
closely by the median NIPTS level after 20 years of exposure. The three
audiometric frequencies presented were speech (average of 0. 5, 1, and 2 kHz),
speech (average of 0. 5, 1, and 4 kHz) and 4 kHz. A Table relating percent
of population with more than a 5 dB NIPTS at 4000 Hz versus exposure is also
developed. The data are presented in the sequence in which reduced so that
a user may, at his discretion, stop and use as a basis of his decision the data
one or more steps before the manipulation that provides the final curves
discussed above.
5. Details of Selected Methodologies.
a) Passchier-Vermeer (1971)
Passchier-Vermeer results are in graph form (see Figure 2).
Tables 3 and 4 are then used to calculate the effects of age and the correction
necessary for considering different percentile levels. The details of the
calculations of the values in Table 2 are as follows:
10
-------
70
60
50
£40
Q
LJ
y
o
LJ
CC
QL
30
20
10
40
SPEECH ^(.5,1,2 KHz)
75PERCENTILE
8hr. DAILY EXPOSURE
FOR 40 YEARS
KRYTER
PASSCHIER
VERMEER
BAUGHN
ROBINSON
O O
A A
D O
50
60
90
100
10
70 80
SPL(dBA)
* KRYTER CONSIDERS THIS HEARING LEVEU FROM TABLE in [KRYTER (ie)]J
Figure 1
11
-------
4000 Hz
CD
TJ
O
I!
o
o^
o
MEDIAN HEARING LOSS CAUSED BY
EXPOSURE TO NOISE FOR 10 YEARS,
AS A FUNCTION OF THE NOISE RATING
FOR 500 TO 2000 HERTZ.
94
104
dBA
Figure 2
12
-------
TABLE 3
(from PaBsehier-Vermeer)
Frequency j Increase; of 1)5070 in relation to 050$ (T = 10)
* .v vr vi *•» w • • ™t/
for exposure time a
500 HR 2 #
1000 "
2000 »
3000 "
4000
6000 "
8000 "
2,5
10
1
0
0.
0.28 (NR-92)
0
0.37 (NR-92)
of at least 10 years 'j
per year
n
it
n
n
n
n
n
n
I
1
NR < 92
NR > 92
NR < 92
NR > 92
TABLE 4
(from PaB«chler-Verme«r)
NR for 500
{o 2000 Hz
75
80
85
90
94
98
Number of decibels to bo addod io Degj, in order io calculate Dyetf
500 Hz
0
0
0
0
0
0
1000 Hz
0
0
0
0
0
o.5
2000 Hz
0
1
2
3
4.5
7
3000 Hz
0
0
2.5
4.5
4.5
4.5
. 4000 Hz
4
3.5
3
2
0.5
0
6000 Hz
0
1
2.5
3.5
4
5
8000 Hz
0
1
2
^
3
3
NR for 500
to 2000 Hz
75
80
85
90
94
98
Number of decibels to bo substrocted from Dg£ ;in order to calculate bft% J
500 Hz
0
0
0
0
0.5
1.5
1000 Hz
0
0
0
0
0.5
1.5
2000 Hz
0
0
0.5
3
4
5
3000 Hz
1
1
2.5
3.5
3.5
3.5
4000 Hz
5
5
5
4
2
1
6000 Hz j SOCO Hz |
1
!•*
7
7.5
8
0
0
S
0
0
13
-------
Reference: "Hearing Loss Due to Exposure to Steady-State Broadband Noise. "
(1) Converted N. R. into dBA by adding formula dBA = N. R. + 4.
(2) Procedure used was outlined in pages 23-25.
(3) Noise-induced shift of hearing level (Dx), not approximation of noise
induced hearing loss (D'x) was calculated.
(4) (Dx) values were obtained from Figure R35-A and Tables A and B.
(5) For 75 dBA, the curves of R35-A were extended slightly by straight lines.
(6) Speech hearing loss was obtained from averaging Dx for 500, 1000, and
2000 Hz frequencies.
(7) Since no method was suggested in her original report for estimating the
10 and 90 percentile levels, the corrections used to estimate the 25 or
75 percentile levels were doubled in order to approximate the 10 or 90
percentile levels. The error of this approximation will be less than
10 percent for a normal distribution. This is in agreement with
Passchier-Vermeer's supplement (1969) to the main report.
In her 1971 paper "Occupational Hearing Loss", Passchier-Vermeer
does provide NIPTS values for the 10 year exposure point. These values
agree with the approximation used in this supplement.
b) Robinson
Robinson provides a formula and a set of Tables (see
Tables 5 and 6) which can be used to calculate NIPTS. A nomogram is also
presented which allows calculation of hearing levels of noise-exposed popu-
lations since the presbycusis correction is included. Details of the calcu-
lations used to obtain the values of Table 2 are as follows:
Reference: "The Relationships Between Hearing Loss and Noise Exposure. "
(1) Used LA = dBA.
(2) Used procedure outlined on page 18 except that the formula:
LA + 10 LOG T/TO -I- Un - Xi
H = 27. 5
1 + TANK
15
was used instead of the nomogram.
(3) Table 5 (page 6 of reference) was used to find \i for TO = I year.
(4) Table 6 (page 7 of reference) was used to find Un, which relates H to a
percentile of the population.
(5) T = time of exposure in years and H = noise induced hearing loss.
(6) Speech hearing loss was calculated from averaging H for 500, 1000 and
2000 Hz frequencies.
c) Baughn
Baughn presents a set of Tables (see Tables 7 and 8) that
give the actual hearing levels of 8 different age groups for 9 percentile levels
under three exposure conditions. Considering the 78 dBA group as non-
exposed groups, the calculations are as follows:
14
-------
TABLE 5
Frequency parameter X in H-function
(from Robinson)
Audiometric
frequency
(kHz)
0. 5
1
2
3
4
6
X (dB)
To = * year
130.0
126.5
120.0
1 14. 5
112.5
115,5
15
-------
TABLE 6
Percentile parameter u in H-function
(from Robinson)
Percentile n
"Sensitive ears"
1 *
2
3
5
7
Deoile 1 0
15
20
Quartile 25
30
40
Median 50
60
70
'Quartile 75
80
85
Deoile 90
93
95
98
99 *
"Resistant ears"
u
13.8
12*1
11.1
9.8
8.7
7.6
6.0
5*0
4*0
3.1
1.5
0
- 1.5
- 3.1
- 4.0
- 5.0
- 6.0
- 7.6
- 8.7
- 9.8
-11.1
-12.1
-13.8
Extrapolated.
16
-------
TABLE 7 (from Baughn)
INTERPOLATED AND EXTRAPOLATED FROM FIELD
(Speech (.5, 1, 2 kHz)
Int. AGE 18 - 23 ASH 24 - 29 AGE 30 - 35 AGH 36 - 4!
