TR-73-53
RELATION BETWEEN DAILY NOISE EXPOSURE
AND HEARING LOSS BASED ON THE EVALUATION
OF 6,835 INDUSTRIAL NOISE EXPOSURE CASES
W.L.BAUGHN.M.D.
GUIDE LAMP DIVISION
GENERAL MOTORS CORPORA T1ON
JUNE 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
-------
26950
DOCUMENT CONTROL DATA - R & D
(Security clattlflcelton at title, body ol abttracl and Indexing annotation mutt fee entered when the overall report I* claitllled)
I. ORIGINATING ACTIVITY (Corporate author)
Guide Lamp Division, General Motors Corporation,
Anderson, Indiana
2*. REPORT SECURITY CLASSIFICATION
UNCLASSIFIED
26. GROUP
N/A
3. REPORT TITLE
RELATION BETWEEN DAILY NOISE EXPOSURE AND HEARING LOSS BASED ON THE EVALUATION OF
6,835 INDUSTRIAL NOISE EXPOSURE CASES
4. DESCRIPTIVE NOTES (Type ol report and Inclusive date*)
Final Report
B. AU THOR(S) (Flrat name, middle Initial, Imal name)
W. L. Baughn, M.D.
e. REPORT DATE
. CONTRACT OR
b. PROJECT NO.
c. Work Unit
d.
June 1973
GRANT NO.
7230
72300001
7m. TOTAL NO. OF PAGES 76. NO. OF REFS
38 7
»«. ORIGINATOR'S REPORT NUMBER(S)
»6. OTHER REPORT NO(S) (Any other number* that may b* maligned
thit report)
AMRL-TR- 73-53
10. DISTRIBUTION STATEMENT
Approved for public release; distribution unlimited
II. SUPPLEMENTARY NOTES
12. SPONSORING MILITARY ACTIVITY
Aerospace Medical Research Laboratory,
Aerospace Medical Div., AF Systems Command,
Wright-Patterson AFB, Ohio 45433
13. ABSTRACT
The present study is designed to display the percent of a population exhibiting
greater than certain specified audiometric hearing levels as a function of specified
exposure levels and duration of exposures to those levels. Audiometric data from
6,835 employees of an industrial plant were taken during the period from 1960 through
1965. The employees were selected only on the criterion that their noise exposures
were reasonably well known. Hearing levels for each of three exposure conditions
(78, 86, and 92 dBA) were obtained for the speech {0.5, 1, and 2 kHz) and the 4 kHz
audiometric frequencies. The data are smoothed and hearing risk tables are presented.
Key Words: hearing risk
hearing damagehearing loss
presbyacusis
NIPTS (noise induced permanent threshold shjft),
DD
FORM
t NOV ««
1473
Security Classification
-------
FOREWORD
During a collaborative effort for the American National Standards Institute,
Working Group 46 on Hearing Conservation, this technical report was completed in
1968 by Dr. William L. Baughn of the Guide Lamp Division of the General Motors
Corporation and transmitted in letter form to Dr. H. 0. Parrack (now deceased) of
this Laboratory. The scientific information in the "letter" has been widely used
as the basis for selecting criteria limits of noise exposure for purposes of hear-
ing conservation. Among the most well-known uses are its basis for the revisions
of both APR 160-3 (APR 161-33) on Hazardous Noise Exposure and for the International
Standards Organization (ISO) Recommendation R1999, "Assessment of Occupational Noise
Exposure for Hearing Conservation Purposes."
The Biodynamics and Bionics Division of Aerospace Medical Research Laboratory
is currently developing a criteria document, "A Scientific Basis for Limiting Noise
Exposure for Purposes of Hearing Conservation," under an Interagency Agreement with
the Environmental Protection Agency. The University of Dayton Research Institute
is providing technical support for this effort under contract F33615-72-C-1402.
Dr. Baughn is a prime consultant to the University of Dayton Research Institute and
his technical information serves as important background information for the criteria
document. However, it was considered mandatory that any material contained in the
criteria document be available in the published literature. The publication of
Dr. Baughn's report, in addition to serving as the basis for the new APR 161-33,
also satisfies the technical information availability requirement for the criteria
document and will allow it to be successfully completed.
The Aerospace Medical Research Laboratory and the Environmental Protection
Agency greatly appreciate Dr. Baughn's and his company's collaboration in making this
extremely valuable technical information available for publication in its complete
form.
This technical report has been reviewed and is approved.
HENNING E. G. von GIERKE, Dr. Ing.
Director
Biodynamics and Bionics Division
Aerospace Medical Research Laboratory
-------
TABLE OF CONTENTS
Page
THE DATA ]
THE NOISE STUDIES 1
METHOD 4
THE PERCENT OF POPULATION TABLES 9
ADDENDUM 30
REFERENCES 33
LIST OF TABLES
Table
1 Mean percent of time spent at each dBA level by subjects in each
exposure group 3
2 Age span and exposure data 5
3 Data from graphic smoothing procedure (3F/3) 8
4 Distribution Coefficients 9
5 Mathematically smoothed decile points (3F/3) 10
6a Interpolated and extrapolated from field (3F/3) 16
6b Extrapolated from field (3F/3) 17
7 Percent of population displaying more than 15 dB hearing level
averaged at 500, 1000, 2000 Hertz (ASA 1951) as a function of age,
years of exposure (assuming years of exposure = age 18) and
exposure level in dBA 28
8 Percent of population exhibiting more than 15 dB (ASA 1951)
hearing level averaged at 500, 1000, and 2000 Hertz as a function
of age, years of exposure (assuming years of exposure = age 18)
and exposure level in dBA (adjusted) 29
9 Mathema/tically smoothed decile points (4 kHz) 31
10 Percent of population exhibiting more than 40 dB hearing level
at 4 kHz as a function of age and exposure level in dBA 32
IV
-------
LIST OF ILLUSTRATIONS
Figure Page
1 Typical distribution of hearing level for a specific age and
exposure group. This distribution was for a group ranging in
age from 60 to 65 years and exposure of 86 dBA 6
2 Typical terminal distribution plotted on log-normal paper.
This distribution was for a group ranging in age from 60 to
65 years and exposure of 86 dBA 7
3 Plot of all median values on rectilinear grid. Note that the
function of hearing level versus exposure is not linear .... 12
4 Plot of data from figure 3 on a log grid. Note that straight
lines adequately approximate the relationship between hearing
level and exposure 13
5 Plot of median values adjusted so (1) the maximum hearing
level is 65 dB at the 130 dBA SPL and (2) the nonexposed
median is anchored by 18-year old new hire males 15
6 Sample plot of the data from table 6a. Exposure level is
105 dBA 18
7 Percent of the population with more than a 5 dB audiometric
hearing level (re 1951 ASA) for the speech frequencies
(0.5, 1, and 2 kHz) 19
8 Percent of the population with more than a 10 dB audiometric
hearing level (re 1951 ASA) for the speech frequencies
(0.5, 1, and 2 kHz) 20
9 Percent of the population with more than a 15 dB audiometric
hearing level (re 1951 ASA) for the speech frequencies
(0.5, 1, and 2 kHz) 21
10 Percent of the population with more than a 25 dB audiometric
hearing level (re 1951 ASA) for the speech frequencies
(0.5, 1, and 2 kHz) 22
11 Percent of the population with more than a 40 dB audiometric
hearing level (re 1951 ASA) for the speech frequencies
(0.5, 1, and 2 kHz) 23
12 Idealized graph drawn to represent a probable pattern from
birth to death of the percent of the population with more
than 15 dB audiometric hearing level (re 1951 ASA) for the
speech frequencies 0.5, 1, and 2 kHz 25
13 Median 3F/3 hearing loss by age for five well-known population
studies 26
-------
RELATION BETWEEN DAILY NOISE EXPOSURE AND HEARING LOSS BASED ON THE EVALUATION
OF
6,835 INDUSTRIAL NOISE EXPOSURE CASES
The present study is designed to display the percent of a population exhibiting
greater than certain specified audiometric hearing levels as a function of
specified exposure levels and duration of exposure to those levels.
