United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-80-174 July 1981
Project Summary
Public Health Effects from
Industrial Amines Production,
A Preliminary Evaluative
Approach
T. Briggs, M. Karaffa, and C. Sawyer
The report gives results of a study to
assess the degree to which available
data on health effects of chemical
exposure, data from the National
Cancer Institute, and information on
production and use of classes of
chemicals suspected of being carcin-
ogenic might be used to identify
possible cancer-related operations
needing field investigation. The search
for exposure intensity data confirmed
that such data are not available from
the open literature. Although published
data are available on national produc-
tion of some large-volume chemicals,
site-specific data on such things as
production rates, process losses, and
process discharges are almost com-
pletely lacking. Until better sources of
information are available, studies such
as this will have to be supplemented
with considerable field investigation
to identify specific operations that are
suitable for in-depth study. Until some
way is established to develop such
information so that it can be used with
existing health data for identification
of potential hazards, kepone and vinyl
chloride incidents will continue to be
dealt with after considerable damage
has already been done. The data anal-
ysis techniques presented can be
useful in identifying process discharges
that must be studied in detail for
definition of their potential environ-
mental impact. Volume 2 of the report
contains five appendices, including a
list of industrial amine chemicals.
This Project Summary was devel-
oped by CPA's Industrial Environmen-
tal Research Laboratory, Research
Triangle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
In the last several years the priorities
of the U.S. Environmental Protection
Agency have shifted to detection and
control of human and environmental
exposures to toxic emissions containing
carcinogenic materials. Chemical anal-
ysis and bioassay testing methods are
being developed and used as valuable
tools in this effort. It seems clear,
however, that some systematic but
inexpensive approach is needed to set
priorities for industries to be investigated
and discharge streams to be tested. This
is especially true for industries as
complex as those involved with the
production and use of synthetic organic
chemicals.
Because it is felt that production and
use of organic chemicals may be in
some way relatable to more frequent
occurrence of some types of cancer, this
project was undertaken. Its purpose isto
assess the degree to which available
data on health effects of chemical
-------
exposure, data from the National Cancer
Institute (NCI), and information on
production and use of classes of chemical
compounds suspected of being carcino-
genic might be used to identify possible
cancer-related operations needing fur-
ther investigation in the field.
Summary
It has been established tha.t industrial
usage of certain amine chemicals has in
the past resulted in cancer in exposed
workers. It also appears that operations
involving amine chemicals may be
contributing to atmospheric nitrosamines
which, according to laboratory animal
studies, are highly potent carcinogens
Hence industrial practices involving
amines were selected for use in this
investigation to define how far informa-
tion in open literature can be used as an
inexpensive means of identifying opera-
tions that appear to warrant costly
characterization studies aimed at full
definition of their potential impacts.
The main elements of the investigation
were:
1) A master list of over 1900 amine
chemicals was developed using
the Tariff Commission report on
synthetic organic chemical pro-
duction and the Stanford Research
Institute Directory of Chemical
Producers. These chemicals were
divided into two groups: primary
amines, which would not react
with N0« in the atmosphere to
form nitrosamines; and secondary
or tertiary amines, which could
react.
2) Health effects literature was re-
viewed to identify studies that
have been conducted to relate
exposure of amines or nitrosamines
to the incidence of cancer.
3) Using information from the study
of commercial amines and back-
ground information from a health
effects investigation, 48 amine
compounds were identified as
candidates for more detailed study.
Consideration was given to a) po-
tential toxicity or carcinogenicity
of the compound, b) volume of
annual production (nigh volume
production was used as an index
of possible exposure to humans),
and c) potential for reaction to
form toxic chemicals.
4) The Stanford Resea'rch Institute
Directory of Chemical Producers
was used to locate sites where one
or more of the amines selected for
study was being manufactured,
and Moody's Industrial Directory
was used to identify sites where it
could be established that they had
been produced for 25 years or
longer. No attempt was made to be
exhaustive in locating such sites.
Investigation of 32 chemicals led
to identification of 12 sites. One
additional county, where amines
have been used in large quantities
for many years, was added to
make a total of 13 considered
suitable for further study.
5) The 13 counties selected were
examined to determine:
a. Whether secondary and tertiary
amines capable of being con-
verted to nitrosamines in the
atmosphere were included in
the population of .amines being
produced.
b. Whether organic dyes and pig-
ment in which amines are used
were produced in any of the
counties.
c. Whether nitric acid plants,
which might contribute to the
conversion of amines to nitros-
amines in the atmosphere, were
present in any of the counties.