Pec. 5173 3173 S17T
Points SO 8590 95 £0 S5~~ 90 95 80 ST~ 90 95 80
1 -.8 -»4 -.1 .4 .2 1.0 1.7 2.7 .6 1.4 2.5 3.9 1.0 2.0 3.1
2 .55 1.1 1.4 1.2 1.7 2.6 3.4 4.6 2.2 3.1 4.3 6.0 2.6 3.S 5.1
3 1.5 2.1 2.4 3.0 2.8 3.8 4.6 5.0 3.3 4.3 5.6 7.5 3.8 5.0 6.5
4 2.5 3.1 3.5 4.1 3.8 4.9 5.8 7.5 4.4 5.5 6.9 8.9 4.9 6.3 7.8
5 3.7 4.4 4.8 5.5 5.2 6.4 7.4 9.0 5.8 7.0 8.6 10.8 6.4 7.9 9.6
6 4.3 5.1 5.6 6.4 6.0 7.4 8.6 10.4 6.7 8.1 10.0 12.5 7.4 9.2 11.1
7 5.0 5.9 6.5 7.4 7.0 8.6 10.0 12.2 7.8 9.5 11.6 14.6 8.6 10.7 13.0
8 6.0 7.1 7.7 8.9 8.4 10.3 11.9 14.5 9.3 11.3. 13.8 17.4 10.3 12.7 15.5
9 7.8 9.2 10.1 11.6 10.3 13.4 15.5 18.9 12.2 14.7 18.1 22.7 13.4 16.6 20.2
AGE 42 -. 47 ACE 48 - 53 AGE 54 - 59 AGE 60 - 65
1 1.6 2;7 3.9 5.3 2.7 3.7 4.9 6.7 4.1 5.3 6.8 8.4 6.8 8.0 9,2
2 3.3 4.6 5.9 7.6 4.5 5.7 7.2 9.2 6.2 7.6 9.3 11.2 9.3 10.7 12,1
3 4.5' 5.9 7.4 9.2 5.9 7.1 8.7 10.9 7.7 9.2 11.0 13.1 11.0 12.6 14.1
4 5.7 7.2 8.8 10.8 7.2 8.6 10.3 12.7 9.2 10.8 12.8 15.0 12,8 14.5 16.1
5 7. 3 8.9 10.7 12.9 8.!) 10.4 12.3 14.9 11.1 12.9 15.0 17.5 15.0 16.9 18.7
6 8.5 10.3 12.4 15.0 10.-! 12.1 14.3 17.3 12.9 15.0 17.4 20.3 17.4 19.6 21.7
7 9.9 12.0 14.4 17.4 12.1) 14.0 16.6 20.1 15.0 17.4 20.3 23.6 20.3 22.8 25.2
8 J1.8 14.3 17.2 20.8 14.:; 16.7 19.8 24.0 17.9 20.8 24.2 28.2 24.2 27.2 30.1
9 15.3 18.7 22.5 27.1 18.7 21.8 25.8 31.3 23.3 27.1 31.5 36.8 31.5 35.5 39.3
-------
00
TABLE 8
4000 Hz
Int.
Dec.
Points
1
2
5
A
5
6
7
8
9
1
2
3
4
5
6
7
8
9
AGE 18 - 23
78
.37
1.44
2.34
3.28
4.1
5. OS
6.85
7.95
10.7
AGE
8.32
13.5
18.2
22.1
26.0
30.2
35. 6
41.3
50.4
W
2.09
4.09
5.66
7.13
8.7
10.6
13.05
16.18
23.66
42 - 47
19.3
26.7
31.6
36.1
41.0
45.1
50.0
56.2
64.8
92
2.8
5.74
8.40
11.48
14.0
17.5
21.8
26.6
37.2
23.9
30.8
35.9
41.4
46.0
50.1
54.3
59.8
66.2
AGE
78
1.37
3.46
5.46
7.46
9.1
11.10
14.2
16.8
22.2
AGE
12.1
18.7
23.9
28.5
32.8
37.4
42.6
48.9
58.4
24 - 29
"TTC
5.7
9.88
12.9
15.96
19.0
22.4
26.6
28.5
45.03
48 - 53
24.1
31.6
37.1
41.8
'46. 4
50.6
54.8
60.3
67.7
92
7.67
13.15
17.54
23.01
27.4
32.61
38.91
46.03
61.38
30.5
37.1
41.7
46.7
50.8
54.4
58.4
63.0
69.1
AGE
78
3.0
6.15
9.15
11.87
14.3
17.12
21.31
25.17
32.6
AGE
17.2
24.5
30.9
35.3
40.1
44.9
49.7
56.1
65.0
(from Baughn)
30 - 35
— B6
9.77
15.6
20.09
23.99
27.9
32.09
37.11
43.52
53.01
54 - 59
30.3
38.2
43.4
48.1
52.3
56.5
60.1
64.9
71.7
92
12.6
18.9
23.8
30.5
35.0
39.9
45.9
52.5
64.8
37.3
43.4
47.2
51.6
54.9
58.2
62.6
65.9
71.4
78
5.2
9.6
13.4
16.8
20.0
23.6
28.6
33.4
42.4
24.0
32.3
39.2
44.1
49.0
53.9
57.8
63.7
70.6
AGE 36 - 41
4 V
"SET
14.7
20.9
25.8
30.4
34.9
39.1
43.2
51.0
60.0
AGE 60 - 65
35. S
43.6
48.1
52.0
55.9
59.3
63.7
66.5
72.7
92
17.9
24.4
29.6
35.7
40.6
45.5
50.3
56.0
64.1
44.0
50.4
53.3
56.3
58. 6
62.1
66.2
09.7
7S.O
-------
(1) Use Table 7 (6a of reference) and Table 8 (9 of reference) from Baughn's
data.
(2) NIPTS for speech was considered as the difference in hearing of a certain
percentile of people, who are exposed to a noise level greater than 80 dBA
minus the. hearing level of that same percentile of people who are exposed
to only 80 dBA.
(3) Percentile levels were given in units of 10 percent only. The 25 and 75
percentile points were obtained by averaging 20 and 30, and 70 and 80 per-
centile values, respectively.
(4) The data was given by age groups with 6 year differences. Linear inter-
polation was used where necessary to obtain exposures for 10, 20 and
40 years.
(5) HL values for 4000 Hz at 80, 85 and 90 dBA calculated from Baughn's
data by linear interpolation between the 78 and 86 dBA data points or the
86 and 92 dBA data points. Values at 95 dBA were obtained by linear
extrapolation from the 86 and 92 dBA points. NIPTS due to some exposure
level, e. g. , 85 dBA, was calculated as the HL at 85 dBA minus the HL
at 78 dBA for the same percentile and age group.
6. Manipulation of Data. These values were manipulated and
simplified as follows: Tables 9, 10 and 11 were constructed by averaging
the NIPTS values of Table 2 over a 40 year lifetime (age 20 to age 60).
After the NIPTS values were averaged over time for various population
percentiles, the results were averaged over the total population. A graphic
method was used to calculate "Average NIHL during 40 Years Exposure".
The 0, 10, 20 and 40 year data points were plotted on graph paper. The area
under the curve drawn through these points was measured and then divided
by 40 to obtain the "average NIHL during 40 Years' Exposure." A graphic
method in which the . 9, • 75, . 5, . 25, and . 1 percentile points were plotted
was used to calculate "Average Loss of Total Population During 40 Years of
Noise Exposure". The area under the resultant curve was measured and
normalized to obtain the desired value.
From this average, Table 12 was developed. Tables 13 and
14 come directly from the data of Table 2. Table 13 provides the expected
NIPTS after 10 years of noise exposure that will not be exceeded by 90 per-
cent of the population (.9 Percentile level). Table 14 depicts the maximum
NIPTS that will be encountered during a typical 40 year exposure which starts
at age 20. Normally this occurs at 60 years o£ age, but for 4000 Hz,
Passchier-Vermeer's method shows that this occurs after both 10 and 40 years
of exposure time, while Baughn's data indicates that this occurs at the 10
year exposure point.
The resulting NIPTS values of Tables 12, 13 and 14 are now
averaged over the three methods. This grand average is presented in Fig-
ures 3-8. Figures 3, 4 and 5 compare the 3 different ways (Max NIPTS,
. 9 percentile; NIPTS after 10 year exposure, . 9 percentile; and average
NIPTS of total population during 40 years) of considering the data at three
19
-------
TABLE 9
Average NIPTS during 40 Years Exposure
1/3 (.5, 1, 2 kHz)
dBA
Passchier-Vermeer 0
80 Robinson
Baughn
Passchier-Vermeer
85 Robinson
Baughn
Population Pe re entiles
.9 .75 .5 .25
.1
Passchier-Vermeer 3.0
90 Robinson
Baughn
Passchier-Vermeer 9.2
95 Robinson
Baughn
Average
Loss of Total
Population
0
2.0
0
0
1.3
0
0
.8
0
0
.4
0
0
.2
0
0
.9
0
.9
3.6
2.8
.5
2.4
2.0
.2
1.4
1.3
.2
.8
1.1
.1
.5
.9
.4
1.6
1.6
3.0
5.5
6.0
2.3
3.2
4.3
1.6
2.1
3.0
1.4
1.2
2.3
1.2
.9
1.9
1.9
2.5
3.5
9.2
11.0
10.2
6.3
7.5
7.2
5.5
4.4
5.0
4.4
3.1
3.8
3.5
2.1
3.4
5.8
5.2
5.7
20
-------
TABLE 10
Average NIPTS during 40 Years Exposure
1/4 {. 5, 1, 2, 4 kHz)
Population Percentiles
dBA
80
85
90
95
Passchier-Vermeer
' Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn
3.