THE DATA
The audiometric data dealt with in this study consists of 6,835 audiograms
of employees in a midwestern industrial plant. This is a little more than
one third of all audiograms taken from this population over the six year
period from 1960 through 1965. About 1WO thirds of the available audiograms
from this period were eliminated from the study because the subjects had
significant unknown or mixed exposures.
Tine audiometric test environment conformed fully with the specifications
of the American Standards Association. The audiometers were Maico H-l models
and were checked against normal experienced ears before each clay's use, and were
calibrated in the laboratory of the Maico Company periodically. They were
never found to be out of the acceptable calibration range.
The same two trained and experienced audiometrists took all the audiograms
used in the study. Prior to the beginning date they had done more than 25,000
audiograms over a period of eight years, all of which had been submitted to
the laboratory of the Subcommittee on Conservation of Hearing, of the American
Academy of Ophthalmology and Otolaryngology in Los Angeles where samples were
subjected to consistency tests and mathematical analysis by Dr. Ann Summerfield.
Similar tests applied to the data used in this study have confirmed its self
consistency.
THE NOISE STUDIES
The noise studies used in this work consist of nearly 15,500 detailed sound
analyses of work-location exposures covering a period of 14 years. Inter-
views and studies of work records, and comparative testing of older with
more recent equipment and processes allowed extension in some subjects back
40 years or more with reasonable confidence that their exposures were known
with sufficient precision to allow their inclusion in the study.
-------
While the noise analyses included octave bands, A, B, and C weightings, along
with SIL and other computed indices in both slow and fast inertial dynamics, and
all these repeated with the General Radio Impact meter, only the A weighting
and slow meter dynamics reading was used in this study. We, and we believe
most others working in the field, are satisfied that the A - slow reading
provides an adequately precise index to the long-term effect of noise on the
hearing function and present evidence is that it more accurately predicts the
effects of noise on hearing than any other available single-number index.
All noise analyses were done with General Radio 1551-B noise level meter and
1550-A octave band analyzer conforming to the applicable A. S. A. apecifica-
tions. All were done by engineers or engineering students under competent
supervision and data were tested for consistency. Readings for each noise
field used in the analysis were logarithmically averaged over the several
noise measurements made on that particular exposure over the years.
Hie three exposure levels used are 78 dBA, 86 dBA, and 92 dBA. It was about
these levels that actual exposures in the environment under study tended to
cluster most closely, thus yielding the largest population samples with the
narrowest exposure distributions. Approximately five thousand "A" - slow
averaged readings were used in assigning exposure levels. Those studies
show that individuals assigned 78 dBA exposure spent 651 of their working
time in exposures no greater than 80 and no less than 74 dBA, 90% no greater
than 81 nor less than 66 dBA. The remaining 10% may have occasionally been
as high as 82 dBA and as low as 42 dBA.
the group assigned 86 dBA spent 651 of their work time at 86 - 2 dBA, 801 -
4 dBA, and not more than 5% at above 92 and below 78 dBA combined.
The group assigned 92 dBA spent 65% of their work time at 92 * 3 dBA, 871 at
92 ± 5 dBA, and not more than 5% at above 100 and below 84 dBA combined. (Table 1)
The noise in all three groups was generally relatively rich in low frequency
components, which is to say it conformed roughly with the inverse of the "A"
weighting characteristic of the noise meter. The 78 dB intensity noises tend
to be located principally in crib, storage, shipping, and office spaces. The
86 dB noises tend to be principally associated with light assembly operations
on thin metal, plastic, wood, and glass. The 92 dB exposures arise largely
from press operations, grinding, and heavier assembly operations. Some
impulsive characteristic is evident, particularly in the 86 and 92 dB exposures,
but no impact sources such as riveting guns or impact wrenches are represented.
The population under study is composed of the employees of a midwestern
industrial plant producing automobile parts. The factory is under one roof
and has occupied its present site for more than 40 years. The employees are
drawn from the surrounding agriculatural-industrial community of about 100,000
population. The work force is very stable with relatively light turnover,
particularly in its older members, providing a high continuity of employment
both invocation and job content. A number having remained in the same work
40 years^ and more. The age range is from 18 to 68 years.
-------
TABLE 1
MEAN PERCENT OF TIME SPENT AT EACH dBA LEVEL BY. SUBJECTS IN EACH EXPOSURE GROUP
dBA 2i J!6. £
65 - 66 .75
66 - 67 1.
67- 68 1.
68 - 69 1.
69 - 70 1.
70 - 71 1.
71 - 72 2.
72 - 73 2.
73 - 74 3.
74 - 75 3.
75 - 76 4.
76 - 77 10.
77 - 78 12. .5
78 - 79 16. 2.
79 - 80 20. 2.
80 - 81 12. 3.
81 - 82 5. 4.
82 - 83 2. 6.
83 - 84 1. 8. .5
84 - 85 .5 10. 2.
85 - 86 13. 2.
86 - 87 14. 2.5
87 - 88 14. 3.5
88 - 89 10. 4.5
89 - 90 8. 6.0
90 - 91 5. 10.0
91 - 92 3. 12.0
92-93 14.0
93 - 94 12.0
94 - 95 10.0
95 - 96 6.0
96 - 97 5.0
97 - 98 3.5
98 - 99 2.0
99 - 100 1.0
100 - 101
-------
Chronological age is used as the uniform measure of exposure duration. Attempts
have been made to modify this measure to accommodate rest periods within the
work day, absences due to lay-offs, vacations, illnesses, etc. The fact remains
that the average employee in this population enters the work force at age 18,
has an average number of rest periods, illnesses, etc, and ends his industrial
employment at age 65 or 68 with an average duration of exposure to industrial
noise directly related to his age. Neither philosophy nor mathematics has given
us any reason to believe another index to duration of exposure is in any way
superior.
Subjects with seriously mixed exposures, or unknown exposures, were categorically
excluded from the study. No other selection was made. Changes in hearing level
reflect all causes of such change.