6) Data from the National Cancer
Institute were used to examine
cancer rates in two ways. First,
rates of cancer in counties con-
tiguous with those being investi-
gated were compared with rates in
the counties of interest. This ap-
proach was used to screen out the
effect of local conditions (e.g.,
drinking water) that might be con-
tributing to high incidence of can-
cer. Second, the absolute values
for incidence of cancer were de-
termined for those counties with
high relative cancer rates.
7) The relative rates of all 14 kinds of
cancer shown in the NCI data base
were analyzed, giving special at-
tention to six kinds believed most
likely to be associated with amine'
or nitrosamine exposure (esoph-
agus, stomach, liver, pancreas,
kidney, and bladder). In addition
the rates of all cancers were com-
pared to provide an index of the
general prevalence of cancer from
all causes; and the rate for lung
cancer, which is the dominant
form in terms of frequency, was
used as a basis for comparison of
frequency of occurrence. A third
index, the "relative risk" factor
defined by the NCI, was used alon
with relative rates and absolut
rates to rank the 13 counties
Specific results of the rankings ar
summarized in Tables 1 -4. Table
presents the key for numbers an
symbols used in the "Cancer pro
file comments" column in Table
1-4.
8) Information was assembled o
factors that might explain observe
cancer rates (relative or absolute;
This material included backgroum
on population, degree of industri
alization, and types of chemica
plants in operation in the 13 se
lected counties and all surroundin)
counties. The most notable facto
not accounted for was age adjust
ment of the cancer rates observed
i.e., it was not possible to deter
mine whether some of the highei
rates could be attributable to gen
erally older populations in some
counties.
9) Concurrent with the epidemiologi
cal analysis, an attempt was made
to obtain information on potentia
for human exposure to amines 01
nitrosamines resulting from pro-
duction or use of chemicals in the
selected counties.
Conclusions
The search for data on intensity of
exposure confirmed that such informa-
tion is not available from the open
literature. Although published data are
available on national production of
some large-volume chemicals, site-
specific data on such things as produc-
tion rates, process losses, and process
discharges are almost completely lack-
ing. Until better sources of information
are available, studies such as this one
will have to be supplemented with
considerable field investigation to iden-
tify specific operations that are suitable
for in-depth study. It is doubtful whether
data collection authority available to
health officials under such legislation
as the Clean Air Act and Amendments
and the Toxic Substances Control Act is
adequate to support the activities needed
to document human and environmental
impacts associated with the production
and use of synthetic organic chemicals.
The specificity of the current legislated
authority, problems of confidentiality,
and other related problems present
obstacles that would appear to make
systematic development of exposure
data difficult. Until some way is estab-
lished to develop such information so
-------
Table 1. Relative Rates of All Cancers Versus Selected Cancers in 13 Counties
All cancers
Selected cancers
Relative rate
County, state
Kanahwa. W. Va.
Erie, N.Y.
Vigo, Ind.
Harris, Tex.
Warren. N.J.
St. Clair. III.
Salem. N.J.
Summit, Ohio
St. Charles. La.
St. Louis, Mo.
Somerset, N.J.
Brazoria, Tex.
Washington, Ala
M
118
120
116
106
105
101
98
94
91
91
89
87
81
F
111
107
108
108
108
102
105
97
99
96
92
92
94
OA
114
114
112
106
106
102
100
95
95
94
91
90
87
County.state
Kanahwa, W. Va.
Erie. N.Y.
Vigo. Ind.
Salem, N.J.
Washington, Ala.
Harris, Tex.
Warren, N.J.
St. Clair, III.
Brazoria, Tex.
Summit, Ohio
Somerset, N.J.
St. Louis, Mo.
St. Charles. La.
Average relative
M
140
125
119
112
78
106
99
96
92
92
88
87
55
F
111
121
114
110
152
104
110
95
93
93
97
97
137
rate
OA
128
124
116
115
97
104
104
99
93
92
91
90
72
Cancer
profile
comments*'*
(1) (2)
(D (2)
(D (2)
(D (2)
(2)
(D (2)
(D (2)
(11
M - Male Rate
F = Female Rate
OA = Overall Rate
'Refers to county shown in "Selected cancers" column.
"See Table 5.
Table 2. Absolute Rates of All Cancers and Selected Cancers in
County, state
Warren. N.J.
Salem, N.J.
Vigo, Ind.
Erie, N.Y.
St. Clair. III.
Summit, Ohio
Somerset, N.J.
St. Louis. Mo.
Kanahwa. W. Va.
Harris, Tex.
Washington, Ala.
Brazoria, Tex.
St. Charles, La.
