vvEPA
EPA/635/R-13/139a
Revised External Review Draft
www.epa.gov/iris
Toxicological Review of Ammonia
(CASRN 7664-41-7)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
August 2013
NOTICE
This document is a Revised External Review draft. This information is distributed solely for the
purpose of pre-dissemination peer review under applicable information quality guidelines. It has
not been formally disseminated by EPA. It does not represent and should not be construed to
represent any Agency determination or policy. It is being circulated for review of its technical
accuracy and science policy implications.
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Toxicological Review of Ammonia
DISCLAIMER
This document is a preliminary draft for review purposes only. This information is
distributed solely for the purpose of pre-dissemination peer review under applicable
information quality guidelines. It has not been formally disseminated by EPA. It does not
represent and should not be construed to represent any Agency determination or policy.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
CONTENTS
AUTHORS | CONTRIBUTORS | REVIEWERS vi
PREFACE viii
PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS xii
EXECUTIVE SUM MARY xxx
LITERATURE SEARCH STRATEGY | STUDY SELECTION AND EVALUATION xxxvi
1. HAZARD IDENTIFICATION 1-1
1.1. SYNTHESIS OF EVIDENCE 1-1
1.1.1. Respiratory Effects 1-1
1.1.2. Gastrointestinal Effects 1-16
1.1.3. Immune System Effects 1-21
1.1.4. Other Systemic Effects 1-25
1.1.5. Carcinogenicity 1-33
1.2. SUMMARY AND EVALUATION 1-36
1.2.1. Weight of Evidence for Effects Other than Cancer 1-36
1.2.2. Weight of Evidence for Carcinogenicity 1-37
1.2.3. Susceptible Populations and Lifestages 1-37
2. DOSE-RESPONSE ANALYSIS 2-1
2.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER 2-1
2.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER THAN CANCER 2-2
2.2.1. Identification of Studies and Effects for Dose-Response Analysis 2-2
2.2.2. Methods of Analysis 2-4
2.2.3. Derivation of the Reference Concentration 2-5
2.2.4. Uncertainties in the Derivation of the Reference Concentration 2-6
2.2.5. Confidence Statement 2-8
2.2.6. Previous IRIS Assessment 2-9
2.3. Cancer Risk Estimates 2-9
REFERENCES R-l
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
TABLES
Table ES-1. Summary of reference concentration (RfC) derivation xxxii
Table LS-1. Details of the literature search strategy employed for ammonia xxxvii
Table 1-1. Evidence pertaining to respiratory effects in humans following inhalation exposure
in industrial settings 1-6
Table 1-2. Evidence pertaining to respiratory effect in humans following inhalation exposure in
cleaning settings 1-9
Table 1-3. Evidence pertaining to respiratory effects in animals 1-12
Table 1-4. Evidence pertaining to gastrointestinal effects in animals 1-18
Table 1-5. Evidence pertaining to immune system effects in animals 1-23
Table 1-6. Evidence pertaining to other systemic effects in humans 1-27
Table 1-7. Evidence pertaining to other systemic effects in animals 1-28
Table 1-8. Evidence pertaining to cancer in animals 1-35
FIGURES
Figure LS-1. Study selection strategy xxxviii
Figure 1-1. Exposure-response array of respiratory effects following inhalation exposure to
ammonia 1-14
Figure 1-2. Exposure-response array of gastrointestinal effects following oral exposure to
ammonia 1-19
Figure 1-3. Exposure-response array of immune system effects following inhalation exposure to
ammonia 1-24
Figure 1-4. Exposure-response array of systemic effects following inhalation exposure to
ammonia 1-32
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT—DO NOT CITE OR QUOTE
-------
ABBREVIATIONS
Toxicological Review of Ammonia
ALT alanine aminotransferase MRM
AST aspartate aminotransferase NCEA
ATSDR Agency for Toxic Substances and Disease
Registry NH3
BCG bacillus Calmette-Guerin NH4+
BMCL 95% lower bound on the benchmark NIOSH
concentration
BMDL 95% lower bound on the benchmark NOAEL
dose NRC
CAC cumulative ammonia concentration ORD
CCRIS Chemical Carcinogenesis Research
Information System PEFR
CERCLA Comprehensive Environmental pOz
Response, Compensation, and Liability POD
Act PPD
CPU colony forming unit RfC
CI confidence interval RfD
DAP diammonium phosphate RTECS
EPA Environmental Protection Agency
FEVi forced expiratory volume in 1 second TSCATS
FVC forced vital capacity
HERO Health and Environmental Research UF
Online UFA
HSDB Hazardous Substances Data Bank UFn
IgE immunoglobulin E UFi
IgG immunoglobulin G UFS
IRIS Integrated Risk Information System UFo
LDso 50% lethal dose VEh
LOAEL lowest-observed-adverse-effect level
MAO monoamine oxidase VEho
MNNG N-methyl-N'-nitro-N-nitrosoguanidine
murine respiratory mycoplasmosis
National Center for Environmental
Assessment
ammonia
ammonium ion
National Institute for Occupational
Safely and Health
no-observed-adverse-effect level
National Research Council
EPA's Office of Research and
Development
peak expiratory flow rate
oxygen partial pressure
point of departure
purified protein derivative
reference concentration
reference dose
Registry of Toxic Effects of Chemical
Substances
Toxic Substance Control Act Test
Submission Database
uncertainty factor
interspecies uncertainty factor
intraspecies uncertainly factor
LOAEL to NOAEL uncertainty factor
subchronic-to-chronic uncertainly factor
database deficiencies uncertainly factor
human occupational default minute
volume
human ambient default minute volume
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
AUTHORS | CONTRIBUTORS | REVIEWERS
4
5
Assessment Team
Audrey Galizia, Dr. PH (Chemical
Manager)
James Ball, Ph.D.
Glinda Cooper, Ph.D.
Louis D'Amico, Ph.D.
Keith Salazar, Ph.D.
Christopher Sheth, Ph.D.
Christopher Brinkerhoff, Ph.D.
U.S. EPA/ORD/NCEA
Edison, NJ
U.S. EPA/ORD/NCEA
Washington, DC
ORISE Postdoctoral Fellow at the U.S. EPA
Washington, DC
Scientific Support Team
Vincent Cogliano, Ph.D.
Samantha Jones, Ph.D.
Jamie Strong, Ph.D.
Ted Berner, MS
Jason Fritz, Ph.D.
Martin Gehlhaus, MPH
John Stanek, Ph.D.
U.S. EPA/ORD/NCEA
Washington, DC
U.S. EPA/ORD/NCEA
Research Triangle Park, NC
Production Team
Maureen Johnson
Vicki Soto
Ellen F. Lorang, MA
U.S. EPA/ORD/NCEA
Washington, DC
U.S. EPA/ORD/NCEA
Research Triangle Park, NC
Contractor Support
Amber Bacom, MS
Fernando Llados, Ph.D.
Julie Stickney, Ph.D.
SRC, Inc., Syracuse, NY
14
Executive Direction
Kenneth Olden, Ph.D., Sc.D., L.H.D. (Center Director)
John Vandenberg, Ph.D. (National Program Director, HHRA)
Lynn Flowers, Ph.D., DABT (Associate Director for Health)
Vincent Cogliano, Ph.D. (IRIS Program Director-acting)
Samantha Jones, Ph.D. (IRIS Associate Director for Science)
Susan Rieth, MPH (Branch Chief)
U.S. EPA/ORD/NCEA
Washington, DC
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Internal Review Team
Marian Rutigliano, MD U.S. EPA/ORD/NCEA
John Whalan Washington, DC
Amanda S. Persad, Ph.D. U.S. EPA/ORD/NCEA
Paul Reinhart, Ph.D. Research Triangle Park, NC
Reviewers
4
5 This assessment was provided for review to scientists in EPA's program and regional Offices.
6 Comments were submitted by:
7
8 Office of Policy, Washington, DC
9 Office of Water, Washington, DC
10 Office of Children's Health Protection, Washington, DC
11 Office of Transportation and Air Quality in the Office of Air and Radiation, Ann Arbor, Michigan
12 Office of Air Quality and Planning Standards in the Office of Air and Radiation, Washington, DC
13 Region 2, New York, New York
14
15 This assessment was provided for review to other federal agencies and the Executive Office of the
16 President. Comments were submitted by:
17
Agency for Toxic Substances and Disease Registry, Centers for Disease Control and
Prevention, Department of Health & Human Services
Council on Environmental Quality, Executive Office of the President
Food Safety and Inspection Service, U.S. Department of Agriculture
18 This assessmentwas released for public commenton June 8, 2012 and comments were due on
19 August 7, 2012. A summary and EPA's disposition of the comments received from the public is
20 included in Appendix G of the Supplemental Information to the Toxicological Review. Comments
21 were received from the following entities:
The American Chemistry Council Washington, DC
The Fertilizer Institute Washington, DC
22
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
3 PREFACE
4
5
6 This Toxicological Review critically reviews the publicly available studies on ammonia in
7 order to identify its adverse health effects and to characterize exposure-response relationships.
8 The assessment covers gaseous ammonia (NHs] and ammonia dissolved in water (ammonium
9 hydroxide, NFUOH). It was prepared under the auspices of the Environmental Protection Agency's
10 (EPA's) Integrated Risk Information System (IRIS) program.
11 Ammonia and ammonium hydroxide are listed as hazardous substances under the
12 Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA) and
13 ammonia is found at about 8% of hazardous waste sites on the National Priorities List (ATSDR,
14 2004]. Ammonia is subject to reporting requirements for the Toxics Release Inventory under the
15 Emergency Planning and Community Right-to-Know Act of 1986 and to emergency planning
16 requirements under section 112(r) of the Clean Air Act
17 This assessment updates a previous IRIS assessment of ammonia that was developed in
18 1991. The previous assessment included only an inhalation reference concentration (RfC) for
19 effects other than cancer. New information has become available, and this assessment reviews
20 information on all health effects by all exposure routes.
21 This assessment was conducted in accordance with EPA guidance, which is cited and
22 summarized in the Preamble to IRIS Toxicological Reviews. The findings of this assessment and
23 related documents produced during its development are available on the IRIS website
24 (http://www.epa.gov/iris/]. Appendices for chemical and physical properties, the toxicity of
25 ammonium salts, toxicokinetic information, and summaries of toxicity studies and other
26 information are provided as Supplemental Information to this assessment (see Appendices A to E).
27 Portions of this Toxicological Review were adapted from the Toxicological Profile for
28 Ammonia developed by the Agency for Toxic Substances and Disease Registry (ATSDR, 2004] under
29 a Memorandum of Understanding that encourages interagency collaboration, sharing of scientific
30 information, and more efficient use of resources.
31
32 Implementation of the 2011 National Research Council Recommendations
33 On December 23, 2011, The Consolidated Appropriations Act, 2012, was signed into law
34 (U.S. Congress, 2011]. The report language included direction to EPA for the IRIS Program related
35 to recommendations provided by the National Research Council (NRC] in their review of EPA's
36 draft IRIS assessment of formaldehyde (NRC. 2011]. The report language included the following:
37
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 The Agency shall incorporate, as appropriate, based on chemical-specific data sets
2 and biological effects, the recommendations of Chapter 7 of the National Research
3 Council's Review of the Environmental Protection Agency's Draft IRIS Assessment of
4 Formaldehyde into the IRIS process...For draft assessments released in fiscal year
5 2012, the Agency shall include documentation describing how the Chapter 7
6 recommendations of the National Academy of Sciences (NAS) have been
7 implemented or addressed, including an explanation for why certain
8 recommendations were not incorporated.
9
10 The NRC's recommendations, provided in Chapter 7 of the review report, offered
11 suggestions to EPA for improving the development of IRIS assessments. Consistent with the
12 direction provided by Congress, documentation of how the recommendations from Chapter 7 of the
13 NRC report have been implemented in this assessment is provided in Appendix F. Where
14 necessary, the documentation includes an explanation for why certain recommendations were not
15 incorporated.
16 The IRIS Program's implementation of the NRC recommendations is following a phased
17 approach that is consistent with the NRC's "Roadmap for Revision" as described in Chapter 7 of the
18 formaldehyde review report. The NRC stated that, "the committee recognizes that the changes
19 suggested would involve a multi-year process and extensive effort by the staff at the National
20 Center for Environmental Assessment and input and review by the EPA Science Advisory Board and
21 others."
22 Phase 1 of implementation has focused on a subset of the short-term recommendations,
23 such as editing and streamlining documents, increasing transparency and clarity, and using more
24 tables, figures, and appendices to present information and data in assessments. Phase 1 also
25 focused on assessments near the end of the development process and close to final posting. The
26 IRIS assessment for ammonia is the first assessment in Phase 2 of implementation, which addresses
27 all of the short-term NRC recommendations (see Appendix F, Table F-l). The IRIS Program is
28 implementing all of these recommendations but recognizes that achieving full and robust
29 implementation of certain recommendations will be an evolving process with input and feedback
30 from the public, stakeholders, and external peer review committees. Chemical assessments in
31 Phase 3 of implementation will incorporate the longer-term recommendations made by the NRC
32 (see Appendix F, Table F-2), including the development of a standardized approach to describe the
33 strength of the evidence for noncancer effects. On May 16, 2012, EPA announced (U.S. EPA. 2012c]
34 that as a part of a review of the IRIS Program's assessment development process, the NRC will also
35 review current methods for weight-of-evidence analyses and recommend approaches for weighing
36 scientific evidence for chemical hazard identification. This effort is included in Phase 3 of EPA's
37 implementation plan.