3.
*
5.
6.
5.
8.
9.
8.
q
4
6
8
1
3
1
1
3
,8
14.7
15,
,8
13.3
.75
2.5
2.3
.7
4.0
4.2
4.0
6.9
6.4
7.2
12.1
11.7
10.7
.
1.
1.
•
2.
5
5
5
7
,9
2.7
3
5
4
6
11
7
8
.5
.7
.3
.0
.1
.7
.5
.25
• •
4
I
1
1
3
4
2
4
9
5
6
2
.8
,6
.6
.6
.0
.3
.7
.9
.4
.3
.9
.1
0 .
.6
.6
.3
1.0
2.7
3.0
1.9
4.3
7.9
3.6
6.4
Average
Loss ot Total
Population
1.4
1.7
<
2
3
3
5
4
6
11
8
9
.7
.9
.2
.7
.7
.9
.3
.1
.5
.0
21
-------
TABLE 11
Average NIPTS during 40 Years Exposure
4000 Hz
Population Percentiles
dBA
80
Average
Los a of Total
Passchier-Vermeer 13.8
Robinson
Baughn
Passchier-Vermeer 17.8
85 Robinson
Baughn
Passchier-Vermeer 23.6
90 Robinson
Baughn
Passchier-Vermeer 31.4
95 Robinson
Baughn
.
-------
TABLE 12
Average Loss of Total Population
during 40 Years of Exposure
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
1/3 (.5, 1, 2 kHz)
75
-
-
-
75
-
-
«•
75
-
-
-
80
0
.9
0
.3
1/4
80
1.4
1.7
.7
1.2
80
5.5
4.2
3.0
4.2
85
.4
1,6
1.6
1.3
(.5, 1
85
2.9
3.2
3.7
3.2
4000
85
10.6
7.4
10.0
9.3
90
1.9
2.5
3.5
2.6
, 2, 4 kHz)
90
5.7
4.9
6.3
5.6
Hz.
90
17.0
12.0
14.7
14.6
95
5.8
5.2
5.7
5.5
95
11.1
8.5
9.0
9.5
95
26.9
18.3
19.0
21.6
23
-------
TABLE 13
Noise Induced Hearing Loss
90 Percentile Level - 10 Years
1/3 (.5, 1, 2 kHz)
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
75
0
.8
0
.3
75
2.5
1..5
0
1.3
75
10.0
3.6
0
4.5
80
0
1.5
0
.5
1/4 (.
80
3.5
2.7
1.3
2.5
80
13.8
6.6
5.3
8.6
85
.9
2.8
2.5
'2.1
5, 1, 2 &
85
5.2
5.0
6.5
5.6
4000 Hz
85
17.8
11.6
18.6
16.0
90
2.4
4.2
5.5
4.0
4 kHz)
90
7.3
7.8
11.6
8.9
90
23.6
18.8
30.1
24.0
95
5.6
8.1
9.6
7.8
95
12.1
13.0
17.6
14.2
95
31.4
27.7
41.2
33.4
24
-------
TABLE 14
Maximum Hearing Loss from Noise .9 Percentile
1/3 (.5. 1, 2 kHz)
Passchier-Vermeer
Robinson
Baughn
Average
Worst Case
Passchier-Vermeer
Robinson
Baughn
Average
Worst Case
Passchier-Vermeer
Robinson
Baughn
Average
Worst Case
*This maximum value is for 10 years
maximum occurs at 40 years).
25
75
0
1.6
0
.5
Use
75
1.9
3.0
0
1.6
Use
75
10.0*
7.5
0
5.8
10.0
80
0
3.2
0
1.1
Robinson
1/4 (.
80
3.5
5.6
1.3
3.5
Robinson
80
13.8*
12.9
5.3*
10.7
.13.8
85
1.1
5.8
3.9
3.6
's Data
5, 1, 2 &
85
5.2
9.5
6.5
7.1
•s Data
4000 Hz
85
17.8*
20.5
18.6*
19.0
20.5
90
4.5
8.6
7.3
6.8
4 kHz)
90
9.5
13.8
11.6
11.6
90
23.6*
29.5
30.1*
27.7
30.1
95
11.6
15.1
12.9
13'.?.
95
15.3
19.6
17.6
17.5
95
31.4*
38.1
41.2*
36.9
41.2
(Otherwise the
-------
SPEECH (.5,1,2)
MAX. NIPTS
.9PERCENTILE
NIPTS AT 10 YR
.9PERCENTILE
AVERAGE NiPTS
OF TOTAL POPULATION
DURING 40YRS.
75
80 85 90
SPL(dBA)
95
100
Figure 3. Predicted NIPTS Averaged over the Methodologies of
Passchier-Vermeer, Baughn and Robinson.
26
-------
20
15
m
3 10
E
z
SPEECH i-(.5,l,2t4)
MAX
NIPTS .9PERCENTILE
NIPTS AT IOYR
.9 PERCENTILE
AVERAGE NIPTS OF
TOTAL POPULATION
DURING 40YRS
JL
70
75
80
85
SPL(dBA)
90
95
100
Figure 4. Predicted NIPTS Averaged over the Methodologies of
Passchier-Vermeer, Baughn and Robinson.
27
-------
30
CO
20
OL
I S
I w»
10
4000 Hz
MAX
NIPTS.9PER.
NIPTS.9PER.
AT 10 YEARS
AVERAGE NIPTS
OF TOTAL
POPULATION
DURING 40 YRS.
70 75 80 85 90
SPL(dBA)
95
100
Figure 5. Predicted NIPTS Average over the Methodologies of
Passchier-Vermeer, Baughn and Robinson.
28
-------
25
20
UJ
o:
log
K-X
UJ
<^to
K6-K
10
70
75
4000 Hz
(.5,1,2,4,^)
(.5,1,2KHz)
80 85 90 95
SPL(dBA)
100
Figure 6.
29
-------
50r
35
30
LU
25
h-
UJ
o
Sao
G_
0>
CL
2
10
4000HZ
^(5,1,2,4 KHz)
I (.5,1,2 KHz)
70 75 80
85 90
SPL(dBA)
95 100
Figure 7.
30
-------
40
39
y 30
P
2
UJ
O
5 25
Q.
O>
CO
Q.
2
70
4000 Hz
.5,1,2,4KHz)
75
80
85 90
SPL(dBA)
100
Figure 8.
31
-------
selected audiometric frequencies. It is these sets of figures, along with a
set of Hearing Risk tables and one other table to be discussed later, that are
considered sufficient to select the permissible A-weighted SPL for the 8 hour
noise exposure. Before such a selection is made, however, certain other
observations should be considered in detail.
7. Considerations.
a) NIPTS at 4000 Hz may decline with exposure for the very
sensitive ears, while increasing for resistant ears. Figures 9, 10, 11 are
a plot of the Hearing Levels of Baughn's data for .9, .5, and . 1 percentile
levels. Figure 12 is a plot of the difference between 85 dBA exposed groups
and 78 dBA exposed groups. As expected, during the first years of exposure
the sensitive ears (. 9 percentile) show a large increase in NIPTS while the
resistant ears (. 1 percentile ) show little increase. After 40 years of ex-
posure, the situation is completely reversed. If only the effect on the sen-
sitive ears is considered, the NIPTS for the noise resistant ears could be
improperly neglected.