This brings into focus a criticism of our work which has been leveled since
our first publication of it in 1966, This is relative to our decision not to
exclude on the basis of historical or objective anatomical ear defects. Had
we excluded on the basis of possibly significant history and possibly signifi-
cant anatomical defects, our numbers would have suffered seriously, and con-
sequently our statistical confidence levels. There comes a time when further
exclusion is counter-productive. Our own work, and that of others, has
indicated that quite small changes in hearing level numbers follow even massive
exclusion based on history and physical examination.
Following the exclusions from the study detailed above, we were left with 6,835
audiograms matched with exposure history in terms of three exposure groups
identified as 78 dBA, 86 dBA, and 92 dBA.
The criteria for defining those members of the population who have suffered
an "impairment" of hearing are based on the thesis that impairment shall be
for the understanding of spoken English in sentence form. The American Academy
of Ophthalmology and Otolaryngology has determined, and the American Medical
Association has concurred, that such impairment begins when the arithmetic
mean of the audiometric hearing levels at 500, 1000, and 2000 cycles per second
exceeds 15 decibels (A. S. A. 1951), or 25 decibels (I. S. 0. 1964) and that
impairment increases at the rate of 1 1/2% for each decibel in excess of 15
(A. S. A.) or 25 ( I. S. 0.) until a maximum of 1001 has been reached at
82 decibels (A. S. A.) or 92 decibels (I. S. 0.).
We have accepted this 15 dB (A. S. A.) as our criterion for beginning impair-
ment. When we identify a certain percent of the population under study as
having a mean hearing level (at the speech frequencies) of more than 15 dB
(A. S. A.), it means that this percent of the population has at least a be-
ginning calculable impairment.
METHOD
(All audiograms were done prior to the end of 1965 and all were done to
A. S. A. 1951 standard audiometric zero calibration. All audiometric,
exposure, and identification data were entered on punched cards and all
sorting and calculations were done by electronic data processing equipment.)
-------
The population under study, after having been stripped of members with mixed
and unknown exposures, was divided into three exposure groups. There were
852 members in the group assigned exposure 78 dBA, 5,150 members of the group
assigned 86 dBA, and 833 members, of the group assigned 92 dBA.
Each exposure group was broken into eight age groups. Each age group covers
a span of six years, the youngest group encompassing ages 18 through 23 years
inclusive and the oldest age group 60 through 65 years inclusive.
TABLE 2. AGE SPAN AND EXPOSURE DATA
Age Group
Number
1
2
3
4
5
6
7
8
Age Span
18 -23
24 -29
30 -35
36 -41
42 -47
48 -53
54 -59
60 -65
Exposure I
78 dBA
N « 10
68
144
148
183
159
95
45
Exposure II
86 dBA
N = 107
476
544
860
1041
1070
723
329
Exposure III
92 dBA
N '- 4
39
76
124
189
197
127
77
Total
121
583
764
1132
1413
1426
145
451
852 5150 833 6835
E. D. P. Cards are punched for each subject carrying the exposure level, age
group, and audiometric data. Audiometric hearing levels at 500, 1000, and
2000 Hz are added for each subject and the sum divided by three. These three
frequency mean hearing levels are printed out as an array by increasing hearing
levels. A break is made at each change of hearing level (each 1 z/3 dB H. L.
for the three frequency average) and the percent of that age-exposure group
lying below this change is noted.
Now the percent-of-the-group below is plotted on some type of distribution
paper (since there are elements of several kinds of distribution present, it
doesn't make any real difference which form of grid we use.) We have chosen
to do the primary graphic interpolation on normal distribution paper (Fig.l
is an example.) Terminal distributions are done, where necessary, on log-normal
paper, since the extremes of the distributions, particularly in higher age groups
tend to be log-normal (Fig. 2.)
Whatever method of interpolative smoothing is used yields a series of crossing
points on the distribution graph (or by formula) as intersections between the
regression line (representing hearing level) and percentage distribution line on
the graph. We have chosen to select the nine inter-decile points for further
work. Quartile or centile points could be chosen, but we feel the deciles give
sufficiently high resolution to exhaust the quality of the data and provide
sufficiently smooth curves for our later work. Now we tabulate all the inter-
decile points from all 24 graphs, Table 3.
Plotting and least squares smoothing is all that is required to complete the
work graphs for a procedure dealing only vdth data within the experimental field,
and was in fact what was done for the initial work on these data which was
reported in 1966.l
-------
O.b
1
2
5
1 0
20
30
40
50
60
70
80
90
95
98
99
-H
/
/ 1
/
/
/
/
/
/
/
7
/
/
J
/
\
S
*£
S
s
/
/
X
/.
xx
i*
^»
x^»
0-50 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 6
99
98
95
90
80
70
60
50
40
30
20
10
5
2
1
0.5
0
Figure 1. Typical distribution of hearing level for a specific age and exposure group.
This distribution was for a group ranging in age from 60 to 65 years and
exposure of 86 dBA.
-------
y y.b
99.8
99.5
99
98
95
90
UJ 80
y 70
z 60
5 50
£ 40
< 30
r
H 20
CO
y 10
5
2
1
0.5
02
O.I
^
/
./
/
&
s
/
*
/
/
/
10
100
HEARING LEVEL IN dB (3F/,)
Figure 2. Typical terminal distribution plotted on log-normal paper. This distribute
was for a group ranging in age from 60 to 65 years and exposure of 86 dBA.
-------
TABLE 3
DATA FRCM GRAPHIC SMOOTHING PROCEDURE
(3F/3)
Int.
Dec.
Points
1
2
3
4
5
6
7
8
9
Int.
Dec.
Points
1
2
3
4
5
6
7
8
9
AGE 18 -
78
-.2
1.3
3.0
3.9
4.4
5.2
6.1
7.1
8.0
86
1.3
3.1
4.1
5.0
5.7
6.4
7.1
8.0
9.7
AGE 42
78
1.0
3.0
4.5
5.8
6.9
8.2
9.7
11.3
14.2
86
3.2
5.0
6.6
7.9
9.2
10.9
12.9
15.5
20.6
23
92
3.0
3.9
4.9
5.8
6.3
7.6
8.9
10.3
13.0
- 47
92
6.6
7*. 6
9.0
10.4
11.8
13.2
15.0
17.3
21.6
AGE 24
78
-1.0
1.2
2.5
3.8
4.9
5.9
7.0
8.4
10.6
86
1.3
3.2
4.6
5.6
6.4
7.2
8.4
9.8
11.9
AGE 48
78
2.0
4.5
6.3
7.9
9.4
11.1
13.1
15.6
19.8
86
3.6
5,6
7.3
8.8
10.2
12.1
14.2
17.3
23.3
- 29
92
5.0
5.2
5.9
6.3
7.3
8.5
10.0
12.6
18.0
- 53
92
6.2
8.3
10.0
11.5
13.2
15.2
17.8
21.5
29.4
AGE 30
78 86
0.0 1.6
2.1 3.5
3.2 4.9
4.2 6.1
5.2 7.2
6.2 8.5
7.3 10.0
8.6 11.9
10.5 15.2
AGE 54
78 86
3.3 6.5
5.6 7.9
7.1 9.8
8.5 11.5
10.0 13.2
11.6 15.2
13.6 17.7
16.5 20.9
21.7 27.2
- 35
92
4.5
6.8
7.9
8.8
9.4
10.6
12.1
14.1
18.2
- 59
92
7.2
10.0
12.1
14.1
16.1
18.3
21.1
24.7
32.0
AGE 36 -
78
.5
2.4
3.6
4.6
5.6
6.6
7.9
9.3
12.0
86
3.0
4.7
6.0
7.0
8.2
9.5
11.1
13.3
18.5
AGE 60 -
78
7.3
9.4
11.1
12.8
14.4
16.3
18.4
21.5
27.6
86
7.4
10.5
13.0
15.2
17.5
20.1
23.4
28.2
38.0
41
92
4.3
6.3
7.9
9.2
10.7
12.3
14.2
16.9
22.1
65
92
9.6
12.4
14.8
16.9
19.2
21.9
25,0
30.0
39.9
8
-------
For all regressions relating exposure to hearing levels by age group, we
use a simple logarithmic relationship:
Log1Q ii. L. « a + b Exposure
For all regressions relating time to hearing levels by exposure group, we
use a cubic parabola:
II. L. - a + b Time + c Time2 + d Time3
Working from the "interpolated raw data" table, we fitted such a cubic curve
to the medians and to each interdecile set of points. By comparing and smoothing
the coefficients we rationalized the interdecile intervals. For the three
frequency mean this worked out to a single set of ratios xvith evidence of well
under two decibels probable error for even the most extreme fields, Tables 4 and 5.