38
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Assessments by Other National and International Health Agencies
2 Toxicity information on ammonia has been evaluated by ATSDR, the National Research
3 Council (NRC), the American Conference of Governmental Industrial Hygienists, the National
4 Institute for Occupational Safety and Health, and the Food and Drug Administration. The results of
5 these assessments are presented in Appendix A of the Supplemental Information. It is important to
6 recognize that these assessments may have been prepared for different purposes and may utilize
7 different methods, and that newer studies may be included in the IRIS assessment.
8
9 Chemical Properties and Uses
10 Ammonia is a corrosive gas with a pungent odor. It is highly soluble in water (up to
11 482 g/L) and is a weak base [Lide. 2008: O'Neil etal.. 2006: Eggeman. 2001: Dean. 1985].
12 Additional information on the chemical and physical properties of ammonia is presented in
13 Appendix B.
14 About 80% of commercially produced ammonia is used in agricultural fertilizers. Ammonia
15 is also used as a corrosion inhibitor, in water purification, as a household cleaner, as an
16 antimicrobial agent in food products, as a refrigerant, as a stabilizer in the rubber industry, in the
17 pulp and paper and metallurgy industries, as a source of hydrogen in the hydrogenation of fats and
18 oils, and as a chemical intermediate in the production of Pharmaceuticals, explosives, and other
19 chemicals. Ammonia is also used to reduce nitrogen oxide emissions from combustion sources such
20 as industrial and municipal boilers, power generators, and diesel engines [HSDB, 2012: Tohnson et
21 al.. 2009: Eggeman. 20011
22 Ammonia is a component of the global nitrogen cycle and is essential to many biological
23 processes. Nitrogen-fixing bacteria convert atmospheric nitrogen to ammonia that is available for
24 uptake into plants. Organic nitrogen released from biota can be converted to ammonia. Ammonia
25 in water and soil can be converted to nitrite and nitrate through the process of nitrification.
26 Ammonia is also endogenously produced in humans and other mammals, where it is an essential
27 metabolite used in nucleic acid and protein synthesis, is necessary for maintaining acid-base
28 balance, and is an integral part of nitrogen homeostasis [Nelson and Cox. 2008: Socolow. 1999:
29 Rosswall. 1981). This assessment compares endogenous levels of ammonia in humans to the
30 toxicity values that it derives.
31
32 Consideration of Ammonium Salts for Inclusion in This Assessment
33 EPA considered whether to include ammonium salts (e.g., ammonium acetate, chloride, and
34 sulfate) in this assessment. These salts readily dissolve in water through dissociation into an
35 ammonium cation (NH4+) and an anion. Oral toxicity studies on ammonium chloride and
36 ammonium sulfate suggest that these salts may differ in toxicity (see Appendix C for a summary of
37 subchronic/chronic toxicity information for selected ammonium salts), but it is not clear whether
38 this reflects differences between the salts or in the effects that were studied. If the toxicity of the
This document is a draft for review purposes only and does not constitute Agency policy.
x DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 salts is affected by the anion, then it would not be correct to attribute toxic effects to the ammonium
2 cation. ATSDR considered this question and concluded, "... that it would be inappropriate to
3 extrapolate findings obtained with ammonium chloride (or any ammonium salt) to equivalent
4 amounts of ammonium, but derived from a different salt" [ATSDR, 2004]. Similarly, the World
5 Health Organization considered ammonium chloride-induced kidney hypertrophy and observed
6 that the extent to which it results from ammonium chloride-induced acidosis or from a direct effect
7 of the ammonium ion is not clear [IPCS, 1986]. Thus, in light of the uncertain influence of the anion
8 on toxicity, ammonium salts were not used in the identification of effects or in the derivation of
9 reference values for ammonia and ammonium hydroxide.
10
11 For additional information about this assessment or for general questions regarding IRIS,
12 please contact EPA's IRIS Hotline at 202-566-1676 (phone], 202-566-1749 (fax], or
13 hotline.iris@epa.gov.
14
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
PREAMBLE TO IRIS TOXICOLOGICAL REVIEWS
3
4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
1. Scope of the IRIS Program
Soon after the EPA was established in
1970, it was at the forefront of developing
risk assessment as a science and applying it
in decisions to protect human health and the
environment The Clean Air Act, for example,
mandates that the EPA provide "an ample
margin of safety to protect public health";
the Safe Drinking Water Act, that "no
adverse effects on the health of persons may
reasonably be anticipated to occur, allowing
an adequate margin of safety." Accordingly,
the EPA uses information on the adverse
effects of chemicals and on exposure levels
below which these effects are not
anticipated to occur.
IRIS assessments critically review the
publicly available studies to identify adverse
health effects from exposure to chemicals
and to characterize exposure-response
relationships. In terms set forth by the
National Research Council [NRC. 1983], IRIS
assessments cover the hazard identification
and dose-response assessment steps of risk
assessment, not the exposure assessment or
risk characterization steps that are
conducted by the EPA's program and
regional offices and by other federal, state,
and local health agencies that evaluate risk
in specific populations and exposure
scenarios. IRIS assessments are distinct from
and do not address political, economic, and
technical considerations that influence the
design and selection of risk management
alternatives.
An IRIS assessment may cover a single
chemical, a group of structurally or
toxicologically related chemicals, or a
complex mixture. These agents may be found
in air, water, soil, or sediment. Exceptions
are chemicals currently used exclusively as
46 pesticides, ionizing and non-ionizing
47 radiation, and criteria air pollutants listed
48 under section 108 of the Clean Air Act
49 (carbon monoxide, lead, nitrogen oxides,
50 ozone, particulate matter, and sulfur oxides).
51 Periodically, the IRIS Program asks other
52 EPA programs and regions, other federal
53 agencies, state health agencies, and the
54 general public to nominate chemicals and
55 mixtures for future assessment or
56 reassessment Agents may be considered for
57 reassessment as significant new studies are
58 published. Selection is based on program
59 and regional office priorities and on
60 availability of adequate information to
61 evaluate the potential for adverse effects.
62 Other agents may also be assessed in
63 response to an urgent public health need.
64 2. Process for developing and peer-
65 reviewing IRIS assessments
66 The process for developing IRIS
67 assessments (revised in May 2009 and
68 enhanced in July 2013) involves critical
69 analysis of the pertinent studies,
70 opportunities for public input, and multiple
71 levels of scientific review. The EPA revises
72 draft assessments after each review, and
73 external drafts and comments become part
74 of the public record (U.S. EPA. 2009).
75 Before beginning an assessment, the IRIS
76 Program discusses the scope with other EPA
77 programs and regions to ensure that the
78 assessment will meet their needs. Then a
79 public meeting on problem formulation
80 invites discussion of the key issues and the
81 studies and analytical approaches that might
82 contribute to their resolution.
83 Step 1. Development of a draft
84 Toxicological Review. The draft
85 assessment considers all pertinent
86 publicly available studies and applies
This document is a draft for review purposes only and does not constitute Agency policy.
xii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 consistent criteria to evaluate study
2 quality, identify health effects, identify
3 mechanistic events and pathways,
4 integrate the evidence of causation for
5 each effect, and derive toxicity values. A
6 public meeting prior to the integration of
7 evidence and derivation of toxicity
8 values promotes public discussion of the
9 literature search, evidence, and key
10 issues.
11 Step 2. Internal review by scientists in
12 EPA programs and regions. The draft
13 assessment is revised to address the
14 comments from within the EPA.
15 Step 3. Interagency science consultation
16 with other federal agencies and the
17 Executive Offices of the President.
18 The draft assessment is revised to
19 address the interagency comments. The
20 science consultation draft, interagency
21 comments, and the EPA's response to
22 major comments become part of the
23 public record.
24 Step 4. Public review and comment,
25 followed by external peer review. The
26 EPA releases the draft assessment for
27 public review and comment A public
28 meeting provides an opportunity to
29 discuss the assessment prior to peer
30 review. Then the EPA releases a draft for
31 external peer review. The peer review
32 meeting is open to the public and
33 includes time for oral public comments.
34 The peer reviewers assess whether the
35 evidence has been assembled and
36 evaluated according to guidelines and
37 whether the conclusions are justified by
38 the evidence. The peer review draft,
39 written public comments, and peer
40 review report become part of the public
41 record.
42 Step 5. Revision of draft Toxicological
43 Review and development of draft IRIS
44 summary. The draft assessment is
45 revised to reflect the peer review
46 comments, public comments, and newly
47 published studies that are critical to the
48 conclusions of the assessment. The
49 disposition of peer review comments
50 and public comments becomes part of
51 the public record.
52 Step 6. Final EPA review and interagency
53 science discussion with other federal
54 agencies and the Executive Offices of
55 the President The draft assessment and
56 summary are revised to address the EPA
57 and interagency comments. The science
58 discussion draft, written interagency
59 comments, and EPA's response to major
60 comments become part of the public
61 record.
62 Step 7. Completion and posting. The
63 Toxicological Review and IRIS summary
64 are posted on the IRIS website [http://
65 www.epa.gov/iris].
66 The remainder of this Preamble
67 addresses step 1, the development of a draft
68 Toxicological Review. IRIS assessments
69 follow standard practices of evidence
70 evaluation and peer review, many of which
71 are discussed in EPA guidelines [U.S. EPA.
72 2005a. b. 2000. 1998. 1996. 1991. 1986a. b)
73 and other methods [U.S. EPA. 2012a. b, 2011.
74 2006a. b, 2002. 1994b). Transparent
75 application of scientific judgment is of
76 paramount importance. To provide a
77 harmonized approach across IRIS
78 assessments, this Preamble summarizes
79 concepts from these guidelines and
80 emphasizes principles of general
81 applicability.
82 3. Identifying and selecting
83 pertinent studies
84 3.1. Identifying studies
85 Before beginning an assessment, the EPA
86 conducts a comprehensive search of the
87 primary scientific literature. The literature
88 search follows standard practices and
89 includes the PubMed and ToxNet databases
90 of the National Library of Medicine, Web of
91 Science, and other databases listed in the
92 EPA's HERO system (Health and
93 Environmental Research Online, http://
This document is a draft for review purposes only and does not constitute Agency policy.
xiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 hero.epa.gov/]. Searches for information on
2 mechanisms of toxicity are inherently
3 specialized and may include studies on other
4 agents that act through related mechanisms.
5 Each assessment specifies the search
6 strategies, keywords, and cut-off dates of its
7 literature searches. The EPA posts the
8 results of the literature search on the IRIS
9 web site and requests information from the
10 public on additional studies and ongoing
11 research.
12 The EPA also considers studies received
13 through the IRIS Submission Desk and
14 studies (typically unpublished) submitted
15 under the Toxic Substances Control Act or
16 the Federal Insecticide, Fungicide, and
17 Rodenticide Act. Material submitted as
18 Confidential Business Information is
19 considered only if it includes health and
20 safety data that can be publicly released. If a
21 study that may be critical to the conclusions
22 of the assessment has not been peer-
23 reviewed, the EPA will have it peer-
24 reviewed.
25 The EPA also examines the toxicokinetics
26 of the agent to identify other chemicals (for
27 example, major metabolites of the agent) to
28 include in the assessment if adequate
29 information is available, in order to more
30 fully explain the toxicity of the agent and to
31 suggest dose metrics for subsequent
32 modeling.
33 In assessments of chemical mixtures,
34 mixture studies are preferred for their
35 ability to reflect interactions among
36 components. The literature search seeks, in
37 decreasing order of preference (U.S. EPA,
38 2000.32.1. 1986b. 32.21:
39 - Studies of the mixture being assessed.
40 - Studies of a sufficiently similar mixture.
41 In evaluating similarity, the assessment
42 considers the alteration of mixtures in
43 the environment through partitioning
44 and transformation.
45 - Studies of individual chemical
46 components of the mixture, if there are
47 not adequate studies of sufficiently
48 similar mixtures.
49 3.2. Selecting pertinent epidemiologic
50 studies
51 Study design is the key consideration for
52 selecting pertinent epidemiologic studies
53 from the results of the literature search.
54 - Cohort studies, case-control studies, and
55 some population-based surveys (for
56 example, NHANES) provide the strongest
57 epidemiologic evidence, especially if they
58 collect information about individual
59 exposures and effects.
60 - Ecological studies (geographic
61 correlation studies) relate exposures and
62 effects by geographic area. They can
63 provide strong evidence if there are
64 large exposure contrasts between
65 geographic areas, relatively little
66 exposure variation within study areas,
67 and population migration is limited.