It was for this reason that the "average NIPTS during 40 year*"
was calculated. For instance, using the results for Table 11 for 85 dBA,
Baughn's method gives approximately 12 dB average NIPTS for the sensitive
(.9) ears and approximately 8 dB average NIPTS for the resistant (. 1) ears.
Apparently the entire population, not just some super-sensitive individuals,
are significantly affected by noise during some part of their lifetime at the
4000 Hz audiometric frequency. Essentially, Table 11 was prepared to
show this effect.
One of the obvious reasons for the decline of NIPTS is seen
from Figure 11. As the total loss of hearing increases, regardless of the
reason, the influence of noise diminishes as there is only so much hearing
to be lost. The unanswerable question that remains is "what causes such a
large hearing loss as evidenced by Baughn's (78 dBA) supposedly non-noise
exposed group? " Is it aging, pathological conditions, non-occupational noise
exposure greater than 80 dBA, the fact that 78 dBA may still be capable of
causing a very significant loss in sensitive ears, or some combination of
these factors? Figure 13 is a plot of Baughn's 78 dBA {. 9) population versus
the 1960-62 Public Health Survey (PHS) data. For the most part, Baughn's
78 dBA (. 9) group shows less hearing loss than the PHS group, until age 50,
at which point the two groups become equal. One can conclude that Baughn's
78 dBA (. 9) group does not differ significantly from the general population.
Baughn did not screen for pathological conditions, so one would definitely
expect that such conditions would be an influence in both groups. The effect
of aging cannot be neglected. The rate of hearing loss for both the 78 dBA
group and the PHS (. 9) group is approximately 1. 5 dB/yr. Such a steep
increase does not occur for median hearing levels for 4000 Hz once a certain
age is reached (such as 50-70 years). It may not, therefore, be so unlikely
that for this sensitive 10 percent of the population, aging alone causes a very
32
-------
80
60
-J O
> <£
LLJ
CD
Z ^
DC ~
< d)
*
40
20
FROM BAUGHN
10 PERCENT HAVE BETTER HEARING
THAN LEVEL INDICATED (.1)
95dB(A)
90dB(A)
85dB(A)
78dB(A)
20 40 6O
AGE (YRS)
Figure 9
33
-------
FROM BAUGHN
80r
60
^O
•^ co
LJ »-«
C£> O
2 Cft
CC
5
40
- 20
50% HAVE BETTER HEARING THAN
LEVEL INDICATED (MEDIAN)
95dB(A)
90dBlA)
85dB(A)
78dB(A)
20 40 60
AGE (YRS)
Figure 10.
34
-------
60
LJ —
> O
LU CO
40
20
FROM BAUGHN
90% HAVE BETTER HEARING THAN LEVEL
INDICATED (9)
90dB(A)
85 dB(A)
78
-------
m
/ k
U)
bJ
LJ
40
m5
til O
y ¥ 30
" 00
LU Q
§iio
cp cx)
E
LU
X
0
FROM BAUGHfJ
4000 Hz
10
20
30 40
AGE(YRS)
50
60
70
Figure 12.
-------
CO
CO
•a
^o
H f£
- 60
^5
t CD
£D "O
CO
« ^O
UJ O
i
^ u
LU UJ
JL. CD
20
i
I
i
FROM BAUGHN
4000 Hz
20 40
AGE(YRS)
60
90dB(A)
Figure 13.
37
-------
80r
4000 Hz
90% HAVE BETTER HEARING THAN LEVEL INDICATED(.9)
60
O
v>
u
> 40-
UJ nw>
o
2
UJ
O—O BAUGHN 78 (.9)
X—X PUBLIC HEALTH
SURVEY 1960-62
I
20 40
AGE (YEARS)
60
Figure 14
38
-------
significant change even in the early years. These arguments are not brought
forth to prove that the rapid loss of hearing at 4000 Hz for this segment of the
population is not largely due to noise exposure, but rather to emphasize the
converse; over-protecting the population against noise exposure to prevent the
rapid rise in hearing loss at 4 Hz for 10 percent of the population may be
entirely futile. Such over-protection could easily come about if one made
the assumption that the 78 dBA is the main cause of the large hearing losses
in the sensitive 10 percentile.
b) Selection of a standard deviation for sensitivity to hearing
loss. Figures 12 and 13 demonstrate the difficulty of considering only mean
data at some exposure time and from these data estimating various percentile
levels by assuming a standard deviation. In order to predict Baughn's data,
the standard deviation must be constantly changed for increasing exposure
time. This emphasizes the care that must be taken if a noise limitation is
selected to protect 90 percent of the population instead of the median. The
90 percentile points can be seriously misestimated.
8. Risk of Noise Relative to Hearing Level Exceeding a Predeter-
mined Level or Fence. Up to this point discussion of hearing
risk, as it relates to an increase of the numbers of individuals who show a
hearing loss greater than some fence value, has not been undertaken. The
use of hearing risk as it relates to fences has been used for some time.
One of the major drawbacks to the use of fences, however, is that a single
fence only considers or protects hearing of individuals whose hearing is al-
ready near the fence values. Since fences have customarily been set relative-
ly high with respect to the median hearing level, the hearing of the majority
of the population is not considered.
Simply stated, the object of the fence is not to protect the
excellent hearing from becoming just good, but the fair hearing from becom-
ing bad. The argument that the excellent hearing will automatically be pro-
tected if the fair hearing is protected may not be true. Figure 15 is such a
counter example. Thus the use of hearing risk should not be the only basis
for selecting a noise limit for hearing conservation. Nevertheless hearing
risk is one way to give meaning to NIPTS values and for this reason Tables
15 and 16 were prepared. Table 15 shows the hearing risk in percentage
as calculated by Robinson. The 87, 92 and 97 dBA values were taken direct-
ly from Robinson and the 80 dBA values were calculated using his method.
Table 16 shows the same data as calculated from Baughn's curves. A typical
curve from Baughn's data is shown in Figure 16. The data agree well only
if a 10 dB is added to each of Robinson's fence values. This, as proposed
by Robinson, will account for the fact that Robins on1 a data have been care-
fully screened for pathological hearing losses while Baughn's data have not.
Baughn's data, in this regard, will certainly be more typical of the normal
population exposed to non-occupational noise. Therefore, the 10 dB correc-
tion will be added to Robinson's fence values in this report.
39
-------
HYPOTHETICAL SITUATION
X=5dB
BEFORE
NOISE
FENCE
25 dB
,.5%
-10 0 10 20 30
FENCE
B
-10 0 10 20 30
AFTER NOISE 10% HEARING RISK
,n * o, MORE SENSITIVE POPULATION
IU'° /0 AFFECTED MORE THAN LESS
_ SENSITIVE.
FENCE
AFTER NOISE 10% HEARING RISK
k vlo-5% ALL EAR AFFECTED WITH THE
NT SAME AMOUNT OF NIPTS.
-10 0 10 20 30
B
-10 0 10 20 30
THERE IS A 5 dB
DIFFERENCE BETWEEN
THE MEAN HEARING
LEVEL DEPENDING ON
WHICH ASSUMPTION
USED.
Figure 15
40
-------
TABLE 15
Robinson's Method
Noise risk for population at various ages for exposure
at constant noise level commencing at age 30.
Fence
Height
(ISO)
80 20
87 10*
92
80 25
87 15*
92
80 30
87 20*
92
80 35
87 25*
92
Noise
22 25
2
3
6
1
2
3
0
0
1
0
0
0
2
5
10
2
2
5
0
1
2
0
0
1
Risk
30
4
8
15
2
4
8
1
2
4
0
0
2
(%) at age
40 50 60
6
14
22
3
7
15
1
4
8
0
1
4
9
17
28
6
13
23
3
7
14
1
3
8
10
18
28
9
19
31
6
13
24
2
7
14
*Use these fence values for non-pathological
population.