National average: All cancers
Selected cancers
All
cancers
(female)
147.640"
146.595"
144.191"
142.230*
138.690*
137.901*
135.655*
131.979*
126.141
124.176
114.009
113.951
112.541
• 129.974 - female
172.733 -male
- 24.232 - female
44.405 - male
All
cancers
(male)
189.201*
185.827*
180.760*
206.966***
195.918**
185.488*
182.711*
181.411*
172.278
188.483*
150.645
165.026
185.710*
13 Counties
Selected
cancers
(female)
28.242***
27.133**
25.953*
27.454**
24.060
25.400*
25.796*
22.881
25.974*
22.069
23.420
21.957
27.04T*
Selected
cancers
(male)
46. 198*
54.221***
42. 105
56. 169***
45.350*
47.927*
50.384**
42.786
43.605
44.348
36.050
42.923
32.900
Cancer
profile
comments"
(1)(2)(3**) (4***)
(1)(2)(3**> (4***)
(1)(2)(3**) (4)
(1)(2)(3***)(4***)
(D (3**)
(3*) (4*)
(3*) (4**)
(3*)
(D(2)
(D(2)(3)
(2)
(3) (4**)
* Above average.
** >1O% above
***>15%above
a See Table 5.
average rate.
average rate.
that it can be used with existing health
data for identification of potential haz-
ards, "kepone" and "vinyl chloride"
incidents will continue to be dealt with
after considerable damage has already
been done. The data analysis techniques
presented here can, however, be* useful
in identifying process discharges that
must be studied in detail for definition of
their potential environmental impact.
Some specific questions seem to require
further county-specific investigations.
These include:
(1) What factors make Salem County
and Warren County, New Jersey,
high in relative rates of cancer,
"all cancer" rates, relative risk of
selected cancers, and relative risk
of bladder cancer despite their
low population, population den-
sity, and degree of industriali-
zation?
(2) Why is Erie County, New York,
highest in cancer rates for all
categories studied? Is there a
connection between these rates
and the known presence of facili-
ties producing and using amines
or producing nitrosamine precur-
sors?
(3) What factors make the cancer
rates of Summit County, Ohio,
high for all cancers, selected
cancers, and bladder cancer? (This
county was included in the study
only because amines are known
to be used in rubber processing.
No production of amines or nitros-
amines has been identified.)
Although other questions are sug-
gested by the assembled data, the
absence of evidence to substantiate the
exposure of humans to amines or nitros-
amines as a result of the investigated
-------
Table 3. Thirteen-County Ranking Based on Relative Risk of All Cancers
County, state
Erie.N.Y.
Warrefi, N.J.
St. Clair, III.
Salem, N.J.
Vigo, Ind.
Summit, Ohio
Somerset, N.J.
St. Louis, Mo.
Harris, Tex.
Kanahwa. W. Va.
St. Charles. La.
Brazoria, Tex.
Washington, Ala.
Combined
average for
male, female
115.1
111.5
110.5
109.9
106.7
106.3
104.6
102.7
102.8
99.4
94.3
92.8
86.8
Male
120.3"
109.6'
114.1"
107.9
103.9
106.6'
105.2
104.0"
109.7
101.0
101.5
96.0
86.1
Female
109.4"
113.8"
106.5*
1 12.4*
109.8*
106.0*
103.8
101.4
95.5
97.6
85.7
88.9
87.7
Cancer
profile
comments*
(1)(2)(3**)(4***)
d ) (2) (3+*) (4***)
d) (3%
(1 ) (2) (3** ) (4)
(3*) (4*)
(3*) (4++)
&)
d)f2)(3)
(D(2)
(3) (4")
(2)
(5+*)
(5")
(5")
(6*)
(5*1
(5)
(5+)
15)
"Statistically significant high relative risk, all cancers.
"See Table 5.
Table 4. Thirteen-County Ranking Based on Relative Risk and Adjusted Relative Risk of Bladder Cancer
County, state
Salem, N.J.
Erie, N.Y.
Warren, N.J.
Summit, Ohio
Somerset, N.J.
St. Louis, Mo.
Vigo, Ind.
St. Charles, La.
Harris, Tex.
St. Clair, III.
Kanahwa, W. Va.
Washington, Ala.
Brazoria, Tex.
Relative risk
male/female
245.8/155.0
147.8/124.7
118.1/141.8
118.1/117.1
1 13.4/128.5
112.9/106.6
103.5/120.3
1 13.4/96.8
106.9/96.1
93.9/102.8
90.9/96.6
86.9/48.5
56.8/95.6
Adjusted relative risk
male/female
2.28/1.38
1.23/1.14
1.08/1.25
1.11/1.10
1.08/1.24
1.09/1.05
<1/1.10
1.12/1.13
<1/1.01
<1/-1
<1/-1
1.01/-1
<1/1.08
Cancer profile
comments"
(1)(2)(3~) (4+")(5+) (6+)
d)(2)(3"+)(4+++)(5++)(6")
(1)12) (3") (4+*) (5") (6)
(3*) (4+) (5*+j(6+)
(3+) (4") (5) (6)
(3+) (5+) (6)
d)(2){3") (4) (5+)
(3) (4)
(1){2)(3) 15)
(1) (3") (5*+)
(DI2)
(2)
"See Table 5.
industrial activities makes further dis-
cussion inappropriate.