68 - Case reports of high or accidental
69 exposure lack definition of the
70 population at risk and the expected
71 number of cases. They can provide
72 information about a rare effect or about
73 the relevance of analogous results in
74 animals.
75 The assessment briefly reviews
76 ecological studies and case reports but
77 reports details only if they suggest effects
78 not identified by other studies.
79 3.3. Selecting pertinent experimental
80 studies
81 Exposure route is a key design
82 consideration for selecting pertinent
83 experimental animal studies or human
84 clinical studies.
85 - Studies of oral, inhalation, or dermal
86 exposure involve passage through an
87 absorption barrier and are considered
88 most pertinent to human environmental
89 exposure.
90 - Injection or implantation studies are
91 often considered less pertinent but may
92 provide valuable toxicokinetic or
93 mechanistic information. They also may
This document is a draft for review purposes only and does not constitute Agency policy.
xiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 be useful for identifying effects in
2 animals if deposition or absorption is
3 problematic (for example, for particles
4 and fibers).
5 Exposure duration is also a key design
6 consideration for selecting pertinent
7 experimental animal studies.
8 - Studies of effects from chronic exposure
9 are most pertinent to lifetime human
10 exposure.
11 - Studies of effects from less-than-chronic
12 exposure are pertinent but less
13 preferred for identifying effects from
14 lifetime human exposure. Such studies
15 may be indicative of effects from less-
16 than-lifetime human exposure.
17 Short-duration studies involving animals
18 or humans may provide toxicokinetic or
19 mechanistic information.
20 For developmental toxicity and
21 reproductive toxicity, irreversible effects
22 may result from a brief exposure during a
23 critical period of development Accordingly,
24 specialized study designs are used for these
25 effects fU.S. EPA. 2006b. 1998. 1996.19911
26 4. Evaluating the quality of
27 individual studies
28 After the subsets of pertinent
29 epidemiologic and experimental studies
30 have been selected from the literature
31 searches, the assessment evaluates the
32 quality of each individual study. This
33 evaluation considers the design, methods,
34 conduct, and documentation of each study,
35 but not whether the results are positive,
36 negative, or null. The objective is to identify
37 the stronger, more informative studies based
38 on a uniform evaluation of quality
39 characteristics across studies of similar
40 design.
41 4.1. Evaluating the quality of
42 epidemiologic studies
43 The assessment evaluates design and
44 methodological aspects that can increase or
45 decrease the weight given to each
46 epidemiologic study in the overall evaluation
47 [U.S. EPA. 2005a. 1998.1996.1994b. 19911:
48
49
50
51
52
53
54
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
- Documentation of study design,
methods, population characteristics, and
results.
Definition and selection of the study
group and comparison group.
Ascertainment of exposure
chemical or mixture.
to the
55 - Ascertainment of disease or health effect.
Duration of exposure and follow-up and
adequacy for assessing the occurrence of
effects.
Characterization
critical periods.
of exposure during
Sample size and statistical power to
detect anticipated effects.
Participation rates and potential for
selection bias as a result of the achieved
participation rates.
- Measurement error (can lead to
misclassification of exposure, health
outcomes, and other factors) and other
types of information bias.
- Potential confounding and other sources
of bias addressed in the study design or
in the analysis of results. The basis for
consideration of confounding is a
reasonable expectation that the
confounder is related to both exposure
and outcome and is sufficiently prevalent
to result in bias.
For developmental toxicity, reproductive
toxicity, neurotoxicity, and cancer there is
further guidance on the nuances of
evaluating epidemiologic studies of these
effects (U.S. EPA. 2005a. 1998.1996. 1991).
83 4.2. Evaluating the quality of
84 experimental studies
85 The assessment evaluates design and
86 methodological aspects that can increase or
87 decrease the weight given to each
This document is a draft for review purposes only and does not constitute Agency policy.
xv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 experimental animal study, in-vitro study, or
2 human clinical study [U.S. EPA. 2005a. 1998.
3 1996, 1991]. Research involving human
4 subjects is considered only if conducted
5 according to ethical principles.
6 - Documentation of study design, animals
7 or study population, methods, basic data,
8 and results.
9 - Nature of the assay and validity for its
10 intended purpose.
11 - Characterization of the nature and extent
12 of impurities and contaminants of the
13 administered chemical or mixture.
14 - Characterization of dose and dosing
15 regimen (including age at exposure) and
16 their adequacy to elicit adverse effects,
17 including latent effects.
18 - Sample sizes and statistical power to
19 detect dose-related differences or trends.
20 - Ascertainment of survival, vital signs,
21 disease or effects, and cause of death.
22 - Control of other variables that could
23 influence the occurrence of effects.
24 The assessment uses statistical tests to
25 evaluate whether the observations may be
26 due to chance. The standard for determining
27 statistical significance of a response is a
28 trend test or comparison of outcomes in the
29 exposed groups against those of concurrent
30 controls. In some situations, examination of
31 historical control data from the same
32 laboratory within a few years of the study
33 may improve the analysis. For an uncommon
34 effect that is not statistically significant
35 compared with concurrent controls,
36 historical controls may show that the effect
37 is unlikely to be due to chance. For a
38 response that appears significant against a
39 concurrent control response that is unusual,
40 historical controls may offer a different
41 interpretation fU.S. EPA. 2005a. §2.2.2.1.3).
42 For developmental toxicity, reproductive
43 toxicity, neurotoxicity, and cancer there is
44 further guidance on the nuances of
45 evaluating experimental studies of these
46 effects [U.S. EPA. 2005a. 1998. 1996. 1991).
47 In multi-generation studies, agents that
48 produce developmental effects at doses that
49 are not toxic to the maternal animal are of
50 special concern. Effects that occur at doses
51 associated with mild maternal toxicity are
52 not assumed to result only from maternal
53 toxicity. Moreover, maternal effects may be
54 reversible, while effects on the offspring may
55 be permanent [U.S. EPA. 1998. §3.1.1.4,
56 1991 §3.1.2.4.5.4).
57 4.3. Reporting study results
58 The assessment uses evidence tables to
59 present the design and key results of
60 pertinent studies. There may be separate
61 tables for each site of toxicity or type of
62 study.
63 If a large number of studies observe the
64 same effect, the assessment considers the
65 study quality characteristics in this section
66 to identify the strongest studies or types of
67 study. The tables present details from these
68 studies, and the assessment explains the
69 reasons for not reporting details of other
70 studies or groups of studies that do not add
71 new information. Supplemental information
72 provides references to all studies
73 considered, including those not summarized
74 in the tables.
75 The assessment discusses strengths and
76 limitations that affect the interpretation of
77 each study. If the interpretation of a study in
78 the assessment differs from that of the study
79 authors, the assessment discusses the basis
80 for the difference.
81 As a check on the selection and
82 evaluation of pertinent studies, the EPA asks
83 peer reviewers to identify studies that were
84 not adequately considered.
85 5. Evaluating the overall evidence
86 of each effect
87 5.1. Concepts of causal inference
88 For each health effect, the assessment
89 evaluates the evidence as a whole to
90 determine whether it is reasonable to infer a
91 causal association between exposure to the
This document is a draft for review purposes only and does not constitute Agency policy.
xvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 agent and the occurrence of the effect This
2 inference is based on information from
3 pertinent human studies, animal studies, and
4 mechanistic studies of adequate quality.
5 Positive, negative, and null results are given
6 weight according to study quality.
7 Causal inference involves scientific
8 judgment, and the considerations are
9 nuanced and complex. Several health
10 agencies have developed frameworks for
11 causal inference, among them the U.S.
12 Surgeon General [CDC. 2004: HEW. 1964).
13 the International Agency for Research on
14 Cancer [2006] , the Institute of Medicine
15 [20081. and the EPA [U.S. EPA. 2010. §1.6,
16 2005a. §2.5). Although developed for
17 different purposes, the frameworks are
18 similar in nature and provide an established
19 structure and language for causal inference.
20 Each considers aspects of an association that
21 suggest causation, discussed by Hill [1965]
22 and elaborated by Rothman and Greenland
23 [1998] [U.S. EPA. 2005a. §2.2.1.7, 1994b.
24 app. C).
25 Strength of association: The finding of a
26 large relative risk with narrow
27 confidence intervals strongly suggests
28 that an association is not due to chance,
29 bias, or other factors. Modest relative
30 risks, however, may reflect a small range
31 of exposures, an agent of low potency, an
32 increase in an effect that is common,
33 exposure misclassification, or other
34 sources of bias.
35 Consistency of association: An inference of
36 causation is strengthened if elevated
37 risks are observed in independent
38 studies of different populations and
39 exposure scenarios. Reproducibility of
40 findings constitutes one of the strongest
41 arguments for causation. Discordant
42 results sometimes reflect differences in
43 study design, exposure, or confounding
44 factors.
45 Specificity of association: As originally
46 intended, this refers to one cause
47 associated with one effect. Current
48 understanding that many agents cause
49 multiple effects and many effects have
50 multiple causes make this a less
51 informative aspect of causation, unless
52 the effect is rare or unlikely to have
53 multiple causes.
54 Temporal relationship: A causal
55 interpretation requires that exposure
56 precede development of the effect.
57 Biologic gradient (exposure-response
58 relationship): Exposure-response
59 relationships strongly suggest causation.
60 A monotonic increase is not the only
61 pattern consistent with causation. The
62 presence of an exposure-response
63 gradient also weighs against bias and
64 confounding as the source of an
65 association.
66 Biologic plausibility: An inference of
67 causation is strengthened by data
68 demonstrating plausible biologic
69 mechanisms, if available. Plausibility
70 may reflect subjective prior beliefs if
71 there is insufficient understanding of the
72 biologic process involved.
73 Coherence: An inference of causation is
74 strengthened by supportive results from
75 animal experiments, toxicokinetic
76 studies, and short-term tests. Coherence
77 may also be found in other lines of
78 evidence, such as changing disease
79 patterns in the population.
80 "Natural experiments": A change in
81 exposure that brings about a change in
82 disease frequency provides strong
83 evidence, as it tests the hypothesis of
84 causation. An example would be an
85 intervention to reduce exposure in the
86 workplace or environment that is
87 followed by a reduction of an adverse
88 effect
89 Analogy: Information on structural
90 analogues or on chemicals that induce
91 similar mechanistic events can provide
92 insight into causation.
93 These considerations are consistent with
94 guidelines for systematic reviews that
This document is a draft for review purposes only and does not constitute Agency policy.
xvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 evaluate the quality and weight of evidence.
2 Confidence is increased if the magnitude of
3 effect is large, if there is evidence of an
4 exposure-response relationship, or if an
5 association was observed and the plausible
6 biases would tend to decrease the magnitude
7 of the reported effect. Confidence is
8 decreased for study limitations,
9 inconsistency of results, indirectness of
10 evidence, imprecision, or reporting bias
11 fGuyattetal.. 2008a: GuyattetaL 2008b).
12 5.2. Evaluating evidence in humans
13 For each effect, the assessment evaluates
14 the evidence from the epidemiologic studies
15 as a whole. The objective is to determine
16 whether a credible association has been
17 observed and, if so, whether that association
18 is consistent with causation. In doing this,
19 the assessment explores alternative
20 explanations (such as chance, bias, and
21 confounding) and draws a conclusion about
22 whether these alternatives can satisfactorily
23 explain any observed association.
24 To make clear how much the
25 epidemiologic evidence contributes to the
26 overall weight of the evidence, the
27 assessment may select a standard descriptor
28 to characterize the epidemiologic evidence
29 of association between exposure to the agent
30 and occurrence of a health effect.
31 Sufficient epidemiologic evidence of an
32 association consistent with causation:
33 The evidence establishes a causal
34 association for which alternative
35 explanations such as chance, bias, and
36 confounding can be ruled out with
37 reasonable confidence.
38 Suggestive epidemiologic evidence of an
39 association consistent with causation:
40 The evidence suggests a causal
41 association but chance, bias, or
42 confounding cannot be ruled out as
43 explaining the association.
44 Inadequate epidemiologic evidence to
45 infer a causal association: The available
46 studies do not permit a conclusion
47 regarding the presence or absence of an
48 association.
49 Epidemiologic evidence consistent with no
50 causal association: Several adequate
51 studies covering the full range of human
52 exposures and considering susceptible
53 populations, and for which alternative
54 explanations such as bias and
55 confounding can be ruled out, are
56 mutually consistent in not finding an
57 association.
58 5.3. Evaluating evidence in animals
59 For each effect, the assessment evaluates
60 the evidence from the animal experiments as
61 a whole to determine the extent to which
62 they indicate a potential for effects in
63 humans. Consistent results across various
64 species and strains increase confidence that
65 similar results would occur in humans.
66 Several concepts discussed by Hill [1965]
67 are pertinent to the weight of experimental
68 results: consistency of response, dose-
69 response relationships, strength of response,
70 biologic plausibility, and coherence [U.S.