41
-------
TABLE 16
Hearing Risk
Baughn's Data
Noise Fence
Level Height
dBA (ISO)
85 15
90
95
85 20
90
95
85 25
90
95
85 35
90
95
22
10
14
22
4
7
13
1
4
6
0
0
1
25
10
14
25
9
13
26
3
8
13
0
1
3
30
10
18
30
13
22
37
5
13
21
1
2
5
40
11
21
30
14
26
38
7
17
29
2
3
8
50
10
15
12
22
36
9
19
32
3
6
12
60 70
8
17
24
11
20
29
6
13
22
42
-------
lOOp
100
95 m
Q
ID
CO
90 g
-
UJ
85
80
GLORIG'S(I960)
NON-NOISE
EXPOSED
GROUP (2518)
10
20
30 40 50
AGE(yrs)
70
Figure 16.
43
-------
9. Percent of the Population with more than a 5 dB NIPTS at
4000 Hz Versus 8 Hour Noise Exposure Level. Since In general
the audiometric frequency at 4000 Hz is the most sensitive indicator of hear-
ing changes, a special table was derived to indicate the percentage of the
population expected to exceed a measurable NIPTS (greater than 5 dB) for a
daily 8 hour noise exposure of more than 40 years. The expected NIPTS
for each of the Sound Pressure levels was calculated or obtained graphically.
The NIPTS values of the three methodologies (Passchier-Vermeer, Baughn,
and Robinson) were averaged for the various percentile points. These points
were plotted on probability paper and a line was drawn through them with a
French curve. The intersect point with the 5 dB NIPTS line gives the per-
cent of the population that will exceed a measurable hearing change at that
exposure level. Table 17 is a summary of such data.
It must be emphasized that this method is approximate only
and is very sensitive to errors in the basic data. To emphasize this vari-
ability Table 18 was constructed in the same way as Table 17 except each
individual methodology was used alone;.
10. Selection of Limit for the 8 Hour Day. Data have been presented
that should allow the setting of a maximum allowable noise exposure (8 hour)
based on several considerations. The considerations emphasized in this
report have been: (a) average NIPTS of total population during 40 years,
(b) NIPTS not exceeded by 90 percent of the population at any time during
their exposure history, (c) percent of the population with a measurable hear-
ing change at 4000 Hz, (d) hearing risk as determined by a permissible hear-
ing loss or fence. If desired, other considerations can be developed from
the data. It is suggested that any recommended noise exposure be accept-
able with respect to all selected considerations.
11. Criticism of Kryter's Method.
a) From Figure 1 it is obvious that there is a very large dis-
parity between the predictions of Kryter and that of other researchers.
While Kryter may make some valid points, it is believed that there are
enough basic errors or inconsistencies in his methodology to make his re-
sulting predictions invalid. Therefore his NIPTS predictions were not con-
sidered in this document.
b) Faults and Inconsistencies of Kryter's Method
(1) Kryter arrives at the conclusion that a non-noise ex-
posed population is that population that has not been exposed to a continuous
8 hour noise of 55 dBA. This is based on extrapolation from Baughn1 s Data
and the Public Health Survey of 1962. The faults of this method are:
(a) Baughn1 s data are for 92, 86, and 78 dBA. From
just these 3 points which span a range of 14 dB only it is very questionable
that it is justifiable to extrapolate another 23 dB downward to determine
44
-------
TABLE 17
Derivation of % of Population with greater than
5dB NIPTS after 40 years exposure.
L
eq
in
U "o
rt Q) a\
fc c •-<
W H X»
< s s
._ 2 o
CO *J S
H *3 «
S u
o
.9
.75
.5
.25
.1
% of Population with more
than 5 dB NIPTS
72
3.8
2.2
.7
.4
0
4
75
5.8
3.6
1.7
•6
.4
15
80
9.2
6.5
4.4
2.2
1.7
44
82
11
8.4
6.4
4.2
3. 1
66
85
13. 5
11. 5
9.8
7.8
5.2
92
45
-------
TABLE 18
Precent of Population with more than 5 dB NIPTS versus L
Individual Methods
eq
Le 5 dB NIPTS *
4000 Hz
Pas schie r -Ve rmee r,
Unmodified
Passchier-Vermeer
Straight Regression
Line
Baughn
Robinson
72
4
14
0
N/A
12
75
15
28
1
N/A
17
80
44
50
21
N/A
54
82
66
66
50
34
66
85
92
78
75
77
83
46
-------
where the threshold SPLi that causes NIPTS is located. Furthermore, most
of the three points do not even align in a straight line, thus requiring the
extrapolation be made by a series of complex curves (see Figure 17).
(b) Kryter uses two different reports, which probably
have different biases, to determine the "NIPTS Threshold. " In fact Baughn
admits that he had a systematic error of at least 5 dBA and perhaps more in
his absolute thresholds. For instance TTS was a problem as Baughn had to
test people during working hours. The problems do not unduly jeopardize
the validity of Baughn1 s data when compared with itself as at least some of the
biases will be expected to cancel. But when Baughn1 s data are compared to
other data, such differences will not tend to cancel and must be fully con-
sidered. Looking at the PHS curves and Baughn1 s 78 dBA curves versus age,
(Figure 18), it can be noted that they look very similar except Baughn1 s
78 dBA curve is displaced upward by 10 dBA. Kryter would attribute this
upward shift to the fact that the 78 dBA exposure was still causing a sub-
stantial hearing loss. But plotted also in Figure 18 is the median of Baughn's
pre-exposure audiograms of new 18 year old employees. Note that even for
this group, there is still an 8 dB variation in the Public Health Survey data
and Baughn1 s. This variation shows that there were indeed systematic
differences between the studies. These differences may have come from
audiometric techniques, differences in the population of this midwest area
versus the nation as a whole, or some other subtle bias; however, it is clear
that the 78 dB exposure is not, a priori, the cause of the 10 dB discrepancy
between Baughn1 s data and the Public Health Survey data.
(c) In order to demonstrate the sensitivity of Kryter's
method to systematic error between the two sets of data, consider that the
hearing levels of Baughn1 s subjects were systematically 10 dB too high.
This 10 dB error has significant implications with respect to Kryter's NIPTS
threshold prediction. See Figure 17 for a typical correction if Baughn's data
are reduced by 10 dB. Such a 10 dB reduction now brings the "NIPTS Thres-
hold" up to 75-80 dBA with far less extrapolation. This puts Kryter more in
line with other researchers. It should also be apparent that the gain in "NIPTS
Threshold" was 20-25 dB for a change of only 10 dB in Baughn1 s raw data.
This indicates that with an arbitrary fence of so many dB, the results obtained
are very sensitive to the absolute thresholds of the data used. One only has
to look at the literature to see how often a 10 dB or greater difference has
occurred between researchers as to what is the median threshold level. The
10 dB difference between the 1951 ANSI standard and the 1969 ANSI standard
for the speech frequencies is an obvious example. It should be noted that
even if the systematic difference in Baughn's data was as small as 5 dB, which
is the minimum amount of error predicted by Baughn, Kryter's methodology
would still predict that the threshold of the effect is at 65-70 dBA, not 55 dBA.
Therefore, even if one would agree with Kryter that his methodology is adequate,
one must correct his threshold value of 55 dBA by at least 10-15 dBA and
probably much more.
47
-------
FROM KRYTER'S SIR
ACTUAL DATA
POINTS
1 1 1
HlJS26dB OR GREATER
- (Average
-------
vO
C3
s
Ul
60
50
40
30
UJ t
XS20
o
111
10
-10
SPEECH (.5,1,2)
Baughn's Pre-Exposure Audiograms
of New Hire 18 yr. old Bnployees
10
20
30
40 50
AGE (YRS)
BAUGHN(78dBA EXPOSED
GROUP)
PUBLIC HEALTH
SURVEY (MALE
RT EARS)
KRYTER
ROBINSON (NON-NOISE EXPOSED)
60
70
80
Figure 18.
-------
(2) On Figure 18, Kryter's recommended presbyacusis
curves are plotted along with Robinson's. Note that Robinson's values are
below Kryter's. Yet Robinson has found that NIPTS for speech (0. 5, 1 and
2 kHz) essentially disappears for less than 75 dBA exposure. This does
not fit with Kryter's assumption that 75 dBA is causing a very significant
shift in hearing.