Final smoothing of the 216 median and interdecile points is accomplished by
use of a statistical method known as "Joint Regression Surfaces." I shall not
describe the technique of this method which smoothes associated data in three
dimensions simultaneously. It is ideally suited to our problem. This method
does not appear in many statistics texts, so we suggest specifically:
Methods of Correlation and Regression Analysis
Ezekiel and Fox
Third Edition
John Wiley 5 Sons, New York
(Chapter 21)
TABLE 4. DISTRIBUTION COEFFICIENTS
Interdecile #1 = .67 (Med. + 10) - 10
ii 2 - .77 " "
M 3 <84 ii ii
4 » ,91 ii i.
" 5 Median
" 6-1.16 Med.
" 7 = 1.35 "
" 8 - 1.61 "
" 9 = 2,10 "
We have now completed the interpolation process and have allowed 216 raw
data points (3 exposure groups x 8 age groups x 9 interdecile points) to
arrange themselves by mutual push and pull into a most probable arrangement
in space. The fact that we had to subjectively choose specific curves for
them to follow prohibits us from saying the most probable arrangement.
Indeed, we may be sure it is not the most probable arrangement. For example,
we know that our median to interdecile ratios tend to slightly understate the
interval between median and first decile at very low ages and exposures (but
not by more than .3 dB at 18 years and 78 dB), and understates the median to
ninth decile interval at very high ages and exposures (but not more than 1.7 dB
at 65 years and 92 dB.)
To proceed: we now have families of deciles which reflect as accurately as is
-------
TABLE 5
MATHEMATICALLY SMOOTHED DECILE POiriTS
Int.
Dec.
Points
1
2
3
4
5
6
7
8
9
Int.
Dec.
Points
1
2
3
4
5
6
7
8
9
AGE 18 -23
78
- .2
1.24
2.26
3.29
4.6
5.34
6.2
7.41
9.66
86"
.39
1.94
3.02
4.11
5.5
6.38
7.4
8.86
11.55
AGE 42 -
78
1.39
3.09
4.28
5.47
7.0
8.12
9.45
11.27
14.7
86
2.80
4.71
6.04
7.38
9.1
10.56
12.29
14.65
19.11
92
.72
2.32
3.44
4.56
6.0
6.96
8.10
9.66
12.6
47
92
4.67
6.86
8.40
9.93
11.9
13.8
16.07
19.16
24.99
AGE
78
-.02
1.47
2.52
3.56
4.9
5.68
6.62
7.89
10.3
AGE
78
2.33
4,17
5.46
6.74
8.4
9.74
11.34
13.52
17.64
(3F/3)
24 - 29 AGE 30 - 35
86 92 78 86 92
1.12 2.06 .25 1.59 3.07
2.78 3.86 1.78 3.32 5.02
3.94 5.12 2.85 4.53 6.38
5.11 6.38 3.92 5.74 7.75
6.6 8.0 5.3 7.3 9.5
7.66 9.28 6.15 8.47 11.02
8.91 10.8 7.16 9.86 12.83
10.63 12.88 8.53 11.75 15.3
13.86 16.8 11.13 15.33 19.95
48 - 53 AGE 54 - 59
86 92 78 86 92
3.74 5.81 3.74 5.48 7.49
5.79 8.17 5.79 7.79 10.1
7<22 9.82 7.22 9.40 11.9
8.66 11.48 8.66 11.02 13.75
10.5 13.6 10.5 13.1 16.1
12.18 15.78 12.18 15.20 18.68
14.18 18.36 14.18 17.69 21.74
16.91 21.89 16.91 21*1' 25.9
22.05 28.56 22.05 27.51 33.8
AGE 36 -41
78
2
3
4
6
6
8
9
12
6
8
10
12
14
16
19
23
30
.72
.32
.44
.56
.0
.96
.10
.66
.6
AGE
78
.42
.87
.58
.30
.5
.82
.58
.35
.45
2
3
5
6
8
9
10
13
17
86
.13
.94
.20
.47
.1
.4
.9
.04
.01
92
3.
5.
7.
8.
10.
12.
14.
17.
22.
87
93
39
84
7
4
45
23
47
60 -65
8F 92
8
10
12
14
17
19
23
27
36
.22
.9
.85
.75
.2
.95
.22
.69
.12
9
12
14
16
19
22
26
31
41
.83
.79
.86
.94
.6
.74
.46
.56
.2
10
-------
practical the relationships existing between age, exposure, and hearing
level. The exposure field represented is "real" and extends from 78 dBA to
92 dBA. We wish to extend these limits to 115 dBA on the upside. The
downside doesn't bother us, the 80 dBA "starting point" is an interpolation
within the experiential field and is as accurate as anything else in this
field. We could simply calculate the extended points from our formulae and
hope for the best. To extrapolate a 14 unit field (78 to 92 dBA) almost
25 units upward, especially with complex formulae, would be dangerous.
However, it happens that we can establish one or two acceptable "anchor
points" in the extrapolated field .which will make it considerably less
hazardous.
We take all the median points from our known field (24 points, three exposure
points for each of 8 age groups) and plot them on a rectilinear grid and study
them. We see that the function is not linear on this grid and that the indicated
curve is concave upward in all cases except that of the youngest age group. A
laying-on of templates (Fig. 3) suggests a logarithmic relationship as likely.
A test has shown that the liklihood of a systematic error of -2 dBA in noise
measurement limited to the 86 dBA level and varying rationally by age groups
is less that 1/100, so we must accept the curvilinearity as real.