71 EPA. 2005a. §2.2.1.7.1994. app. C).
72 In weighing evidence from multiple
73 experiments, U.S. EPA [2005a. §2.5)
74 distinguishes
75 Conflicting evidence (that is, mixed positive
76 and negative results in the same sex and
77 strain using a similar study protocol)
78 from
79 Differing results (that is, positive results
80 and negative results are in different
81 sexes or strains or use different study
82 protocols).
83 Negative or null results do not invalidate
84 positive results in a different experimental
85 system. The EPA regards all as valid
86 observations and looks to explain differing
87 results using mechanistic information (for
88 example, physiologic or metabolic
89 differences across test systems) or
90 methodological differences (for example,
91 relative sensitivity of the tests, differences in
This document is a draft for review purposes only and does not constitute Agency policy.
xviii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 dose levels, insufficient sample size, or
2 timing of dosing or data collection).
3 It is well established that there are
4 critical periods for some developmental and
5 reproductive effects [U.S. EPA. 2006b.
6 2005a. b, 1998. 1996. 19911 Accordingly,
7 the assessment determines whether critical
8 periods have been adequately investigated.
9 Similarly, the assessment determines
10 whether the database is adequate to
11 evaluate other critical sites and effects.
12 In evaluating evidence of genetic
13 toxicity:
14 - Demonstration of gene mutations,
15 chromosome aberrations, or aneuploidy
16 in humans or experimental mammals
17 [in vivo] provides the strongest evidence.
18 - This is followed by positive results in
19 lower organisms or in cultured cells
20 [in vitro] or for other genetic events.
21 - Negative results carry less weight, partly
22 because they cannot exclude the
23 possibility of effects in other tissues
24 flARC. 20061.
25 For germ-cell mutagenicity, The EPA has
26 defined categories of evidence, ranging from
27 positive results of human germ-cell
28 mutagenicity to negative results for all
29 effects of concern [U.S. EPA. 1986a. 52.31.
30 5.4. Evaluating mechanistic data
31 Mechanistic data can be useful in
32 answering several questions.
33 - The biologic plausibility of a causal
34 interpretation of human studies.
35 - The generalizability of animal studies to
36 humans.
37 - The susceptibility of particular
38 populations or lifestages.
39 The focus of the analysis is to describe, if
40 possible, mechanistic pathways that lead to a
41 health effect. These pathways encompass:
42 - Toxicokinetic processes of absorption,
43 distribution, metabolism, and
44 elimination that lead to the formation of
45 an active agent and its presence at the
46 site of initial biologic interaction.
47 - Toxicodynamic processes that lead to a
48 health effect at this or another site (also
49 known as a mode of action].
50 For each effect, the assessment discusses
51 the available information on its modes of
52 action and associated key events [key events
53 being empirically observable, necessary
54 precursor steps or biologic markers of such
55 steps; mode of action being a series of key
56 events involving interaction with cells,
57 operational and anatomic changes, and
58 resulting in disease). Pertinent information
59 may also come from studies of metabolites
60 or of compounds that are structurally similar
61 or that act through similar mechanisms.
62 Information on mode of action is not
63 required for a conclusion that the agent is
64 causally related to an effect [U.S. EPA. 2005a.
65 §2.5).
66 The assessment addresses several
67 questions about each hypothesized mode of
68 action [U.S. EPA. 2005a. 32.4.3.41.
69 1) Is the hypothesized mode of action
70 sufficiently supported in test animals?
71 Strong support for a key event being
72 necessary to a mode of action can come
73 from experimental challenge to the
74 hypothesized mode of action, in which
75 studies that suppress a key event
76 observe suppression of the effect
77 Support for a mode of action is
78 meaningfully strengthened by consistent
79 results in different experimental models,
80 much more so than by replicate
81 experiments in the same model. The
82 assessment may consider various
83 aspects of causation in addressing this
84 question.
85 2) Is the hypothesized mode of action
86 relevant to humans? The assessment
87 reviews the key events to identify critical
88 similarities and differences between the
89 test animals and humans. Site
90 concordance is not assumed between
91 animals and humans, though it may hold
92 for certain effects or modes of action.
This document is a draft for review purposes only and does not constitute Agency policy.
xix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Information suggesting quantitative
2 differences in doses where effects would
3 occur in animals or humans is
4 considered in the dose-response
5 analysis. Current levels of human
6 exposure are not used to rule out human
7 relevance, as IRIS assessments may be
8 used in evaluating new or unforeseen
9 circumstances that may entail higher
10 exposures.
11 3) Which populations or lifestages can
12 be particularly susceptible to the
13 hypothesized mode of action? The
14 assessment reviews the key events to
15 identify populations and lifestages that
16 might be susceptible to their occurrence.
17 Quantitative differences may result in
18 separate toxicity values for susceptible
19 populations or lifestages.
20 The assessment discusses the likelihood
21 that an agent operates through multiple
22 modes of action. An uneven level of support
23 for different modes of action can reflect
24 disproportionate resources spent
25 investigating them (U.S. EPA. 2005a.
26 §2.4.3.3). It should be noted that in clinical
27 reviews, the credibility of a series of studies
28 is reduced if evidence is limited to studies
29 funded by one interested sector [Guyatt et
30 al.. 2008bj.
31 For cancer, the assessment evaluates
32 evidence of a mutagenic mode of action to
33 guide extrapolation to lower doses and
34 consideration of susceptible lifestages. Key
35 data include the ability of the agent or a
36 metabolite to react with or bind to DNA,
37 positive results in multiple test systems, or
38 similar properties and structure-activity
39 relationships to mutagenic carcinogens [U.S.
40 EPA. 2005a. 32.3.51.
41 5.5. Characterizing the overall weight
42 of the evidence
43 After evaluating the human, animal, and
44 mechanistic evidence pertinent to an effect,
45 the assessment answers the question: Does
46 the agent cause the adverse effect? [NRG.
47 2009. 1983]. In doing this, the assessment
48 develops a narrative that integrates the
49 evidence pertinent to causation. To provide
50 clarity and consistency, the narrative
51 includes a standard hazard descriptor. For
52 example, the following standard descriptors
53 combine epidemiologic, experimental, and
54 mechanistic evidence of carcinogenicity [U.S.
55 EPA. 2005a. 32.5).
56 Carcinogenic to humans: There is
57 convincing epidemiologic evidence of a
58 causal association (that is, there is
59 reasonable confidence that the
60 association cannot be fully explained by
61 chance, bias, or confounding); or there is
62 strong human evidence of cancer or its
63 precursors, extensive animal evidence,
64 identification of key precursor events in
65 animals, and strong evidence that they
66 are anticipated to occur in humans.
67 Likely to be carcinogenic to humans: The
68 evidence demonstrates a potential
69 hazard to humans but does not meet the
70 criteria for carcinogenic. There may be a
71 plausible association in humans,
72 multiple positive results in animals, or a
73 combination of human, animal, or other
74 experimental evidence.
75 Suggestive evidence of carcinogenic
76 potential: The evidence raises concern
77 for effects in humans but is not sufficient
78 for a stronger conclusion. This
79 descriptor covers a range of evidence,
80 from a positive result in the only
81 available study to a single positive result
82 in an extensive database that includes
83 negative results in other species.
84 Inadequate information to assess
85 carcinogenic potential: No other
86 descriptors apply. Conflicting evidence
87 can be classified as inadequate
88 information if all positive results are
89 opposed by negative studies of equal
90 quality in the same sex and strain.
91 Differing results, however, can be
92 classified as suggestive evidence or as
93 likely to be carcinogenic.
This document is a draft for review purposes only and does not constitute Agency policy.
xx DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Wot likely to be carcinogenic to humans:
2 There is robust evidence for concluding
3 that there is no basis for concern. There
4 may be no effects in both sexes of at least
5 two appropriate animal species; positive
6 animal results and strong, consistent
7 evidence that each mode of action in
8 animals does not operate in humans; or
9 convincing evidence that effects are not
10 likely by a particular exposure route or
11 below a defined dose.
12 Multiple descriptors may be used if there
13 is evidence that carcinogenic effects differ by
14 dose range or exposure route [U.S. EPA.
15 2005a.§2.51
16 Another example of standard descriptors
17 comes from the EPA's Integrated Science
18 Assessments, which evaluate causation for
19 the effects of the criteria pollutants in
20 ambient air [U.S. EPA. 2010. 51.61.
21 Causal relationship: Sufficient evidence to
22 conclude that there is a causal
23 relationship. Observational studies
24 cannot be explained by plausible
25 alternatives, or they are supported by
26 other lines of evidence, for example,
27 animal studies or mechanistic
28 information.
29 Likely to be a causal relationship:
30 Sufficient evidence that a causal
31 relationship is likely, but important
32 uncertainties remain. For example,
33 observational studies show an
34 association but co-exposures are difficult
35 to address or other lines of evidence are
36 limited or inconsistent; or multiple
37 animal studies from different
38 laboratories demonstrate effects and
39 there are limited or no human data.
40 Suggestive of a causal relationship: At
41 least one high-quality epidemiologic
42 study shows an association but other
43 studies are inconsistent
44 Inadequate to infer a causal relationship:
45 The studies do not permit a conclusion
46 regarding the presence or absence of an
47 association.
48 Not likely to be a causal relationship:
49 Several adequate studies, covering the
50 full range of human exposure and
51 considering susceptible populations, are
52 mutually consistent in not showing an
53 effect at any level of exposure.
54 The EPA is investigating and may on a
55 trial basis use these or other standard
56 descriptors to characterize the overall
57 weight of the evidence for effects other than
58 cancer.
59 6. Selecting studies for derivation
60 of toxicity values
61 For each effect where there is credible
62 evidence of an association with the agent,
63 the assessment derives toxicity values if
64 there are suitable epidemiologic or
65 experimental data. The decision to derive
66 toxicity values may be linked to the hazard
67 descriptor.
68 Dose-response analysis requires
69 quantitative measures of dose and response.
70 Then, other factors being equal:
71 - Epidemiologic studies are preferred over
72 animal studies, if quantitative measures
73 of exposure are available and effects can
74 be attributed to the agent
75 - Among experimental animal models,
76 those that respond most like humans are
77 preferred, if the comparability of
78 response can be determined.
79 - Studies by a route of human
80 environmental exposure are preferred,
81 although a validated toxicokinetic model
82 can be used to extrapolate across
83 exposure routes.
84 - Studies of longer exposure duration and
85 follow-up are preferred, to minimize
86 uncertainty about whether effects are
87 representative of lifetime exposure.
88 - Studies with multiple exposure levels are
89 preferred for their ability to provide
90 information about the shape of the
91 exposure-response curve.
This document is a draft for review purposes only and does not constitute Agency policy.
xxi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 - Studies with adequate power to detect
2 effects at lower exposure levels are
3 preferred, to minimize the extent of
4 extrapolation to levels found in the
5 environment
6 Studies with non-monotonic exposure-
7 response relationships are not necessarily
8 excluded from the analysis. A diminished
9 effect at higher exposure levels may be
10 satisfactorily explained by factors such as
11 competing toxicity, saturation of absorption
12 or metabolism, exposure misclassification,
13 or selection bias.
14 If a large number of studies are suitable
15 for dose-response analysis, the assessment
16 considers the study characteristics in this
17 section to focus on the most informative
18 data. The assessment explains the reasons
19 for not analyzing other groups of studies. As
20 a check on the selection of studies for dose-
21 response analysis, the EPA asks peer
22 reviewers to identify studies that were not
23 adequately considered.
24 7. Deriving toxicity values
25 7.1. General framework for dose-
26 response analysis
27 The EPA uses a two-step approach that
28 distinguishes analysis of the observed dose-
29 response data from inferences about lower
30 doses [U.S. EPA. 2005a. S3).
31 Within the observed range, the preferred
32 approach is to use modeling to incorporate a
33 wide range of data into the analysis. The
34 modeling yields a point of departure (an
35 exposure level near the lower end of the
36 observed range, without significant
37 extrapolation to lower doses) (sections 7.2-
38 7.3).
39 Extrapolation to lower doses considers
40 what is known about the modes of action for
41 each effect (Sections 7.4-7.5). If response
42 estimates at lower doses are not required, an
43 alternative is to derive reference values,
44 which are calculated by applying factors to
45 the point of departure in order to account
46 for sources of uncertainty and variability
47 (section 7.6).
48 For a group of agents that induce an
49 effect through a common mode of action, the
50 dose-response analysis may derive a relative
51 potency factor for each agent A full dose-
52 response analysis is conducted for one well-
53 studied index chemical in the group, then the
54 potencies of other members are expressed in
55 relative terms based on relative toxic effects,
56 relative absorption or metabolic rates,
57 quantitative structure-activity relationships,
58 or receptor binding characteristics (U.S. EPA.
59 2005a. 33.2.6. 2000. 34.41.
60 Increasingly, the EPA is basing toxicity
61 values on combined analyses of multiple
62 data sets or multiple responses. The EPA
63 also considers multiple dose-response
64 approaches if they can be supported by
65 robust data.