(3) Another inconsistency of Kryter's NIPTS predictions
can be seen if these values are compared to the actual hearing levels of
Baughn's workers. Figure 19 is such a comparison. Somehow Kryter has
taken Baughn's data and manipulated the data such that the predicted NIPTS
is the same as the total hearing loss of these individuals. Since hearing
loss consists of both NIPTS and aging, the only way to predict such a large
value of NIPTS, as I see it, is to predict that hearing will not change with
age. This is clearly wrong, of course, and even Kryter predicts 15 dB loss
from presbyacusis at age 65.
12. D-Versus A-Weighting of Frequency. At first glance, the use
of a D-weighting scale instead of an A-weighting might seem attractive. The
D-weighting added approximately a 10 dB penalty to the frequencies that are
more likely to cause NIPTS at the super-sensitive 4000 Hz audiometric
frequency. If one's goal is to protect the 3, 4 and 6 kHz frequencies equally
with the lower frequencies of 0. 5, 1 and 2 kHz, then perhaps the D-weighting
would be desirable. However, D-weighting also emphasizes the frequencies
above 5600 Hz by 6-9 dB, and thus would tend to give these high frequencies
more influence than they properly deserve. The very low frequencies are
also emphasized more. Thus protection of the speech frequencies of 0. 5,
1 and 2 kHz is slightly deemphasized. Qualitatively, the argument reduces
to this: if one desires that the risk of hearing loss should be equal for the
speech frequencies of 0. 5, 1 and 2 kHz and for the frequency of 4 kHz, then
the D-scale may be a slightly better approximation. If one is willing to
allow 5 dB more loss at 4 kHz than at the speech frequencies (0. 5, I and 2)
then the dBA is the better approximation. The general feeling among most
investigators is that the frequencies of 0. 5, I and 2 are aornewhat more
essential; therefore it is recommended that the A-scale be used for purposes
of hearing conservation. The D-scale can be used to predict the effects of
noise on hearing, but the proper adjustments must be made to provide the
same safety to the lower speech frequencies.
13. Duration of the Exposure.
(1) Less than 8 hours. The relationships between NIPTS and
SPL discussed up to this point have been based on an 8 hour working day ex-
posure. The auditory system can tolerate higher SPLs provided that the
exposure time is shorter (6). It is not entirely clear, but it is suspected
that the SPL should be reduced: if the ear is exposed to noise for durations
greater than 8 hours.
50
-------
70r
».
50
CD
T3
CO
40
Q
LU
o
LJ
oc
o_
20-
10
0
40
SPEECH3(.5,l,2KHz)
75 PERCENTILE
8hr. DAILY EXPOSURE
FOR 40 YEARS
HEARING LEVEL OF
THE EMPLOYEES OF
BAUGHN'S STUDY
(RE 1964 ISO-USE
SCALE AT LEFT AS
HEARING LEVEL)
KRYTER
PASSCHIER
VERMEER
BAUGHN
ROBINSON
A
D
50
60
90
100
no
70 80
SPL(dBA)
* KRYTER CONSIDERS THIS HEARING LEVEL (FROM TABLE III
[KRYTER (16)])
Figure 19
51
-------
The decision as how to relate SPL to duration in order to
obtain equally noxious noise exposure depends upon how the auditory damage
progresses with time. Three popular theories are equal energy (ISO stand-
ard for example), equal pressure (Kryter for example) or a compromise
between equal energy and equal pressure (NIOSH for example). The equal
energy rule predicts an equal hazard if the SPL is reduced 3 dB for each
doubling of duration (SPL varies inversely as 10 log t). The equal pressure
rule dictates that the SPL must be reduced 6 dB for each doubling of time
(SPL varies inversely as 20 log t). The NIOSH compromise suggests that the
SPL should be reduced by 5 dB for each doubling of time (SPL varies inverse-
ly as 16. 6 log t). The selection of one rule over another is not a trivial
question. For instance, considering the 8 hour exposure as the baseline,
equal pressure allows the permissible SPL for a one-minute exposure to be
27 dB higher than that allowed for equal energy.
There is a lack of unequivocal NIPTS data that would sug-
gest which rule to use. Therefore, equal TTS has been the only method for
assessing equal hazard. This is why a considerable effort was given in the
main criteria document to the relationship of TTS (via animal and human
studies) to NIPTS.
Experimental results have not yet completely clarified the
problem. Spieth and Trittipoe (7) indicate that the equal pressure rule pro-
vides equal TTS for high level, short duration exposures. Ward (8) has
found that equal energy best predicted an equal amount of TTS for chinchilla
during 4 exposure conditions.
Some sense can be made out of the apparent contradictions
if the CHABA curves are studied. Figure 20 is a replot of the CHABA
curves that relate equal TTS at various Sound Pressure Levels (SPL), dura-
tions and audiometric frequencies. All curves, only for thepurpoees of com-
parison, were related to the same SPL value for the 8 hour duration. Vari-
ous schemes for relating SPL to duration are then plotted. The results show
two main points. These are, (1) No simple function of log t best matches
the CHABA values for all time durations and (2) the selection of the function
used varies with the audiometric frequency that is to be protected. At this
time, it is not suggested that a function other than the log t be used since it
would effectively eliminate the ability to provide dosimeters and perhaps
unduly complicate the situation. The use of equal noxious TTS values is not
that firmly secure to warrant such refinements. Spieth and Trittipoe results
can be explained, however, since the durations with which they were con-
cerned were short. For exposures of 16 minutes and less, TTS at 4 kHz
does start to follow the equal pressure law.
Using Figure 20 as a basis, the decision as to which rule
to use reduces to which audiometric frequencies will be protected. If 4000
Hz is to be protected, then the equal energy rule will be the best approxima-
tion. If only the speech frequencies of 0. 5, 1 and 2 kHz are to be protected,
52
-------
01
150
cc 140
m 130
•a
CO
•O
2 120
laJ
-J
a
2
Z>
110
ioo
£ 95
o
uj 90
i 85
80
FROM PARRACK(UNPUBUSHED)
l r~1 ' "
A-WEIGHTED OVERALL &)UND LEVELS
.SHORT DURATION LEVELS EQUIVALENT TO 85 dBA(OR OTHER BASE
LEVELS) WHEN EXPOSURE DURATION IS 8HRS/DAY D
- EQUATION S>
IO
1*3
);
a
® LEQUIVsLBASE + IOI°9|o| ^EQUAL ENERGY RULE) O
= 8HRS/DAY
t = 8 HRS OR SHORTER DURATION AS SELECTED
LEQUIV=LBASE
a
Q250HzBAND
C4000HZ BAND
y 1000 Hz BAND
l t
I II I ii i I il
I
810 16 2432486496126 240
6080 128 256
8hr. 2hr. 30mia
RATIOS OF EXPOSURE DURATIONS
75min225sec.
Imin.
Figure 20.
-------
the NIOSH rule of 5 dB change in SPL for each doubling of time ia a very
good compromise. Either rule will overprotect for short time durations
and as such will add an additional safety factor into any standard for hearing
conservation. It should be noted that given an exposure level and duration,
Figure 20 can be used to directly predict the relation between such a condi-
tion and the dBA SPL of 8 hours duration that will cause the same amount of
TTS (or therefore NIPTS). The usefulness of such a figure is limited, how-
ever, as typically a total daily noise exposure does not occur in such a simple
manner. Therefore, some approximation scheme such as equal energy must
be used. Correction factors for such variables as the intermittency of the
noise are then required.
(2) Durations more than 8 hours. There is a noticeable lack
of actual NIPTS data on 24 hour exposure situations, therefore most of what
is known is based upon TTS data.