We now replot the data on a log. grid and strike straight lines as nearly as
possible through the points (Fig. 4 .) This process brings to light three
important points about what is now a rather neat family of regression lines:
1. There is a convergence to a crossing point centering on about 130 dBA
at about 47 dB H. L. 3F/3. This involves age groups 3, 4, 5, 6, 7, and 8
(ages 30 to 65).
2. There is a crossing point at about 71 dBA/4.8 dB H. L. 3F/3 for age
groups 1 and 2 (ages 18 to 29).
3. There is a slight unresolved curvilinearity in seven of the eight regression
lines even on the log, grid.
Regarding these anomalies, we reasoned as follows:
Some kind of a crossing point at the upper end of the graph is to be expected
as a matter of limits. After all, only so much hearing exists to be lost and
only so much biologically effective noise exposure is possible. As to this
latter, we know that as exposure levels increase above about 125 dBA, a marked
change takes place in the character of the ear's response to the increasing
level. Non-linear distortion rapidly increases, pain develops, increases and
changes in character. We believe that this area of disintegrating auditory
response at 125 - 140 dBA exposure represents a limit to the rational relation-
ship between exposure intensity, time, and progressive degradation of cochlear
function.
We do not believe that the location of our crossing point between 125 and
135 dBA is a matter of chance or coincidence. We will place an anchor at
130 dBA, the center of this range.
The other coordinate of this upper crossing point is at about 47 dB H. L.
3F/3. This doesn't yield so quickly to reflection on the known facts. The
11
-------
75 78 80
8582
90 92 95
dB(A)
100
105
110
115
Figure 3. Plot of all median values on rectilinear grid.
hearing level versus exposure is not linear.
Note that the function of
-------
200
00
o
LJ
LJ
-I
Z
cr
LJ
X
70 75 80 85
95 100 105 110
EXPOSURE dBA
115 120 125 130 135 140
Figure 4. Plot of data from figure 3 on a log grid. Note that straight lines
adequately approximate the relationship between hearing level and exposure.
-------
initial implication is that regardless of age or exposure intensity, the
median hearing level cannot exceed 47 dB. We know by experience that this
is not true. We have in our files, for example, some excellent audiograms
secured from Stewart Nash a number of years ago indicating median levels of
65 dB for extreme exposures. Glorig and his co-workers have demonstrated
that about 65 dB H. L. is a limit from noise exposure.*
If we look now at the third anomaly this comes clear. Age groups 4, 5, 6,
7, and 8 (ages 36-65 inclusive) show a residual curvature, concave upward -
that is, they reveal a slightly more than logarithmic relationship between
exposure and H. L. Age group3 (W^ 35 years) is linear or precisely logar-
ithmic, and age groups 1 and 2 reveal a downward concavity or something less
than a pure log. relationship. If we now carefully lay on log-curve templates
(Fig. 5) we will find that the crossing point at 130 dBA appears to be at
about 65 dB H. L. for all groups above the age of 36 years (18 years exposure.)
With less than 18 years exposure, there is a progressively lower terminal level
regardless of exposure level.
Note that we have selected 65 dB H. L. as the Y limit but that this precise
point is not necessary. If we chose 75 or even 80 dB H. L. as the limit it
would change our extrapolations very little at even 115 dBA exposure.
As soon as the indicated curvilinearity is reestablished the crossing point
at the lower end of the graph (low exposure end) disappears. However, we
were not happy with the low age segments of our median regressions and parti-
cularly with the compression taking place between ages at 78 dBA exposure.
We were anchoring our curves to age group 1 and this is on the face of it
incorrect. The mean subject in age group 1 already has three years exposure
(average of 6 year group, 18 through 23) and three years is a sizable exposure
period especially at high exposure levels. The origin of our curves should
be at a precise point where all subjects have identical (or average identical)
exposures. One such point does exist and it is available. The 18 year old
new hire males employed during the time the other data were being collected.
We determined the pre-employment H. L. 3F/3 for this group and used that
(2.4 dB H. L. 3F/3) for our new X - Y anchor for all medians. This changed
the curve significantly, particularly for low age and high exposure.
All these changes are reflected on the graph of Fig. 5. Having picked off
the median point for each age group at each exposure level from 80 dBA to
115 dBA in 5 dBA steps from this master graph, we enter them in Tables 6a
and 6b and plot them as decile families on a series of linear grids of
which Fig. 6 is an example. Now we lay on an age scale across a given
decile family at a given H, L. "fence" and plot on another linear grid, laid
out by years of age on the abscissa, and percent of population on the ordinate,
the interdecile intersection points with this "fence." Least distance curves
are struck through these points by use of a Copenhagen ship curve and the
final product of our procedure appears. (Figs. 7 through 11.)
A first glance at the finished % of population graphs may be disconcerting.
We have, in fact, two deleterious effects operating independently in their
attack on audition. It is the interaction of these two forces which produces
the complex progression in what one might expect to be steady progress toward
extinction of the hearing function. In high exposures the noise induced effect
Personal communication
14
-------
200
130
EXPOSURE (dBA)
Figure 5.
Plot of median values adjusted so (1) the maximum hearing level is 65 dB
at the 130 dBA SPL and (2) the nonexposed median is anchored by 18-year
old new hire males.
-------
TABLE 6a
INTERPOLATED AND EXTRAPOLATED FRCM FIELD
(3F/3)
Int. AGE 18 - 25 AGE 24 - 29 AGE 50 - 35 AGE 36 - 41
Dec.
Points 80 85 90 95 80 85 90 95 80 BS" 90 95 80 85 90 95
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 4.7
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 3.8 5.1 6.9
3 1.5 2.1 2.4 3.0 2.8 3.8 4.6 6.0 3.3 4.3 5.6 7.5 3.8 5.0 6.5 8.4
4 2.5 3.1 3.5 4.1 3.8 4.9 5.8 7.3 4.4 5.5 6.9 8.9 4.9 6.3 7.8 9.9
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 11.9
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 13.8
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 16.1
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 19.2
9 7.8 9.2 10.1 11.6 10.9 13.4 15.5 18.9 12.2 14.7 18.1 22.7 13.4 16.6 20.2 25.0
AGE,42 - 47 AGE 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 10.8
2 3.3 4.6 5.9 7.6 4.6 5.7 7.2 9.2 6.2 7.6 9.3 11.2 9.3 10.7 12.1 13.9
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 16.0
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 18.2
5 7.3 8.9 10.7 12.9 8.9 10.4 12.3 14.9 11.1 12.9 15.0 17.5 15.0 16.9 18.7 21.0
6 8.5 10.3 12.4 15.0 10.4 12.1 14.3 17.3 12.9 15.0 17.4 20.3 17.4 19.6 21.7 24.4
7 9.9 12.0 14.4 17.4 12.0 14.0 16.6 20.1 15.0 17.4 20.3 23.6 20.3 22.8 25.2 28.4
8 11.8 14.3 17.2 20.8 14.3 16.7 19.8 24.0 17.9 20.8 24.2 28.2 24.2 27.2 30.1 33.8
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 44.1
-------
TABLE 6b
EXTRAPOLATED FROM FIELD
Int. AGE 18 - 23
Dec.