66 7.2. Modeling dose to sites of biologic
67 effects
68 The preferred approach for analysis of
69 dose is toxicokinetic modeling because of its
70 ability to incorporate a wide range of data.
71 The preferred dose metric would refer to the
72 active agent at the site of its biologic effect or
73 to a close, reliable surrogate measure. The
74 active agent may be the administered
75 chemical or a metabolite. Confidence in the
76 use of a toxicokinetic model depends on the
77 robustness of its validation process and on
78 the results of sensitivity analyses (U.S. EPA.
79 2006a. 2005a. §3.1.1994b. §4.3).
80 Because toxicokinetic modeling can
81 require many parameters and more data
82 than are typically available, the EPA has
83 developed standard approaches that can be
84 applied to typical data sets. These standard
85 approaches also facilitate comparison across
86 exposure patterns and species.
87 - Intermittent study exposures are
88 standardized to a daily average over the
89 duration of exposure. For chronic effects,
90 daily exposures are averaged over the
91 lifespan. Exposures during a critical
92 period, however, are not averaged over a
This document is a draft for review purposes only and does not constitute Agency policy.
xxii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 longer duration[U.S. EPA. 2005a. §3.1.1,
2 1991. §3.21.
3 - Doses are standardized to equivalent
4 human terms to facilitate comparison of
5 results from different species.
6 - Oral doses are scaled allometrically
7 using mg/kg3/4-d as the equivalent
8 dose metric across species.
9 Allometric scaling pertains to
10 equivalence across species, not
11 across lifestages, and is not used to
12 scale doses from adult humans or
13 mature animals to infants or children
14 fU.S. EPA. 2011. 2005a. §3.1.31
15 - Inhalation exposures are scaled
16 using dosimetry models that apply
17 species-specific physiologic and
18 anatomic factors and consider
19 whether the effect occurs at the site
20 of first contact or after systemic
21 circulation [U.S. EPA. 2012a. 1994b.
22 §3).
23 It can be informative to convert doses
24 across exposure routes. If this is done, the
25 assessment describes the underlying data,
26 algorithms, and assumptions [U.S. EPA,
27 2005a. 33.1.41.
28 In the absence of study-specific data on,
29 for example, intake rates or body weight, the
30 EPA has developed recommended values for
31 use in dose-response analysis [U.S. EPA,
32 19881
33 7.3. Modeling response in the range
34 of observation
35 Toxicodynamic ("biologically based")
36 modeling can incorporate data on biologic
37 processes leading to an effect Such models
38 require sufficient data to ascertain a mode of
39 action and to quantitatively support model
40 parameters associated with its key events.
41 Because different models may provide
42 equivalent fits to the observed data but
43 diverge substantially at lower doses, critical
44 biologic parameters should be measured
45 from laboratory studies, not by model fitting.
46 Confidence in the use of a toxicodynamic
47 model depends on the robustness of its
48 validation process and on the results of
49 sensitivity analyses. Peer review of the
50 scientific basis and performance of a model
51 is essential [U.S. EPA. 2005a. 33.2.21.
52 Because toxicodynamic modeling can
53 require many parameters and more
54 knowledge and data than are typically
55 available, the EPA has developed a standard
56 set of empirical ("curve-fitting") models
57 (http://www.epa.gov/ncea/bmds/) that can
58 be applied to typical data sets, including
59 those that are nonlinear. The EPA has also
60 developed guidance on modeling dose-
61 response data, assessing model fit, selecting
62 suitable models, and reporting modeling
63 results (U.S. EPA. 2012b). Additional
64 judgment or alternative analyses are used if
65 the procedure fails to yield reliable results,
66 for example, if the fit is poor, modeling may
67 be restricted to the lower doses, especially if
68 there is competing toxicity at higher doses
69 fU.S. EPA. 2005a. 33.2.31.
70 Modeling is used to derive a point of
71 departure (U.S. EPA. 2012b. 2005a. §3.2.4).
72 (See section 7.6 for alternatives if a point of
73 departure cannot be derived by modeling.)
74 - If linear extrapolation is used, selection
75 of a response level corresponding to the
76 point of departure is not highly
77 influential, so standard values near the
78 low end of the observable range are
79 generally used (for example, 10% extra
80 risk for cancer bioassay data, 1% for
81 epidemiologic data, lower for rare
82 cancers).
83 - For nonlinear approaches, both
84 statistical and biologic considerations
85 are taken into account.
86
87
88
89
90
91
92
93
94
95
For dichotomous data, a response
level of 10% extra risk is generally
used for minimally adverse effects,
5% or lower for more severe effects.
For continuous data, a response level
is ideally based on an established
definition of biologic significance. In
the absence of such definition, one
control standard deviation from the
control mean is often used for
This document is a draft for review purposes only and does not constitute Agency policy.
xxiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
minimally adverse effects, one-half
standard deviation for more severe
effects.
4 The point of departure is the 95% lower
5 bound on the dose associated with the
6 selected response level.
7 7.4. Extrapolating to lower doses and
8 response levels
9 The purpose of extrapolating to lower
10 doses is to estimate responses at exposures
11 below the observed data. Low-dose
12 extrapolation, typically used for cancer data,
13 considers what is known about modes of
14 action fU.S. EPA. 2005a. §3.3.1. §3.3.21
15 1) If a biologically based model has been
16 developed and validated for the agent,
17 extrapolation may use the fitted model
18 below the observed range if significant
19 model uncertainty can be ruled out with
20 reasonable confidence.
21 2) Linear extrapolation is used if the dose-
22 response curve is expected to have a
23 linear component below the point of
24 departure. This includes:
25 - Agents or their metabolites that are
26 DN A-re active and have direct
27 mutagenic activity.
28 - Agents or their metabolites for which
29 human exposures or body burdens
30 are near doses associated with key
31 events leading to an effect
32 Linear extrapolation is also used when
33 data are insufficient to establish mode of
34 action and when scientifically plausible.
35 The result of linear extrapolation is
36 described by an oral slope factor or an
37 inhalation unit risk, which is the slope of
38 the dose-response curve at lower doses
39 or concentrations, respectively.
40 3) Nonlinear models are used for
41 extrapolation if there are sufficient data
42 to ascertain the mode of action and to
43 conclude that it is not linear at lower
44 doses, and the agent does not
45 demonstrate mutagenic or other activity
46 consistent with linearity at lower doses.
47 Nonlinear approaches generally should
48 not be used in cases where mode of
49 action has not ascertained. If nonlinear
50 extrapolation is appropriate but no
51 model is developed, an alternative is to
52 calculate reference values.
53 4) Both linear and nonlinear approaches
54 may be used if there a multiple modes of
55 action. For example, modeling to a low
56 response level can be useful for
57 estimating the response at doses where a
58 high-dose mode of action would be less
59 important.
60 If linear extrapolation is used, the
61 assessment develops a candidate slope
62 factor or unit risk for each suitable data set
63 These results are arrayed, using common
64 dose metrics, to show the distribution of
65 relative potency across various effects and
66 experimental systems. The assessment then
67 derives or selects an overall slope factor and
68 an overall unit risk for the agent, considering
69 the various dose-response analyses, the
70 study preferences discussed in section 6, and
71 the possibility of basing a more robust result
72 on multiple data sets.
73 7.5. Considering susceptible
74 populations and lifestages
75 The assessment analyzes the available
76 information on populations and lifestages
77 that may be particularly susceptible to each
78 effect A tiered approach is used [U.S. EPA,
79 2005a. 33.51.
80 1) If an epidemiologic or experimental
81 study reports quantitative results for a
82 susceptible population or lifestage, these
83 data are analyzed to derive separate
84 toxicity values for susceptible
85 individuals.
86 2) If data on risk-related parameters allow
87 comparison of the general population
88 and susceptible individuals, these data
89 are used to adjust the general-population
90 toxicity values for application to
91 susceptible individuals.
This document is a draft for review purposes only and does not constitute Agency policy.
xxiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 3) In the absence of chemical-specific data,
2 the EPA has developed age-dependent
3 adjustment factors for early-life exposure
4 to potential carcinogens that have a
5 mutagenic mode of action. There is
6 evidence of early-life susceptibility to
7 various carcinogenic agents, but most
8 epidemiologic studies and cancer
9 bioassays do not include early-life
10 exposure. To address the potential for
11 early-life susceptibility, the EPA
12 recommends [U.S. EPA. 2005b. 55):
13 - 10-fold adjustment for exposures
14 before age 2 years.
15 - 3-fold adjustment for exposures
16 between ages 2 and 16 years.
17 7.6. Reference values and uncertainty
18 factors
19 An oral reference dose or an inhalation
20 reference concentration is an estimate of an
21 exposure (including in susceptible
22 subgroups) that is likely to be without an
23 appreciable risk of adverse health effects
24 over a lifetime (U.S. EPA. 2002. §4.2).
25 Reference values are typically calculated for
26 effects other than cancer and for suspected
27 carcinogens if a well characterized mode of
28 action indicates that a necessary key event
29 does not occur below a specific dose.
30 Reference values provide no information
31 about risks at higher exposure levels.
32 The assessment characterizes effects
33 that form the basis for reference values as
34 adverse, considered to be adverse, or a
35 precursor to an adverse effect For
36 developmental toxicity, reproductive
37 toxicity, and neurotoxicity there is guidance
38 on adverse effects and their biologic markers
39 fU.S. EPA. 1998.1996.19911
40 To account for uncertainty and
41 variability in the derivation of a lifetime
42 human exposure where adverse effects are
43 not anticipated to occur, reference values are
44 calculated by applying a series of uncertainty
45 factors to the point of departure. If a point of
46 departure cannot be derived by modeling, a
47 no-observed-adverse-effect level or a
48 lowest-observed-adverse-effect level is used
49 instead. The assessment discusses scientific
50 considerations involving several areas of
51 variability or uncertainty.
52 Human variation. The assessment accounts
53 for variation in susceptibility across the
54 human population and the possibility
55 that the available data may not be
56 representative of individuals who are
57 most susceptible to the effect. A factor of
58 10 is generally used to account for this
59 variation. This factor is reduced only if
60 the point of departure is derived or
61 adjusted specifically for susceptible
62 individuals (not for a general population
63 that includes both susceptible and non-
64 susceptible individuals) (U.S. EPA. 2002.
65 §4.4.5, 1998. §4.2, 1996. §4, 1994b.
66 §4.3.9.1.1991. §3.4).
67 Animal-to-human extrapolation. If animal
68 results are used to make inferences
69 about humans, the assessment adjusts
70 for cross-species differences. These may
71 arise from differences in toxicokinetics
72 or toxicodynamics. Accordingly, if the
73 point of departure is standardized to
74 equivalent human terms or is based on
75 toxicokinetic or dosimetry modeling, a
76 factor of lO1^ (rounded to 3) is applied
77 to account for the remaining uncertainty
78 involving toxicokinetic and
79 toxicodynamic differences. If a
80 biologically based model adjusts fully for
81 toxicokinetic and toxicodynamic
82 differences across species, this factor is
83 not used. In most other cases, a factor of
84 10 is applied (U.S. EPA. 2011. 2002.
85 §4.4.5, 1998. §4.2, 1996. §4, 1994b.
86 §4.3.9.1,1991 §3.4).
87 Adverse-effect level to no-observed-
88 adverse-effect level. If a point of
89 departure is based on a lowest-
90 observed-adverse-effect level, the
91 assessment must infer a dose where
92 such effects are not expected. This can be
93 a matter of great uncertainty, especially
94 if there is no evidence available at lower
95 doses. A factor of 10 is applied to
96 account for the uncertainty in making
This document is a draft for review purposes only and does not constitute Agency policy.
xxv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 this inference. A factor other than 10
2 may be used, depending on the
3 magnitude and nature of the response
4 and the shape of the dose-response
5 curve [U.S. EPA. 2002. §4.4.5, 1998. §4.2,
6 1996, §4,1994b, §4.3.9.1,1991, §3.4).
7 Subchronic-to-chronic exposure. If a point
8 of departure is based on subchronic
9 studies, the assessment considers
10 whether lifetime exposure could have
11 effects at lower levels of exposure. A
12 factor of 10 is applied to account for the
13 uncertainty in using subchronic studies
14 to make inferences about lifetime
15 exposure. This factor may also be
16 applied for developmental or
17 reproductive effects if exposure covered
18 less than the full critical period. A factor
19 other than 10 may be used, depending
20 on the duration of the studies and the
21 nature of the response (U.S. EPA. 2002.
22 §4.4.5,1998, §4.2,1994b, §4.3.9.1).
23 Incomplete database. If an incomplete
24 database raises concern that further
25 studies might identify a more sensitive
26 effect, organ system, or lifestage, the
27 assessment may apply a database
28 uncertainty factor [U.S. EPA.
29 2002334.4.5. 1998. §4.2, 1996. §4,
30 1994b. §4.3.9.1, 1991. §3.4). The size of
31 the factor depends on the nature of the
32 database deficiency. For example, the
33 EPA typically follows the suggestion that
34 a factor of 10 be applied if both a
35 prenatal toxicity study and a two-
36 generation reproduction study are
37 missing and a factor of 101/2 if either is
38 missing [U.S. EPA. 2002. §4.4.5).