Smith et al. (9) exposed groups of men for 25 hours to a
70 Hz tone or a 300 Hz tone at 113 dB SPL, In general TTS ranged from
0 to 20 dB. Yuganov et al. (10) simulated a 24 hour space mission with an
ambient noise of about 75 dB (not enough details are given to convert to dBA
but a rough estimate would be 80 dBA) and found a TTS of 10 to 20 dB with
recovery in 1-2 hours. Mills (11) exposed himself to a 93 dB SPL signal
for about 30 hours and measured 25-27 dB TTS which required 2-4 days for
total recovery. Melnick (12) exposed subjects for 16 hours to the 300-600
Hz octave band at 95 dB SPL and found the maximum TTS to be 15-20 dB.
Recovery was complete within 20 hours past exposure. The Environmental
Protection Agency (EPA) is currently sponsoring research at the Aerospace
Medical Research Laboratory (AMRL) to further investigate this question
with human subjects. At this time, however, there is no evidence that the
effect of continuous noise is more noxious that what would be predicted by
use of the logarithm of time. In fact, several investigators (Mills, Melnick)
have suggested that TTS reaches limiting value that may occur between 16-
48 hours. Studies accomplished on animals (Mills and Talo (13); Melnick
(12); and Carder and Miller, (14) all predict that TTS will reach an asymptote
or a limiting value. Exposures have been for as long as three weeks to three
months, with the TTS reaching its limit within the first day (Carder and
Miller (14) and Mills (in Press)). What is not so clear is the question,
Does hearing damage stop when such a limiting value that is independent of
duration is reached? " Based on Carder and Miller's animal findings that
similar recovering curves occurred once the asymptotic values were reached,
the answer appears to be a qualified yes if the TTS is less than 20-30 dB.
Recent work not yet published (Mills (in Press)) indicates that for greater TTS
than 30 dB, such recovery may change with exposure time. Since TTS will
normally be less than 30 dB only for exposures leas than 85 dBA, this limit
will be considered valid only for exposures less than 85 dBA. The signifi-
cance of such a limit is that there may be little1 difference between a con-
tinuous lifetime exposure (24 hours exposure daily with no quiet periods) or
24'hour exposures with rest periods in between each exposure. Up to now, the
term 24 hour exposure has been used rather loosely to mean either case. We
54
-------
will continue to use it in this context for exposures less than 85 dBA with
the justification that the asymptotic behavior of TTS allows such an approxi-
mation to be made.
The equal energy rule would predict that the 24 hour ex-
posure should be 5 dB less than the 8 hour exposure. The NIOSH rule
would predict an 8 dB difference. The animal results of Carder and Miller
show better correlation with the NIOSH rule. The results of Melnick (1972)
on humans show that the equal energy hypothesis gives a better correlation
(it is even slightly conservative).
Preliminary results at AMRL have not shown the necessity
of deviating from the equal energy concept. Therefore a 5 dB reduction in
dBA is considered the best approximation at this time for extrapolating
8 hour data to 24 hours.
If the SPL is below the value which causes measurable
TTS at 8 hours, then there is no evidence that there will be measurable
TTS at 24 hours.
14. Estimation of the Accuracy in Relating NIPTS to Noise Exposure.
a) Underestimation Errors.
(1) Worst case of three methods.
Averaging the NIPTS predictions over the three methods
will provide in some cases lower NIPTS predictions than one method by itself.
In order to estimate the worst conceivable situation, the worst case values
are included in Table 14. This table already consists of the maximum NIPTS
expected for the . 9 percentile level during some part of a 40 year exposure
lifetime. Therefore selecting the highest predicted NIPTS value of the three
methods should set an approximate upper bound on the possible estimation
of NIPTS. That such an upper bound varies at the maximum byr only 4 dB
from the average provides additional confidence that any prediction errors
in the average data presented are not likely to underestimate the risk of
noise by more than 4 dB.
(2) Percentile estimates.
The estimation of NIPTS for some percentile has been
accomplished by subtracting the hearing level of that percentile of the non-
noise exposed group from the hearing level of the respective percentile of the
noise exposed group. The . 9 percentile group is thus that group whose
hearing level is worse than 90 percent of the population. If the . 9 percentile
point moves 10 dB because of noise exposure, then it is considered that the
. 9 percentile group had NIPTS of 10 dB. However, this 10 dB shift could
have been caused by some of the exposed ears shifting from a . 1 percentile
hearing level to the . 9 percentile hearing levels before the noise exposure,
then these exposed ears would have received a true NIPTS of 30 dB. Un-
doubtably there are a few individuals who have this occur. There is no way
55
-------
to account for such individual susceptability and it must be emphasized that
all estimates are for statistical groups of the population, not individuals.
Changes in the . 9 percentile hearing level is still considered the best indica-
tor of the true NIPTS not exceeded by 90 percent of the population, however,
for two reasons. First, the .9 percentile in a noise situation normally
does exhibit the greatest shift when exposed to noise. Apparently the people
that make up this group are those most sensitive to the noise exposure.
Second, changes in the . 9 percentile hearing level should be considered
more significant in that the hearing of this group is already worse than 90
percent of the population. A shift in this percentile point is thus liable to
have more significance than a shift in the . 1 percentile point.
It can be noted that the average NIPTS over 40 years
of exposure circumvents this problem. The errors introduced in saying that
90 percent of the population will have less NIPTS than some value X when
this NIPTS value was obtained by changes in the . 9 percentile hearing level
are difficult to estimate. If the changes in the . 9 percentile hearing level
are small, then one can reasonably expect that the error will be small.
But as stated earlier, a better way to look at this problem is to consider
that the . 9 percentile hearing level changes are the most important measure.
In this light, we will not unduly worry about this error.
b) Over estimation Errors.
(1) "Least effect" of three methods.
Averaging over the three methods will also provide
higher NIPTS predictions than some one method alone. Similiar to the
worst case discussed previously, the maximum difference between a single
method and the average is small. In fact this difference is < 2 dB for the
speech frequencies (either 1/3 (0. 5, 1, 2 kHz) or 1/4 (0. 5, 1, 2, 4 kHz)
and < 6 dB for 4000 Hz.
(2) Bias introduced in manipulation of the basic data.
Figure 21 shows how Passchier-Vermeer used the
data available to her for NIPTS at 4000 Hz. On this figure a curved line is
used to connect the data points represented. One criticism of her work is
that a linear least squares regression line could have been used just as well.
As can be seen in Figure 21, a linear regression line will predict that the
median NIPTS threshold is at 80 dBA, not 7 or 8 dB lower as would be ex-
pected by extrapolating Passchier-Vermeer's existing curve. It can only
be left up to individual judgement as to which approach is correct. Using
a linear regression line, the NIPTS {. 9 percentile) would be expected to be
0 dB for 75 dBA (8 hour) exposure and 8 dB.for an 80 dBA (8 hour) expo sure.
This compares to a NIPTS (. 9) of 10 dB for 75 dBA and 13. 8 for 80 dBA.
At 85 dBA either approach predicts the same amount of NIPTS. Therefore
the greatest possibility of error at the 4000 Hz audiometric frequency is
below 85 dBA. The average of the three methods produced 6 dB for 75 dBA,
56
-------
MEDIAN AND MEAN HEARING LOSS CAUSED BY EXPOSURE TO NOISE
FOR AT LEAST 10 YEARS, AS A FUNCTION OF SOUND LEVEL.
( From Passchier-Vermeer )
05
a
A\
A\
r
0
O
a
o
80
90 100
*» SOUND- LEVEL IN dB (A)
104
Figure 21.
57
-------
so the maximum error at 75 dBA is 6 dB. Likewise, it can be shown that at
80 dBA this possible error is 3 dB. Note that the magnitude of these errors
is the same as was obtained by looking at the "least effect" of the three
methods.
c) In summary, the 4 kHz and (. 9 percentile) data presented
in Table 17 can reasonable be considered accurate within a range of +4 dB
and -6 dB (or more simply Hh 5 dB) of the values given as long as the L
range under consideration is between 70 and 90 dBA.