Points 100 105 110 115
AGE 24 - 29
(3F/3)
100 105 110 115
AGE 30 - 55
100 105 110 115
AGE 36 - 41
100 105 110 115
1 .9 1.4 2.0 2.7
2 2.5 3.1 3.8 4.6
3 3.6 4.3 5.0 6.0
4 4.7 5.5 6.3 7;3
5 6.2 7.0 7.9 9.0
6 7.2 8.1 9.2 10.4
7 8.4 9.5 10.7 12.2
8 10.0 11.3 12.7 14.5
9 13.0 14.7 16.6 18.9
AGE 42 - 47
1 7.6 10.2 14.2 18.2
2 10.2 13.3 17.8 22.4
3 12.0 15.4 20.3 25.4
4 13.8 17.5 22.9 28.3
5 16.2 20.2 26.1 32.1
6 18.8 23.4 30.3 37.2
7 21.9 27.3 35.2 43.3
8 26.1 32.5 42.0 51.7
9 34.0 42.4 54.8 67.4
4.1 5.5 7.4 9.4 5.6 8.0 11.0 14.5
6.2 7.9 10.0 12.3 7.9 10.6 14.1 18.1
8.1 10.0 11.8 14.4 9.6 12.5 16.3 20.7
9.1 11.1 13.7 16.4 11.2 14.4 18.5 23.2
11.0 13.2 16.0 19.0 13.3 16.8 21.3 26.5
12.8 15.3 18*6 22.0 15.4 19.5 24.7 30.7
14.9 17.8 21.6 25.6 18.0 22.7 28.8 35.8
17.7 21.3 25.8 30.6 21.4 27.0 34.9 42.7
23.1 27.7 33.6 39.9 27.9 35.3 44.7 55.7
AGE 48 - 53
AGE 54 - 59
6.6 9.2 12.8 17.0
9.1 12;1 16.3 21.0
10.8 14.1 18.6 23.9
12.6 16.1 21.0 26.7
14.8 18.7 24.1 30.3
17.2 21.7 28.0 35.1
20.0 25.2 32.5 40.9
23.8 30.1 38.8 48.8
31.1 39.3 50.6 63.6
AGE 60 - 65
8.8 11.6 15.3 19.3 10.4 13.0 16.5 20.4 12.8 15.3 18.1 21.5
11.6 14.8 19.1 23.1 13.4 16.5 20.4 24.9 16.2 19.1 22.3 26.2
13.5 17.0 21.8 26.8 15.5 18.9 23.2 28.1 18.6 21.8 25.3 29.5
15.5 19.3 24.4 29.9 17.7 21.3 25.9 31.2 20.9 24.4 28.2 32.8
18.0 22.2 27.8 33.8 20.4 24.4 29.5 35.3 24.0 27.8 32.0 37.0
20.9 25.8 32.2 39.2 23.7 28.3 34,2 40.9 27.8 32.2 37.1 42.9
24.3 30.0 37.5 45.6 27.5 32.9 39.8 47.7 32.4 37.5 43.2 50.0
29.0 35.7 44.8 54.4 32.8 39.3 47.5 56.8
37.8 46.6 58.4 71.0 42.8 51.2 62.0 74.1
38.6 44.8 51.5 59.6
50.4 58.4 67.2 77.7
-------
,00
90
80
QQ
fO
LJ
70
60
50
1 40
<
ui
30
20
10
9
8
CO
o
0.
.73
o
UJ
6 ?
o:
.5 H
z
.4
.3
.2
234567
6 YEAR AGE GROUPS (STARTING AT AGE 18)
8
Figure 6. Sample plot of the data from table 6a. Exposure level is 105 dBA.
18
-------
IOO
II II
EXPOSURE IN
30 40
AGE(yrs)
Figure 7. Percent of the population with more than a 5 dB audiometric hearing level
(re 1951 ASA) for the speech frequencies (0.5, 1, and 2 kHz).
19
-------
100
EXPOSURE IN dB A
Figure 8.
30 40
AGE(yrs)
Percent of the population with more than a 10 dB audiometric hearing level
(re 1951 ASA) for the speech frequencies (0.5, 1, and 2 kHz).
20
-------
IOO
00
95
UJ
CO
90 O
x
UJ
85
80
GLORIG'S(I960)
NON-NOISE
EXPOSED
GROUP (2518) 3
10 20 30 40
AGE(yrs)
50 60
70
Figure 9. Percent of the population with more than a 15 dB audiometric hearing level
(re 1951 ASA) for the speech frequencies (0.5, 1, and 2 kHz).
21
-------
100
10
20
30 40
AGE(yrs)
Figure 10. Percent of the population with more than a 25 dB audiometric hearing level
(re 1951 ASA) for the speech frequencies (0.5, 1, and 2 kHz).
22
-------
100
10
30 40
AGE(yrs)
50
60
<
00
LlJ
o:
3
ts>
O
Q.
X
UJ
70
Figure 11. Percent of the population with more than a 40 dB audiometric hearing level
(re 1951 ASA) for the speech frequencies (0.5, 1, and 2 kHz).
23
-------
has the overwhelming advantage. Intense noise produces such high losses
so rapidly that the contribution of aging (which nevertheless is steadily
producing its changes) is completely lost to view. In a number of years,
however, (15 - 18 - 20) the noise induced component decreases and then is
lost and the age component - wMch has been steadily progressing at an
accelerating rate begins to catch up. Depending on the height of the plateau
(./ exposure intensity) the aging component would catch up sooner (low
exposure) or later (high exposure) and then the aging contribution would
(and does) supervene. Our figures indicate that if a whole population
could be kept alive to age 86 it would make no difference what the exposure
history of the members of that population had been, they would all have passed
some specific criterion of hearing loss.
If we look at the percent of population graph for a fence of 15 dB H. L.
3F/3 and look at the 115 dBA exposure line we see that up to the limit of
our graph (65 years of age) aging lias not overtaken - nor even nearly over-
taken - noise loss.
If we look at the 80 dBA line we will see that noise exposure has made no
visible impression on it and it follows the curve of Glorig's "non-noise
exposed" population.* Now if we carefully study the 100 dBA exposure line
we can see a very tiny concavity upward (to the left) at 18 to 23 years or
so which implies a slight aging coinponent but which is nearly lost in the
overwhelming advance of noise loss. Now note that as the rate of noise
induced loss decreases the line straightens, and begins another upward trend
as the plateau becomes fully developed. Eventually it flattens again as the
1001 of population limit is approached.
(As a philosophical aside, we conceive the whole story to be something like
the idealized graph of Fig. 12. This is drawn to represent our idea of a
birth to death (age 0 to 100) graph of the percent of population picture at
the 15 dB H. L. 3F/3 "fence." We are personally satisfied that it is correct
as a generalization although, of course, we don't claim precision of the exact
lines.)