39 In this way, the assessment derives
40 candidate values for each suitable data set
41 and effect that is credibly associated with the
42 agent These results are arrayed, using
43 common dose metrics, to show where effects
44 occur across a range of exposures [U.S. EPA.
45 1994b. 34.3.91.
46 The assessment derives or selects an
47 organ- or system-specific reference value for
48 each organ or system affected by the agent.
49 The assessment explains the rationale for
50 each organ/system-specific reference value
51 (based on, for example, the highest quality
52 studies, the most sensitive outcome, or a
53 clustering of values). By providing these
54 organ/system-specific reference values, IRIS
55 assessments facilitate subsequent
56 cumulative risk assessments that consider
57 the combined effect of multiple agents acting
58 at a common site or through common
59 mechanisms fNRC. 20091.
60 The assessment then selects an overall
61 reference dose and an overall reference
62 concentration for the agent to represent
63 lifetime human exposure levels where
64 effects are not anticipated to occur. This is
65 generally the most sensitive organ/system-
66 specific reference value, though
67 consideration of study quality and
68 confidence in each value may lead to a
69 different selection.
70 7.7. Confidence and uncertainty in the
71 reference values
72 The assessment selects a standard
73 descriptor to characterize the level of
74 confidence in each reference value, based on
75 the likelihood that the value would change
76 with further testing. Confidence in reference
77 values is based on quality of the studies used
78 and completeness of the database, with more
79 weight given to the latter. The level of
80 confidence is increased for reference values
81 based on human data supported by animal
82 data [U.S. EPA. 1994b. §4.3.9.2).
83 High confidence: The reference value is not
84 likely to change with further testing,
85 except for mechanistic studies that might
86 affect the interpretation of prior test
87 results.
88 Medium confidence: This is a matter of
89 judgment, between high and low
90 confidence.
91 Low confidence: The reference value is
92 especially vulnerable to change with
93 further testing.
This document is a draft for review purposes only and does not constitute Agency policy.
xxvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 These criteria are consistent with
2 guidelines for systematic reviews that
3 evaluate the quality of evidence. These also
4 focus on whether further research would be
5 likely to change confidence in the estimate of
6 effect (Guyatt etal..2008a).
7 All assessments discuss the significant
8 uncertainties encountered in the analysis.
9 The EPA provides guidance on
10 characterization of uncertainty [U.S. EPA.
11 2005a. §3.6). For example, the discussion
12 distinguishes model uncertainty (lack of
13 knowledge about the most appropriate
14 experimental or analytic model) and
15 parameter uncertainty (lack of knowledge
16 about the parameters of a model).
17 Assessments also discuss human variation
18 (interpersonal differences in biologic
19 susceptibility or in exposures that modify
20 the effects of the agent).
21 References
22 CDC. (Centers for Disease Control and
23 Prevention). (2004). The health
24 consequences of smoking: A report of the
25 Surgeon General. Washington, DC: U.S.
26 Department of Health and Human
27 Services.
28 http://www.surgeongeneral.gov/library
29 /smokingconsequences/
30 Guyatt. GH: Oxman. AD: Vist. GE: Kunz. R:
31 Falck-Ytter. Y: Alonso-Coello. P:
32 Schiinemann. HI. (2008a). GRADE: An
33 emerging consensus on rating quality of
34 evidence and strength of
35 recommendations. BMJ 336: 924-926.
36 http://dx.doi.0rg/10.1136/bmj.39489.4
37 70347.AD
38 Guyatt. GH: Oxman. AD: Kunz. R: Vist. GE:
39 Falck-Ytter. Y: Schiinemann. HI. (2008b).
40 GRADE: What is "quality of evidence"
41 and why is it important to clinicians?
42 [Review]. BMJ 336: 995-998.
43 http://dx.doi.0rg/10.1136/bmj.39490.5
44 51019.BE
45 HEW. (U.S. Department of Health, Education
46 and Welfare). (1964). Smoking and
47 health: Report of the advisory committee
48 to the surgeon general of the public
49 health service. Washington, DC: U.S.
50 Department of Health, Education, and
51 Welfare.
52 http://profiles.nlm.nih.gov/ps/retrieve/
53 ResourceMetadata/NNBBMQ
54 Hill. AB. (1965). The environment and
55 disease: Association or causation? Proc R
56 SocMed 58: 295-300.
57 I ARC. (International Agency for Research on
58 Cancer). (2006). Preamble to the IARC
59 monographs. Lyon, France.
60 http://monographs.iarc.fr/ENG/Preamb
61 le/
62 IOM. (Institute of Medicine), (2008)
63 Improving the presumptive disability
64 decision-making process for veterans. In
65 JM Samet; CC Bodurow (Eds.).
66 Washington, DC: National Academies
67 Press.
68 NRC. (National Research Council). (1983).
69 Risk assessment in the federal
70 government: Managing the process.
71 Washington, DC: National Academies
72 Press.
73 NRC. (National Research Council). (2009).
74 Science and decisions: Advancing risk
75 assessment Washington, DC: National
76 Academies Press.
77 Rothman. KT: Greenland. S. (1998). Modern
78 epidemiology (2nd ed.). Philadelphia, PA:
79 Lippincott, Williams, & Wilkins.
80 U.S. EPA. (U.S. Environmental Protection
81 Agency). (1986a). Guidelines for
82 mutagenicity risk assessment [EPA
83 Report]. (EPA/630/R-98/003).
84 Washington, DC.
85 http://www.epa.gov/iris/backgrd.html
86 U.S. EPA. (U.S. Environmental Protection
87 Agency). (1986b). Guidelines for the
88 health risk assessment of chemical
89 mixtures. Fed Reg 51: 34014-34025.
90 U.S. EPA. (U.S. Environmental Protection
91 Agency). (1988). Recommendations for
92 and documentation of biological values
This document is a draft for review purposes only and does not constitute Agency policy.
xxvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 for use in risk assessment [EPA Report].
2 (EPA/600/6-87/008). Cincinnati, OH.
3 http://cfpub.epa.gov/ncea/cfm/recordis
4 play.cfm?deid=34855
5 U.S. EPA. (U.S. Environmental Protection
6 Agency). (1991). Guidelines for
7 developmental toxicity risk assessment
8 [EPA Report]. (EPA/600/FR-91/001).
9 Washington, DC: U.S. Environmental
10 Protection Agency, Risk Assessment
11 Forum.
12 http: //www. ep a. go v/ir is/backgr d. html
13 U.S. EPA. (U.S. Environmental Protection
14 Agency). (1994b). Methods for
15 derivation of inhalation reference
16 concentrations and application of
17 inhalation dosimetry [EPA Report].
18 (EPA/600/8-90/066F). Research
19 Triangle Park, NC.
20 http://cfpub.epa.gov/ncea/cfm/recordis
21 play.cfm?deid=71993
22 U.S. EPA. (U.S. Environmental Protection
23 Agency). (1996). Guidelines for
24 reproductive toxicity risk assessment
25 [EPA Report]. (EPA/630/R-96/009).
26 Washington, DC.
27 http://www.epa.gOV/raf/publications/p
28 dfs/REPR051.PDF
29 U.S. EPA. (U.S. Environmental Protection
30 Agency). (1998). Guidelines for
31 neurotoxicity risk assessment [EPA
32 Report]. (EPA/630/R-95/001F).
33 Washington, DC.
34 http://www.epa.gOV/raf/publications/p
35 dfs/NEUROTOX.PDF
36 U.S. EPA. (U.S. Environmental Protection
37 Agency). (2000). Supplementary
38 guidance for conducting health risk
39 assessment of chemical mixtures [EPA
40 Report]. (EPA/630/R-00/002).
81 U.S. EPA. (U.S. Environmental Protection
82 Agency). (2009). EPAs Integrated Risk
83 Information System: Assessment
84 development process [EPA Report].
85 Washington, DC.
86 http://epa.gov/iris/process.htm
41 Washington, DC.
42 http://cfpub.epa.gov/ncea/cfm/recordis
43 play.cfm?deid=20533
44 U.S. EPA. (U.S. Environmental Protection
45 Agency). (2002). A review of the
46 reference dose and reference
47 concentration processes [EPA Report].
48 (EPA/630/P-02/002F). Washington, DC.
49 http://cfpub.epa.gov/ncea/cfm/recordis
50 play.cfm?deid=51717
51 U.S. EPA. (U.S. Environmental Protection
52 Agency). (2 005 a). Guidelines for
53 carcinogen risk assessment [EPA
54 Report]. (EPA/630/P-03/001F).
55 Washington, DC.
56 http://www.epa.gov/cancerguidelines/
57 U.S. EPA. (U.S. Environmental Protection
58 Agency). (2005b). Supplemental
59 guidance for assessing susceptibility
60 from early-life exposure to carcinogens
61 [EPA Report] (Vol. 113). (EPA/630/R-
62 03/003F). Washington, DC.
63 http://www.epa.gOV/cancerguidelines/g
64 uidelines-carcinogen-supplementhtm
65 U.S. EPA. (U.S. Environmental Protection
66 Agency). (2006a). Approaches for the
67 application of physiologically based
68 pharmacokinetic (PBPK) models and
69 supporting data in risk assessment (Final
70 Report) [EPA Report]. (EPA/600/R-
71 05/043F). Washington, DC.
72 http://cfpub.epa.gov/ncea/cfm/recordis
73 play.cfm?deid=157668
74 U.S. EPA. (U.S. Environmental Protection
75 Agency). (2006b). A framework for
76 assessing health risk of environmental
77 exposures to children [EPA Report].
78 (EPA/600/R-05/093F). Washington, DC.
79 http://cfpub.epa.gov/ncea/cfm/recordis
80 play.cfm?deid=158363
87 U.S. EPA. (U.S. Environmental Protection
88 Agency). (2010). Integrated science
89 assessment for carbon monoxide [EPA
90 Report]. (EPA/600/R-09/019F).
91 Research Triangle Park, NC.
92 http://cfpub.epa.gov/ncea/cfm/recordis
93 play.cfm?deid=218686
This document is a draft for review purposes only and does not constitute Agency policy.
xxviii DRAFT—DO NOT CITE OR QUOTE
-------
1 U.S. EPA. (U.S. Environmental Protection
2 Agency). (2011). Recommended use of
3 body weight 3/4 as the default method
4 in derivation of the oral reference dose
5 [EPA Report]. (EPA/100/R11/0001).
6 Washington, DC.
7 http://www.epa.gOV/raf/publications/i
8 nterspecies-extrapolation.htm
9
10
11
23
U.S. EPA. (U.S. Environmental Protection
Agency). (2012a). Advances in inhalation
gas dosimetry for derivation of a
Toxicological Review of Ammonia
12 reference concentration (rfc) and use in
13 risk assessment [EPA Report].
14 (EPA/600/R-12/044). Washington, DC.
15 http://cfpub.epa.gov/ncea/cfm/recordis
16 play.cfm?deid=244650
17 U.S. EPA. (U.S. Environmental Protection
18 Agency). (2012b). Benchmark dose
19 technical guidance. (EPA/100/R-
20 12/001). Washington, DC.
21 http://www.epa.gOV/raf/publications/p
22 dfs/benchmark_dose_guidance.pdf
24 August 2013
25
This document is a draft for review purposes only and does not constitute Agency policy.
xxix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
EXECUTIVE SUMMARY
4
5
6 Occurren ce an d Health Effects
1
8 Ammonia occurs naturally in air, soil, and water and is produced by humans
9 and other animals as part of normal biological processes. Ammonia is also used as
10 an agricultural fertilizer. Exposure to ammonia occurs primarily through breathing
11 air containing ammonia gas, and may also occur via diet or direct skin contact
12 Health effects observed at levels exceeding naturally-occurring
13 concentrations are generally limited to the site of direct contact with ammonia
14 (skin, eyes, respiratory tract, and digestive tract). Short-term exposure to high
15 levels of ammonia in humans can cause irritation and serious burns on the skin and
16 in the mouth, lungs, and eyes. Chronic exposure to airborne ammonia can increase
17 the risk of respiratory irritation, cough, wheezing, tightness in the chest, and
18 reduction in the normal function of the lung in humans. Studies in experimental
19 animals similarly suggest that breathing ammonia at sufficiently high
20 concentrations can result in effects on the respiratory system. Animal studies also
21 suggest that exposure to high levels of ammonia in air or water may adversely affect
22 other organs, such as the stomach, liver, adrenal gland, kidney, and spleen. There is
23 inadequate information to evaluate the carcinogenicity of ammonia.