As indicated earlier it is felt that the
results have confirmed the usefulness
'of health effects studies and cancer rate
data in analysis of potential health
impacts associated with the production
and use of chemicals. Also the results
tend to support certain conclusions as to
how such studies should be designed. It
appears from the results for bladder
cancer studies that it is important to
identify a form of cancer that is strongly
associated with the chemical(s) being
investigated and that the chemical be
fairly specific to that form of cancer. The
results of the study of selected cancers
are similarly interesting from the stand-
point of the lack of association with high
rates attributed to smoking. It was
known however that nitrosamines are
"highly versatile"; i.e., apparently cap-
able of producing cancer in animals via
any route of exposure and in any organ.
It is not surprising, therefore, that
selected cancer rates tended to follow
all cancer rates and are apparently
relatable to general overall cancer rates.
This could be because (1) there is a lack
of real association of these cancers with
nitrosamine exposure, and (2) nitros-
amines are so versatile and pervasive in
the environment that their impact can-
not be distinguished from general con-
tamination that is producing high cancer
rates, etc. Whatever the reason, the
results demonstrate the importance of
selection of cancer forms for studies of
the type reported here.
Recommendations
Investigations of possible causes of
high bladder cancer rates should be
undertaken in Salem County, New
Jersey; Warren and Somerset Counties,
New Jersey; Erie County, New York; and
Summit County, Ohio. The potential of
the general population exposure to
amines, nitrosamines, or other poten-
tially carcinogenic discharges (e.g., coke
oven emissions) should be defined.
Such investigations should consider all
the industries in the counties named
and the discharges from them. Dis-
charges should be tested from facilities
that appear the most likely to be causing
human exposure to a potential carcinogen
and the air and water quality in the
vicinity of these facilities should be
monitored.
Health records of localities where
potentially carcinogenic discharges are
believed to be occurring should be
studied further (e.g., death certificate
information on cases involving bladder
cancer should be reviewed to determine
-------
Table 5. Key for Cancer Profile Comments, Tables 1 through 4
(1) Above average relative rate, all cancers.
(2) Above average relative rate, selected cancers.
(3) Above average absolute rate, all cancers, male or female.
(3*) Above average absolute rate, all cancers, male and female.
(3") More than 10% over absolute rate for all cancers, male, female, or both.
(3***) More than 15% over absolute rate for all cancers, male, female, or both.
(4) Above average absolute rate, selected cancers, male or female.
(4*) Above average absolute rate, selected cancers, male, and female.
(4**) More than 10% over absolute rate, selected cancers, male, female, or both.
(4***) More than 15% over absolute rate, selected cancers, male, female, or both.
(5) Higher than average relative risk, all cancers.
•(5*) Significantly high relative risk, all cancers, male or female.
(5**) Significantly high relative risk, all cancers, male and female.
(6) Higher than average relative risk for bladder cancer, and adjusted relative
risk >1.0, male and female.
(6*) Significantly high relative risk for bladder cancer, male or female.
(6**) Significantly high relative risk for bladder cancer, male and female.
whether occupational exposures to
amines or nitrosamines might be in-
volved).
The data collection capabilities of the
EPA should be examined to see if they
are adequate for collection of data to
assess the general population exposure
to industrial discharges.
Industrial activity and community
health should be studied further to
evaluate the methodology that has been
developed. A study of the possible
association between plants producing
organic dyes and pigments and the high
incidence of bladder cancer might be
appropriate.
T-. Briggs, M. Karaffa. and C. Sawyer are with PEDCo Environmental. Inc..
Cincinnati, OH 45246; P. Spaite is with Paul W. Spaite Consultants. Cincinnati,
OH.
John O. Smith is the EPA Project Officer (see below).
The complete reports covered by this Project Summary:
"Public Health Effects from Industrial Amines Production, A Preliminary
Evaluative Approach: Volume 1," (Order No. PB 81-129 207; Cost: $12.50.
subject to change).
"Public Health Effects from Industrial^ Amines Production, A Preliminary
Evaluative Approach: Volume 2. Appendices." (Order No. PB 81-129 215;
Cost: $ 17.OO. subject to change).
The above reports will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
* UAOOVERNMENrfMNTWO OFFICE 1W1 -757-012/7232
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 0000329
U S KWVTK PRQTfelCl'lUS
S DEARBORN
CHICAGO Hi 60604
------- |