Fig. 13 is a display of median 3F/3 H. L.'s by age for five very well-known
population studies conducted by expert tearcs over a period of thirty years. 3"6
Glorig's non-noise exposed is the only one with a controlled exposure element* a
It would be expected that these studies would agree within fractions of a
decibel, but note that at no age is there a range of less than 8 dB H. L. and
the range goes up to 26 dB at the higher ages! Any one of these surveys could
certainly be considered "authoritative." If we were to perform percent of
population analysis based on each of these medians and its associated distribu-
tion, we would have estimates of such percent varying by as much as 50% or more
of population at certain ages. Now imagine each of these investigating teams,
using exactly the same equipment and teclinicians, doing a survey on populations
with carefully graded exposures; regardless of where their baseline or median
lay, the interval from each exposure to the exposure 5 dB above would remain
constant. Now all surveys would agree on how much each step lay above the
other. In other words, A % Pop./dB exp. - K (or fK.) Either a constant or a
rational function of a constant would be common to all properly done surveys
regardless of systematic variables which might shift the raw data up or down
on the scale. Now, we have only to agree on a baseline. I think we are already
24
-------
ro
CJ1
80
90
100
Figure 12.
Idealized graph drawn to represent a probable pattern from birth to death
of the percent of the population with more than 15 dB audiometric hearing
level (re 1951 ASA) for the speech frequencies 0.5, 1, and 2 kHz-
-------
Ol
50
40
00
io 30
u_
ro
Ul
LU 20
LJ
10
-10
NAT. HEALTH
SURVEY 1935-36
WISC. ST. FAIR 4
1954
JOHANSEN 5
1957
GLORIG
N.N.EXP. 1961-62
H.E.W. SURVRY6
1960-61
20
50
60
70
30 40
AGE(yrs)
Figure 13. Median 3F/3 hearing loss by age for five well-known population studies.
-------
agreed on one - "No exposure below 80 dBA of ordinary mixed industrial noise
produces significant loss of hearing which can be attributed to the industrial
exposure.11 If we are agreed on this, then all that is needed is for the user
to establish points witli his own equipment, his own teclinicians, in his own
population exposed to carefully measured 80 dBA and lower noise. By application
of K or fK he can predict absolute numbers of this population who will experience
selected amounts of hearing loss from higher exposures. He may feel confident
(assuming always that the work is competently done) that his figures will be
consistent with those being developed elsewhere even against different baselines
unique to other investigators, other instrument clusters, and under different
environmental conditions.
In this framework of adjusting baselines, it may be noted that we have in
this report adjusted our own baseline once (adjusting the unexposed median
to that of incoming 18 year olds, a correction of -2.4 dB.) Other adjustments
could be properly made in these data - in fact, I suggest that they be made.
In the first place, our audiograms are taken throughout the day with only a
20 minute (average) quiet rest period preceding. This means there is some
residual temporary threshold shift in our data and we have quantified this
as about 2.3 dB at the mean of the medians. Then there is truncation by
the audiometer at -10 dB. This truncation produces a positive error of unknown
but possibly consequential size (Dr. Douglas Robinson's work in England with
extended range audiometers suggests the error may be significant.)7 This
particular error also affects distributions about the median by introducing a
skewness at the lower signal levels. Also, our recent change from single wall
to double wall audiometric rooms with 10 dB greater attenuation has revealed
some slight residual low frequency masking in the test environment at the time
these data were collected. In short, it appears that at least a 5 dB adjustment,
perhaps considerably more, could be justified.
THE PERCENT OF POPULATION TABLES
When this percent of population display method was first presented in 1966,
the display was presented in only its graphic form. It was implicit, of course,
that numbers could be picked off the graphs and placed in tabular form, and
in fact, this had been done in a working paper for the Intersociety Committee
on Guidelines for Noise Exposure Control. The warm reception of the percent of
population method for the purpose of displaying protection criteria, and interest
in the tabular rather than the graphic display is the reason for this report.
The actual construction of the table is simple. One simply goes to the percent
of population graph based on the desired criterion (e. g. $ of population with
more than 15 dB H. L.), enters at the age in question (e. g. 63 years), proceeds
to the intersection with an exposure (e. g. 80 dBA) and enters the indicated
number (50%) in his tabular grid. Entry of a certain number of such numbers
produces a table of a certain resolution. We have felt that 5 year intervals of
age and 5 dBA intervals of exposure produce a useful table.
We are appending two such tables to this report. The first is constructed from
the data as they appear in this report (Table 7) and the second a table adjusted
to a base of zero dB H. L. at age 20 with 80 dBA exposure (Table 8.)
27
-------
TABLE 7
PO
00
Percent of Population displaying more than
at 500, 1000, 2000
Hertz (ASA 1951)
posure (Assuming Years of Exposure
Age
Exp, Years
Exp. Level
80 dBA
Exp. Level
85 dBA
Ex'.i, Level
90 dBA
Exr>. Level
95
Exo. Level
100 dBA
Exp. Level
105
Exp. Level
110
Exp. Level
115
(Age - 18)
Total % Expected
i Due. to No-c&e
% Due to Other
Total \
i No-iae
% Other
Total %
% Other
Total %
i Nol&e.
% Other
Total %
% Other
Total %
i Wo-ae
% Other
Total %
% Other
Total %
i No-Ut
% Other
18
0
.5
0
.5
.5
0
.5
.5
0
.5
.5
0
.5
.5
0
.5
.5
0
,5
.5
0
.5
.5
0
.5
23
5
1.7
0
1.7
2.5
.£
1.7
6
4.3
1.7
9.0
7.3
1.7
12.3
1.7
20
IS. 3
1.7
28
26.3
1.7
38
36.3
1.7
28
10
3
0
3
6
3
3
13
10
3
20
17
3
32
29
3
45
42
3
58
55
3
71
3
15 dB Hearing Level Averaged
as a Function of Agef Years of Ex-
= Age -
33
15
4.5
0
4.5
9
4.5
4.5
18
J3.5
28
23.5
U.5
42
36.5
U.5
57
52.5
U.5
75
70.5
87
£3.5
18) and
38
20
6.5
0
6.5
12,5
6
6.5
22
75.5
6.5
34
27.5
6.5
48
41.5
6.5
64
57.5
6.5
84
77.5
6.5
93
£6.5
6.5
Exposure
25
9.7
0
9.7
16.5
6.t
9.7
26
16.3
9.7
39
29.3
9.7
53
43.3
9.7
70
60.3
9. '7
88
7£.3
9.7
94
£4.3
9.7
Level in dBA.
48
30
14
0
11
22
£
32
IS
14
45
31
58
44
14
76
62
91
77
95
£1
14
53
35
21
0
21
30
9
21
41
20
21
53
32
21
65
44
21
82
61
21
93
72
21
96
75
21
58
40
33
0
33
43
10
31
54
21
33
62
29
33
74
41
23
87
54
33
95
62
33
97
64
33
63
45
50
0
50
57
7
50
65
15
50
73
23
50
83
33
50
91
41
50
95
45
50
97
47
50
-------
TABLE 8
Percent of Population exhibiting more than IS db A.S.A. 1951? Hearing Level Averaged
at S00t lOOOj and jopojlertz as a Function of Age» Years of Exposure (Assuming years
of Exposure * Age - 18) and Exposure Level in dBA. (Adjusted) **
(I Noise » "Risk" as defined in docunent.)