24
25 Effects Other Than Cancer Observed Following Oral Exposure
26 There are few oral toxicity studies for ammonia. Gastric toxicity may be a hazard for
27 ammonia based on evidence from case reports in humans and mechanistic studies in experimental
28 animals. Evidence in humans is limited to case reports of individuals suffering from
29 gastrointestinal effects from ingesting household cleaning solutions containing ammonia or from
30 biting into capsules of ammonia smelling salts; the relevance of these acute findings to chronic, low-
31 level ammonia exposure is unclear. The experimental animal toxicity database for ammonia lacks
32 standard toxicity studies that evaluate a range of tissues/organs and endpoints. In rats,
33 gastrointestinal effects, characterized as increased epithelial cell migration in the mucosa of the
34 stomach leading to decreased thickness of the gastric mucosa, were reported following short-term
35 and subchronic exposures to ammonia via ingestion [Hataetal., 1994: Tsujii etal., 1993: Kawano et
36 al.. 1991). While these studies provide consistent evidence of changes in the gastric mucosa
37 associated with exposure to ammonia in drinking water, the investigators reported no evidence of
38 microscopic lesions, gastritis, or ulceration in the stomachs of these rats.
39 Given the limited scope of toxicity testing of ingested ammonia and questions concerning
40 the adversity of the gastric mucosal findings in rats, the available oral database for ammonia was
This document is a draft for review purposes only and does not constitute Agency policy.
xxx DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 considered insufficient to characterize toxicity outcomes and dose-response relationships, and an
2 oral reference dose (RfD) for ammonia was not derived.
o
J
4 Effects Other Than Cancer Observed Following Inhalation Exposure
5 Respiratory effects have been identified as a hazard following inhalation exposure to
6 ammonia. Evidence for respiratory toxicity associated with inhaled ammonia comes from studies
7 in humans and animals. Cross-sectional occupational studies involving chronic exposure to
8 ammonia in industrial settings provide evidence of an increased prevalence of respiratory
9 symptoms [Rahman etal., 2007: Ballal etal., 1998] and decreased lung function [Rahman etal.,
10 2007: Ali etal.. 2001: Ballal etal.. 1998: Bhat and Ramaswamy. 19931 Other occupational studies
11 of exposure to ammonia when used as a disinfectant or cleaning product, for example in health care
12 workers and cleaning workers, provide additional evidence of effects on asthma, asthma symptoms,
13 and pulmonary function, using a variety of study designs [Arif and Delclos, 2012: Dumas etal.,
14 2012: Lemiereetal.,2012: Vizcaya etal.. 2011: Zock etal.. 2007: Medina-Ramon etal.. 2006:
15 Medina-Ramon et al., 2005]. Additional evidence of respiratory effects of ammonia is seen in
16 studies of pulmonary function in livestock workers, specifically in the studies that accounted for
17 effects of co-exposures to other agents such as endotoxin and dust [Donham etal., 2000: Reynolds
18 etal.. 1996: Donham etal.. 1995: Preller etal.. 1995: Heederiketal.. 1990]. Controlled volunteer
19 studies of ammonia inhalation and case reports of injury in humans with inhalation exposure to
20 ammonia provide support for the respiratory system as a target of ammonia toxicity. Additionally,
21 respiratory effects were observed in several animal species following short-term and subchronic
22 inhalation exposures to ammonia.
23 The experimental toxicology literature for ammonia also provides evidence that inhaled
24 ammonia may be associated with toxicity to target organs other than the respiratory system,
25 including the liver, adrenal gland, kidney, spleen, heart, and immune system, at concentrations
26 higher than those associated with respiratory system effects. Little evidence exists for these effects
27 relative to the evidence for respiratory effects.
28
This document is a draft for review purposes only and does not constitute Agency policy.
xxxi DRAFT—DO NOT CITE OR QUOTE
-------
1
2
o
J
4
Toxicological Review of Ammonia
Inhalation Reference Concentration (RfC) for Effects Other Than Cancer
Table ES-1. Summary of reference concentration (RfC) derivation
Critical effect
Decreased lung function and
respiratory symptoms
Occupational epidemiology studies
Holness et al. (1989), supported by
Rahman et al. (2007), Ballal et al.
(1998), and AN et al. (2001)
Point of departure3
NOAELADJ: 3.1 mg/m3
UF
10
Chronic RfC
0.3 mg/m3
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
aBecause the study involved workplace exposure conditions, the NOAEL of 8.8 mg/m3 was adjusted for
continuous exposure based on the ratio of VEho (human occupational default minute volume of 10 m breathed
during an 8-hour workday) to VEh (human ambient default minute volume of 20 m3 breathed during the entire
day) and an exposure of 5 days out of 7 days.
NOAEL = no-observed-adverse-effect level; U F = uncertainty factor
The study of ammonia exposure in workers in a soda ash plant by Holness etal. [1989],
with support from three studies in urea fertilizer plants by Rahman et al. [2007], Ballal et al.
[1998], and Ali etal. [2001], was identified as the principal study for RfC derivation. Respiratory
effects, characterized as increased respiratory symptoms (including cough, wheezing, and other
asthma-related symptoms] and decreased lung function in workers exposed to ammonia, were
selected as the critical effect. Holness etal. [1989] found no differences in the prevalence of
respiratory symptoms or lung function between workers (mean exposure 6.5 mg/m3] and the
control group, and no differences when stratified by exposure level (highest exposure group,
>8.8 mg/m3]. Rahman etal. [2007] observed an increased prevalence of respiratory symptoms and
decreased lung function in workers exposed in a plant with a mean ammonia concentration of
18.5 mg/m3, but not in workers in a second plant exposed to a mean concentration of 4.9 mg/m3.
Ballal etal. [1998] observed an increased prevalence of respiratory symptoms among workers in
one factory with exposures ranging from 2 to 27.1 mg/m3,1 but no increase in another factory with
exposures ranging from 0.02-7 mg/m3. A companion study by Ali etal. [2001] also observed
decreased lung function among workers in the higher exposure factory.
Considerations in selecting the principal study for RfC derivation include the higher
confidence placed in the measures of ammonia exposure in Holness etal. [1989], evaluation of both
respiratory symptoms and lung function parameters in this study, and the fact that the estimate of
the no-observed-adverse-effect level (NOAEL] for respiratory effects of 8.8 mg/m3 from Holness et
al. [1989] was the highest of the studies with adequate exposure-response information. Because a
high level of control of exposures in the plant studied by Holness etal. [1989] resulted in relatively
IThis concentration range does not include exposures in the urea store (number of employees = 6; range of
ammonia concentrations = 90-130.4 mg/m3] because employees in this area were required to wear full
protective clothing, thus minimizing potential exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxii DRAFT— DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 low ammonia levels in this facility, the Holness etal. [1989] study does not demonstrate a
2 relationship between ammonia exposure and respiratory effects. Therefore, the Holness et al.
3 [1989] study is identified as the principal study only as part of a collection of epidemiology studies
4 of industrial settings that includes studies with higher workplace ammonia concentrations in which
5 respiratory effects were observed.
6 In summary, the study of ammonia exposure in workers in a soda ash plant by Holness etal.
7 [1989] was identified as the principal study for RfC derivation, with support from Rahman et al.
8 [2007]. Ballaletal. [1998]. and Ali etal. [2001]. and respiratory effects were identified as the
9 critical effect The NOAEL of 8.8 mg/m3 (NOAELADj = 3.1 mg/m3, i.e., adjusted to continuous
10 exposure] from the Holness etal. [1989] study was used as the point of departure [POD] for RfC
11 derivation.
12 An RfC of 0.3 mg/m3 was calculated by dividing the POD (adjusted for continuous
13 exposure, i.e., NOAELADj] by a composite uncertainty factor [UF] of 10 to account for potentially
14 susceptible individuals in the absence of data evaluating variability of response to inhaled ammonia
15 in the human population.
16
17 Confidence in the Chronic Inhalation RfC
18
19 Study - medium
20 Database - medium
21 RfC - medium
22
23 Consistent with EPA's Methods for Derivation of Inhalation Reference Concentrations and
24 Application of Inhalation Dosimetry [U.S. EPA. 1994]. the overall confidence in the RfC is medium
25 and reflects medium confidence in the principal study (adequate design, conduct, and reporting of
26 the principal study; limited by small sample size and identification of a NOAEL only] and medium
27 confidence in the database, which includes occupational and volunteer studies and studies in
28 animals that are mostly of subchronic duration. There are no studies of developmental toxicity, and
29 studies of reproductive and other systemic endpoints are limited; however, reproductive,
30 developmental, and other systemic effects are not expected at the RfC because it is well
31 documented that ammonia is endogenously produced in humans and animals, ammonia
32 concentrations in blood are homeostatically regulated to remain at low levels, and ammonia
33 concentrations in air at the POD are not expected to alter homeostasis.
34
35 Evidence for Carcinogenicity
36 Consistent with EPA's Guidelines for Carcinogen Risk Assessment [U.S. EPA, 2005a], there is
37 "inadequate information to assess carcinogenic potential" for ammonia, based on the absence
38 of ammonia carcinogenicity studies in humans and a single lifetime drinking water study of
39 ammonia in mice Toth [1972] that showed no evidence of carcinogenic potential. There is limited
40 evidence that ammonia may act as a cancer promoter [Tsujiietal.. 1995: Tsujiietal.. 1992b]. The
This document is a draft for review purposes only and does not constitute Agency policy.
xxxiii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 available genotoxicity studies are inadequate to characterize the genotoxic potential of ammonia. A
2 quantitative cancer assessment for ammonia was not conducted.
o
J
4 Susceptible Populations and Lifestages
5 Hyperammonemia is a condition of elevated levels of circulating ammonia that can occur in
6 individuals with severe diseases of the liver or kidney or with hereditary urea [CO(NH2)2] cycle
7 disorders. These elevated ammonia levels can predispose an individual to encephalopathy due to
8 the ability of ammonia to cross the blood-brain barrier; these effects are especially marked in
9 newborn infants. Thus, individuals with disease conditions that lead to hyperammonemia may be
10 more susceptible to the effects of ammonia from external sources, but there are no studies that
11 specifically support this susceptibility.
12 Studies of the toxicity of ammonia in children or young animals compared to other
13 lifestages that would support an evaluation of childhood susceptibility have not been conducted.
14
15 Key Issues Addressed in Assessment
16 Endogenous Ammonia
17 Ammonia, which is produced endogenously, has been detected in the expired air of healthy
18 volunteers. Ammonia concentrations in breath exhaled from the mouth or oral cavity (0.085-
19 2.1 mg/m3) are higher and more variable than concentrations measured in breath exhaled from the
20 nose and trachea (0.0092-0.1 mg/m3) (Appendix E, Section E.I (Elimination) and Table E-l).
21 Concentrations exhaled from the mouth and oral cavity are largely attributed to the production of
22 ammonia via bacterial degradation of food protein in the oral cavity or gastrointestinal tract, and
23 can be influenced by factors such as diet, oral hygiene, and age. In contrast, the lower ammonia
24 concentrations measured in breath exhaled from the nose and trachea appear to better represent
25 levels at the alveolar interface of the lung or in the tracheo-bronchial region and are thought to be
26 more relevant to understanding systemic levels of ammonia than ammonia in breath exhaled from
27 the mouth.
28 The studies of ammonia in exhaled breath were conducted in environments with
29 measureable levels of ambient (exogenous) ammonia and not in ammonia-free environments.
30 Because concentrations of trace compounds in exhaled breath may be correlated with their
31 ambient concentrations (e.g., Spaneletal. (2013) found that approximately 70% of inhaled
32 ammonia is retained in exhaled breath), it is likely that ammonia concentrations in breath exhaled
33 from the nose would be lower if the inspired air were free of ammonia Therefore, levels of
34 ammonia in exhaled breath reported in the literature would need to be adjusted if they were to be
35 used as a measure of systemic ammonia.
36 Ammonia concentrations measured in breath exhaled from the nose and trachea,
37 considered to be more representative of systemic ammonia levels than breath exhaled from the
38 mouth, are lower than the ammonia RfC of 0.3 mg/m3 by a factor of threefold or more. Although
39 the RfC falls within the range of concentrations measured in the mouth or oral cavity, ammonia
40 exhaled by an individual is rapidly diluted in the larger volume of ambient air and would not
This document is a draft for review purposes only and does not constitute Agency policy.
xxxiv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 contribute significantly to ammonia exposure. Further, such endogenous exposures existed in the
2 occupational epidemiology studies that served as the basis for the ammonia RfC.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxv DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
2 LITERATURE SEARCH STRATEGY | STUDY
3 SELECTION AND EVALUATION
4
5
6 The primary, peer-reviewed literature pertaining to ammonia was identified through a
7 keyword search of the databases listed in Table LS-1. The detailed search string used for searching
8 these databases is provided in Appendix D, Table D-l. The original literature search was conducted
9 through March 2012; an updated literature search was conducted using the same strategy from
10 March 2012 through March 2013. References from health assessments developed by other national
11 and international health agencies were also examined. References were also identified by
12 reviewing the list of references cited in key health effects studies of ammonia ("backwards
13 searching"), and a "forward search" of studies citing the development of an asthma-specific job
14 exposure matrix [Kennedy etal., 2000]: see Appendix D for additional search strategy details.