B. Requirement for "Quiet"
Recent work by Ward (15) has shown that the quiet intervals bet-
ween high intensity noise-bursts must be below 60 dB SPL for the octave
band centered at 4000 Hz if recovery from Temporary Threshold Shift (TTS)
produced is to be independent of the quiet period SPL. Ward suggests
55 dB SPL as the point where the "effective quiet" might be. Assuming then
that (1) TTS recovery from a 90 dBA (8 hour) occupational exposure also
requires this same level of effective quiet for some part of the 16 hours
between the exposure the following day, and (2) total TTS recovery is impor-
tant in order to prevent TTS from becoming NIPTS, noise exposure should
be controlled in order to reasonably insure an effective quiet of 55 dB SPL
at the 4000 Hz octave band (approximately 62-65 dBA). The population
exposed to TTS producing sources (both occupational and non-occupational)
will be guaranteed by such control the availability of a quiet period of less
than 60 dBA. That such a quiet period is really required is not absolutely
proven, of course, but there is enough evidence to suggest at this time that
this approach is advisable.
in. SUMMARY
Selection of a permissible 24 hour exposure will be 5 dB below the
permissible 8 hour exposure SPL if equal energy is to be used. Table 19
summarizes the effects, as baaed on the 8 hour exposure, of exposures of
either 8 or 24 hours for different SPLs. The expected absolute error is
estimated to be well within 5 dB for the NIPTS values predicted. For Hear-
ing Risk, a fence of 25 dB (1964 ISO) is used. Baughn's and Robinson's
Hearing Risk values are averaged. For the 85 and 90 dBA (8 hour) exposure
conditions, the resulting average is within jh 3 percentage points of Hear-
ing Risk predicted by either method. For an 80 dBA condition, Robinson's
estimate (10 percent) and Bsughn's estimate (0 percent) were averaged to
'obtain 5 percent. While these values might seem rather divergent, it is
noteworthy that NIOSH predicted 3 percent for this level. The Hearing Risk
at 60 years of age was used. Hearing Risks at younger ages are less than
these values (see Tables 15 and 16).
58
-------
Table .19- Summary of effects expected for continuous noise
exposure of 8 hours to the levels stated.
Max NIPTS (.9)
NIPTS at 10 yr (.9)
Average NIPTS
Max Hearing Risk*
75 dBA (70 dBA for 24 hrs)
Max NIPTS (.9)
NIPTS at 10 yr (.9)
Average NIPTS
Max Hearing Risk*
Max NIPTS (.9)
NIPTS at 10 yr (.9)
Average NIPTS
Max Hearing Risk*
Max NIPTS (.9)
NIPTS at 10 yr (.9)
Average NIPTS
Max Hearing Risk*
25 dB ISO Fence
Speech (.5,
1 dB
) 0
0
N/A
1, 2) Speech (.5, 1, 2, 4)
2 dB
1
0
N/A
4K
6 dB
5
1
N/A
Speech ( . 5 ,
1 dB
) 1
0
5%
80 dBA (75 dBA for 24 hrs)
1, 2) Speech (.5, 1, 2, 4)
4 dB
3
1
N/A
4K
11 dB
9
4
N/A
Speech ( . 5 ,
4 dB
) 2
1
12%
85 dBA (80 dBA for 24 hrs)
1. 2) Speech (.5, 1, 2, 4)
7 dB
6
3
N/A
4K
19 dB
16
9
N/A
Speech ( . 5 ,
7 dB
) 4
3
22.3%
90 dBA (85 dBA for 24 hrs)
1, 2) Speech (.5, 1, 2, 4)
12 dB
9
6
N/A
4K
28 dB
24
15
N/A
59
-------
IV. CONCLUSIONS
The main purposes for preparing this report were twofold.
(1) The first purpose was to resolve the question of what and/or whose
data should be used to depict the relationship between loss of hearing sensi-
tivity and noise. The question was resolved by using three leading predictive
methodologies and averaging the results. This averaging has been criticized
by some as unscientific. The argument is that one should pick the most
scientifically sound method and use it alone. But the problem then remains
of how to select the single best method. Averaging the three methods avoid*
such a selection. But even more important, averaging the three methods
prevents the possibility of selecting the worst method. Therefore, the
averaging technique was considered as the best way to handle the problem
of data selection.
(2) The second purpose of this supplement was to discuss the method-
ology of Kryter (16). Criticism of Kryter's paper is provided by several
reviewers in the same issue of the Journal of the Acoustical Society of
America. At this time there are too many basic inconsistencies in Kryter's
method for his results to be included in this report.
60
-------
REFERENCES
1. Passchier-Vermeer, W. Hearing Loss Due to Exposure to Steady-State
Broadband Noise. Rept. No. 35, Institute for Public Health Eng., The
Netherlands, 1968.
2. Robinson, D.W. The Relationships Between Hearing Loss and Noise
Exposure. National Physical Laboratory Aero Report Ae32, England,
1968.
3. Baughn, W. L. Relation Between Daily Noise Exposure and Hearing
Loss as Based on the Evaluation of 6835 Industrial Noise Exposure
Cases. In Press as AMRL-TR-73-53, Aerospace Medical Research
Laboratory, Wright-Patterson AFB, Ohio.
4. Passchier-Vermeer, W. "Steady-State and Fluctuating Noise: Its
Effects on the Hearing of People" in Occupational Hearing Loss,
D.W. Robinson, Ed. Academic Press, N. Y., 1971.
5. Robinson, D.W. "Estimating the Risk of Hearing Loss due to Continu-
ous Noise!1. In Occupational Hearing Loss, D. W. Robinson, Ed, Aca-
demic Press, N. Y., 1971.
6. Kryter K. D. W. , J. D. Miller, and D. H. Eldridge. Hazardous Expo-
sure to Intermittent and Steady-State Noise. J. Ac oust. Soc. Am.,
Vol. 39, No. 3, p. 451, 1966.
7. Spieth, W. and W. J. Trittipoe. Intensity and Deviation of Noise Ex-
posure and Temporary Threshold Shifts. J. Acoust. Soc. Am. , 30,
710, 1958.
8. Ward, W. D. and D. A. Nelson. On the Equal-Energy Hypothesis
Relative to Damage-Risk Criteria in the Chinchilla. In Occupational
Hearing Loss, D.W. Robinson, Ed. Acad. Press, N. Y., 1971.
9. Smith, P. F., M. S. Harris, J. S. Russotti and C. K. Myers, Effects
of Exposure to Intense Low Frequency Tones on Hearing and Perform-
ance. Submarine Medical Research Laboratory, Naval Submarine
Medical Center Report No. 610, 1970.
10. Yuganov, Ye, M. et_al. Standards for Noise Levels in Cabins of Space-
craft During Long Duration Flights. (Tech. Transl. F-529, National
Aeronautics and Space Administration, Washington, D. C., 1969. )
61
-------
11. Mills, John H. , Roy W. Gengel, Charles S. Watson and James D. Miller,
"Temporary Changes of the Auditory System Due to Exposure to Noiae
For One or Two Days". . J. Acoust. Soc. Am. , Vol. 48, pp. 524-530,
1970.
12. Melnick, W. Investigation of Human Temporary Threshold Shift (TTS)
from Noise Exposure of 16 Hours Duration. (Presented at the Acous-
tical Society of America, December 1972 meeting.)
13. Mills, J. H. and S. A. Talo. Temporary Threshold Shifts Produced by
Exposure to High-Frequency Noise. Journal of Speech and Hearing
Research, Vol. 15, September 1972, pp. 624-631.
14. Carder, H. M. and J. D. Miller, "Temporary Threshold Shifts From
Prolonged Exposure to Noise, " Journal of Speech and Hearing Research,
Vol. 15, September 1972, pp. 603-623.
15. Ward, W. Dixon. The concept of "Effective Quiet, " presented at the
85th Meeting of the Acoustical Society of America, April 1973.
16. Kryter, K. W., "Impairment to Hearing from Exposure to Noise".
J. Acoust. Soc. Am., Vol. 53, No. 5, May 1973, pp. 1211-1234.
*U.S.Government Printing Office: 1973 - 758-426/78
62
------- |