**
ro
vo
Age
Exp. Years
Exp. Level
80 dBA
Exp. Level
85
Exp. Level
90
Exp. Level
95
Exp. Level
100
Exp. Level
105
Exp. Level
110
Exp. Level
115
(Age - 18)
Total % Rxpccted
1 Vuz to Wo-tae
I Due to Other
Total t
$ Due. to No&e.
% Due to Other
Total *
% Noi&e.
% Other
Total *
I N*c*e
% Other
Total %
1 Ata&e
% Other
Total %
1 Afotae
% Other
Total %
* Wp^e
% Other
Total %
i Wc-ae
% Other
18
0
0.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
.7
0.0
.7
23
S
1.0
0.0
1.0
2.0
7.0
1.0
4.0
3.0
1.0
6.7
5.7
1.0
10.0
9.0
1.0
14.2
13.2
1.0
20.0
19.0
1.0
27.0
26. 0
1.0
28
10
1.3
0.0
1.3
3.9
2.6
1.3
7.9
6.6
1.3
13.6
12.3
1.3
22.0
20.7
1.3
33.0
31.7
1.3
47.5
46.2
1.3
62.5
61.2
1.3
33
15
2.0
0.0
2.0
6.0
4.0
2.0
12.0
10.0
2.0
20.2
IS. 2
2.0
32.0
30.0
2.0
46.0
44.0
2.0
63.0
61.0
2.0
81.0
79.0
2.0
38
20
3.1
0.0
3.1
8.1
5.0
3.1
15,0
n.9
3.1
24.5
23.4
3.1
39.0
35.9
3.1
53.0
49.9
3.1
71.5
68.4
3.1
87.0
83.9
3.1
43
25
4.9
0.0
4.9
11.0
6.1
4.9
18.3
13.4
4.9
29.0
24.1
<4.9
43.0
3S.7
«».9
59.0
54. 1
"*.9
78.0
73. 1
U.9
91.0
86.1
«».9
4S
30
7.7
0.0
7.7
14.2
6.5
7.7
23.3
15.6
7.7
34.4
26.7
7.7
48.5
40. S
7.7
65.5
57. *
7.7
81.5
73.8
7.7
92.0
84.3
7.7
S3
35
13.5
0.0
13.5
21.5
*.0
13.5
31 .-0
17.5
13.5
41.8
2*. 3
13.5
55.0
4T.5
13.5
71.0
57.5
13.5
85.0
7J.5
13.5
93.0
79.5
13.5
58
40
24.0
0.0
24. 0
32.0
8.0
2U.O
42.0
18.0
2«*.0
52.0
28.0
24.0
64.0
40.0
2U.O
78.0
54.0
2«KO
88.0
64.0
2«t.O
94.0
70.0
24.0
63"
45
40.0
0.0
10.0
46.5
6.5
HO.O
54.5
14.5
40.0
64.0
24.0
UO.O
75.0
35.0
40.0
84.5
44.5
40. 0
91.5
51.5
40.0
95.0
55.0
40.0
* 25 dB A. N. S. I.
-------
ADDENDUM
Any frequency, or any combination of frequencies, may be dealt with as we have
dealt with the three frequency mean. There has been some interest expressed
in the behavior of the ear at 4 Kilohertz, so we are including Table 9, which
defines this behavior. We will not detail its derivation which is parallel
to the development of Table 5. There is the difference that rationalization
of the inter-decile points is much more complex, yielding a different ratio
for eadi point, at each age, for each exposure rather than the neat formulae
(Table 4) applicable to the three frequency mean.
Extrapolation and joint regression surface smoothing have not been done
but we include the table for 40 dB II. L. (4 Kilohertz) at 78, 86, and 92
dBA exposures. Table 10.
30
-------
TABLE 9
MATHEMATICALLY SMOOTHED DECILE POINTS
Int.
Dec.
Points
1
2
3
4
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
5.08
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
m~
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
(4kHz)
24 - 29
815
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
30 - 35
86
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 -
86
14.7
20.9
25.8
30.4
34.9
39.1
43.2
51.0
60.0
AGE 60 -
35.8
43.6
48.1
52.0
55.9
59.3
63.7
66.5
72.7
41
92
17.9
24.4
29.6
35.7
40.6
45.5
50.3
56.0
64.1
65
44.0
50.4
53.3
56.3
58.6
62.1
66. Z
j
f>$.7
75.0
-------
CO
ro
TABLE 10
Percent of Population exhibiting more than 40 dB H. L. at 4 Kilohertz as a function
of Age and Exposure Level in
Exp. Level
78 dBA
Exp. Level
86 dBA
Exp. Level
92 dBA
Total % Expected
1 Due to Noise
% Due to Other
Total
Noise
Other
Total
Noise
Other
18
.5
0,0
.5
.5
0.0
.5
.5
0.0
.5
23
1.1
0.0
1.1
6.9
5.8
1.1
20.5
19.4
1.1
dBA.
28
2.9
0.0
2.9
16.1
13.2
2.9
31.2
28.3
2.9
33
6.0
0.0
6.0
27.0
21.0
6.0
40.0
34.0
6.0
38
12.0
0.0
12.0
38.0
26.0
12.0
49.1
37.1
12.0
43
19.0
0,0
19.0
48.0
29.0
19.0
58.7
39.7
19.0
48
29.5
0.0
29.5
60.1
30.6
29.5
68.4
38.9
29.5
53
41.0
0.0
41.0
71.0
30.0
41.0
78.0
37.0
41 TO
58
55.0
0.0
55.0
79.0
24.0
55.0
86.5
31.5
55.0
63
72.0
0.0
72.0
84.2
12.2
72,0
92.7
20.7
72.0
-------
REFERENCES
1. Baughn, W. L., Noise control--percent of population protected. International
Audiology, Vol. V, No. 3 September 1966, pp 331-338.
2. Glorig, A., and Nixon, J., Hearing loss as a function of age. The Laryngo-
scope, Vol. LXXII, No. 11, November 1962, pp 1956-1610.
3. Beasley, W. C., Normal Hearing for Speech at Each Decade of Life. National
Health Survey Hearing Study Series Bull. 3., USPHS, Washington, D. C., 1938.
4. Glorig, A., Wheeler, D., Quiggle, R., Grings, W., and Summerfield, A., 1954
Wisconsin State Fair Hearing Survey: Statistical Treatment of Clinical and
Audiometrie Data.American Academy Ophthalmology and Otolaryngology and
Research Center Subcommittee on Noise in Industry, Los Angeles, California,
1957.
5. Johansen, H., Loss of Hearing due to Age, Munksgaard, Pub., Copenhagen,
Denmark, p 165.
6. National Center for Health Statistics, Hearing Levels of Adults by Age and
Sex. United States. 1960-1962. Vital and Health Statistics. PHS Pub. No.
1000-Series 11-No. 11. Public Health Service. Washington. U.S. Government
Printing Office, October 1965.
7. Robinson, D. W., The general problems of the control of noise. Symposium
No. 12, The Control of Noise, National Physical Laboratory, London, England,
1962. '
33
------- |