15 Other peer-reviewed information, including review articles and literature necessary for the
16 interpretation of ammonia-induced health effects, were retrieved and included in the assessment
17 where appropriate. EPA requested the public submit additional data on December 21, 2007 and
18 November 2, 2009 fU.S. EPA. 2009b. 20071: no submissions were received.
19 Figure LS-1 depicts the literature search and study selection strategy and the number of
20 references obtained at each stage of literature screening. Approximately 23,000 references were
21 identified with the initial keyword search. Based on a secondary keyword search followed by a
22 preliminary manual screen of titles or abstracts by a toxicologist, approximately 1,032 references
23 were identified that provided information potentially relevant to characterizing the health effects
24 or physical and chemical properties of ammonia. A more detailed review of titles, abstracts, and/or
25 papers, and a review of references within identified papers, pared this to 40 epidemiological
26 studies (i.e., studies of workers exposed to ammonia in industrial settings or through the use of
27 ammonia in cleaning products, livestock farmers, or short-term exposure in volunteers as well
28 background epidemiology method papers), 44 case reports, 61 unique oral or inhalation animal
29 studies and 105 other studies (e.g., studies that provided supporting information on physical and
30 chemical properties, mode of action, and toxicokinetics). The majority of the toxicokinetics studies
31 came from the ATSDR (2004) Toxicological Profile of Ammonia2 or were identified based on a
32 focused keyword search (e.g., for studies on ammonia in exhaled breath or ammonia in fetal
33 circulation).
34
2Portions of this Toxicological Review were developed under a Memorandum of Understanding with the
Agency for Toxic Substances and Disease Registry (ATSDR) and were adapted from the Toxicological Profile
for Ammonia (ATSDR. 2004) and the references cited in that document as part of a collaborative effort in the
development of human health lexicological assessments for the purposes of making more efficient use of
available resources and to share scientific information.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxvi DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table LS-1. Details of the literature search strategy employed for ammonia
Database
Keywords3
Pubmed
Toxcenter
Toxline
Current Contents
(2008 and 2010 only)
Chemical names (CASRN): ammonia (7664-41-7); ammonium hydroxide (1336-21-6)
Synonyms: spirit of hartshorn; aquammonia
Initial keyword search
Standard toxicology search (see Appendix D for specific keywords used)
toxicity (including duration, effects to children and occupational exposure); development;
reproduction; teratogenicity; exposure routes; pharmacokinetics; toxicokinetics; metabolism;
body fluids; endocrinology; carcinogenicity; genotoxicity; antagonists; inhibitors
Chemical-specific keywords
respiration; metabolism; breath tests; inhalation; air; breath; exhalation; biological markers;
analysis
Secondary keyword search0
reproductive; developmental; teratogen; gastrointestinal; stomach; gastric AND mucosa,
cancer OR tumor; genotoxicity; kidney OR spleen AND toxicity; exhaled breath; respiratory
irritation, symptom OR disease, including dyspnea, bronchitis, pneumonitis, asthma; lung;
pulmonary function; chest tightness; inflammation; congestion; edema; hemorrhage;
discharge; epithelium; immune; immunosuppression; hypersensitivity; skin lesion; erythema;
host resistance; bacterial colonization; T-cell; liver function OR toxicity; fatty liver; clinical
chemistry; adrenal; heart AND toxicity; myocardium; lacrimation; ocular symptoms; blood
pH; brain AND amino acid; neurotransmitter
The following terms were used to filter out reference not relevant to the evaluation of the
health effects of ammonia: hyperammonemia; ammonemia; hepatic coma; liver failure; Reye
syndrome; hepatic encephalopathy; cirrhosis; fish; daphnia; crustaceans; amphibians
TSCATS
Searched by chemical names (including synonyms) and CASRNs
ChemID
Chemfinder
CCRIS
HSDB
GENETOX
RTECS
aThe use of certain keywords in a given database was contingent on number and type of results. The large number
of search results required restriction of search terms to filter out references not relevant to evaluation of ammonia
health effects and limiting metabolism results to studies in animals and humans.
bAs discussed in the Preface, literature on ammonium salts was not included in this review because of the
uncertainty as to whether the anion of the salt can influence the toxicity of the ammonium compound (see also
Appendix C, Table C-l).
Secondary keywords were selected from an understanding of the targets of ammonia toxicity gained from review
of papers identified in literature searches conducted at the start of document development and relevant review
documents.
CASRN = Chemical Abstracts Service Registry Number; CCRIS = Chemical Carcinogenesis Research Information
System; HSDB = Hazardous Substances Data Bank; RTECS = Registry of Toxic Effects of Chemical Substances;
TSCATS = Toxic Substance Control Act Test Submission Database
This document is a draft for review purposes only and does not constitute Agency policy.
xxxvii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1
2
3
4
Referencesidentified based on initial keyword search (see Table LS-1): ~23,000
Referencesexcluded based on secondary keyword search (see Table LS-1):
~13,270
Referencesidentified based on secondary keyword search (see Table LS-1): ~9,130
Reference excluded based on preliminary manual screen of
titles/abstracts: ~8,700
Reasons/or excluding references included the following:
• Topics not relevant to ammonia toxicity
• Co-exposure with other chemicals
Referencesconsidered for inclusion in the Toxicological Review: 1,032
Note: References maybe cited in more than one subsection; thus the sum of the
subsections maybe greater than the number of unique references
Human studies: 227
Animal studies (oral & inhalation): 206
Othersupportingstudies: 612
Including:
• Reviews
• Background and physical/chemical properties
• Animal studies by routes other than oral & inhalation
• Studies of H. pylori and ammonia
• Studies related to mode of action
Other search strategies
• Backward searching
• Referencesidentified
to support
interpretation of
ammonia health
effects literature
Referencesexcluded based on manual review of papers/abstracts: 737
Types of papers evaluated and not considered further.
• Concernsabout ethical conduct (Kalandarov et al., 1984)
• Not relevant to ammonia toxicity
• Inadequate information to characterize exposure
• Exposure route not relevant
• Co-exposure with other chemicals
• Nonstandard animal model (e.g., nonmammalian species, cattle, etc.)
• Pathogenic effects of H. pylori infection
• Reviewpaper
• Abstract
• Not available in English and, based on abstract, judged not to be
informative
• Duplicate
Referencescited in the Toxicological Review: 295
Note: References maybe cited in more than one subsection; thus the sum of the subsections may be greater than the
number of unique references
Human studies/reports: 84
• Epidemiologic studies: 40
• Occupational studies (6)
• Studiesin volunteers (12)
• Studiesin livestock workers
(10)
• Cleaningstudies (7)
• Background (methods) (5)
Note: Epidemiology methods
papers are not ammonia-
specific
• Case reports: 44
Animal studies: 61
• Oral: 13
• Acute (3)
• Subchronic(7)
• Chronic (3)
• Inhalation: 51
• Acute/short-term (30)
• Subchronic(9)
• Reproductive/
developmental (1)
• Immunotoxicity (11)
Othersupporting studies: 115
• Background and physical & chemical
properties: 16
• Studies related to mode of action,
includinggenotoxicity: 18
• Toxicokinetic studies: 80
• Miscellaneous: 3
Assessment by others: 7
Guidance: 27
Note: Guidances are not ammonia-
specific
Figure LS-1. Study selection strategy.
This document is a draft for review purposes only and does not constitute Agency policy.
xxxviii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Selection of studies for inclusion in the Toxicological Review was based on consideration of
2 the extent to which the study was informative and relevant to the assessment and general study
3 quality considerations. In general, the relevance and scientific quality of the available studies was
4 evaluated as outlined in the Preamble and in EPA guidance (i.e., A Review of the Reference Dose and
5 Reference Concentration Processes [U.S. EPA. 2002] and Methods for Derivation of Inhalation
6 Reference Concentrations and Application of Inhaled Dosimetry [U.S. EPA. 1994]].
7
8 Considerations for evaluation of epidemiology studies
9 Case reports are often anecdotal and describe unusual or extreme exposure situations,
10 providing little information that would be useful for characterizing chronic health hazards.
11 Ammonia case studies were only briefly reviewed; representative citations from the collection of
12 case reports are provided as supplemental information in Appendix E, Section E.2.
13 Epidemiology studies of chronic exposure to ammonia have primarily focused on industrial
14 worker populations, workers exposed to ammonia as a cleaning or disinfectant product, and
15 livestock farmers. The observational epidemiology studies identified in Figure LS-1 (i.e., the studies
16 considered most informative for evaluating ammonia toxicity from chronic exposure] are
17 summarized in evidence tables (i.e., Tables 1-1,1-2, and 1-7). Evaluation of the studies summarized
18 in the evidence tables is provided in Appendix D (Tables D-2, D-3, and D-4 corresponding to Tables
19 1-1,1-2, and 1-7, respectively). This evaluation process addressed aspects relating to the selection
20 of study participants, exposure parameters, outcome measurement, confounding, and statistical
21 analysis, as discussed below for each set of studies.
22
23 Studies of Industrial Settings
24 Selection of study participants
25 All of the studies were cross-sectional analyses in occupational settings. The workers were
26 healthy enough to remain in the work area for a considerable time; with one exception, mean
27 duration ranged from 52 months to 18 years. One study (Bhat and Ramaswamy, 1993] grouped
28 workers into those exposed for up to 10 years and those with more than 10 years of exposure; a
29 minimum exposure duration was not provided. In general, these designs may result in a "healthy
30 worker" bias. In addition, the workers in these studies are not representative of the general
31 population, as they do not include children or women. These aspects of the study design may result
32 in an underestimate of the risk of health effects of ammonia exposure, as the worker population
33 may not exhibit health effects (such as decreased lung function or increased prevalence of
34 respiratory symptoms) to the same degree that would be seen in the general population under the
35 same conditions.
36
37 Exposure parameters
38 Exposure methods differ across these occupational studies, which makes comparison of
39 ammonia measurements among the studies difficult. Spectrophotometric absorption measures of
40 areas samples (Alietal.. 2001: Ballal etal.. 1998] are not directly comparable to direct-reading
41 diffusion methods used to analysis personal samples (Rahman etal., 2007] or to the NIOSH-
This document is a draft for review purposes only and does not constitute Agency policy.
xxxix DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 recommended protocol for personal sampling and analysis of airborne contaminants [Holness etal.,
2 1989]. In the study by Rahman etal. [2007], exposure concentrations were determined by both the
3 Drager tube and Drager PAC III methods. The Drager tube method yielded concentrations of
4 ammonia in the two plants studied that were approximately fourfold higher than the
5 concentrations obtained by the Drager PAC III method; a strong correlation between measurements
6 by the two methods was reported. Rahman et al. [2007] stated that their measurements indicated
7 only relative differences in exposures between workers and production areas, and did not identify
8 one analytical measure as the more valid of the two. Based on communication with technical
9 support at Drager Safety Inc. [Bacom and Yanosky, 2010], EPA considered the PAC III instrument to
10 be a more sensitive monitoring technology than the Drager tubes. Ammonia concentrations based
11 on the PAC III method were also in line with concentrations reported in other studies. Therefore,
12 exposure levels based on PAC III air measurements of ammonia were used in the current health
13 assessment to characterize the exposure-response relationship in the Rahman etal. [2007] study.
14 In the Hamid and El-Gazzar [1996] study, no direct measurement of ammonia exposure was
15 made; blood urea was used as a surrogate measure of ammonia exposure. The correlation of blood
16 urea with ammonia is not reported by the authors. EPA considered this a major limitation of this
17 study, based on other data indicating no correlation between ammonia levels in air and serum urea
18 levels in a study of six groups of workers with varying types of exposure [Giroux and Ferrieres,
19 1998]. No exposure measurements of ammonia were used in the study by Bhat and Ramaswamy
20 [1993]: EPA considers the lack of exposure measure in this study to be a major limitation.
21
22 Outcome measurement
23 Assessment of respiratory symptoms in these studies [Rahman etal., 2007: Ballal etal.,
24 1998: Holness etal.. 1989] was based on three different questionnaires; each of these, however, is a
25 standardized, validated questionnaire. Self-reporting of types and severity of respiratory
26 symptoms could be biased by the knowledge of exposure, for example, in studies comparing factory
27 workers to office workers. EPA evaluated this non-blinded outcome assessment as a potential bias.
28 In each of these studies, comparisons were made across exposure categories among the exposed;
29 EPA concluded that the non-blinded outcome assessment as a potential bias is unlikely in these
30 types of comparisons. One study also compared exposed to nonexposed, and observed little
31 differences in symptom prevalence between these groups [Holness etal.. 1989]. Thus, EPA
32 concluded that the non-blinded outcome assessment was not a major bias in this analysis either.
33 Assessment of lung function was performed by standard spirometry protocols in four studies
34 [Rahman etal.. 2007: Ali etal.. 2001: Bhat and Ramaswamy. 1993: Holness etal.. 1989]. EPA did
35 not consider any of these procedures to be a source of bias or limitation.
36
This document is a draft for review purposes only and does not constitute Agency policy.
xl DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Confounding
1 Co-exposures to other ambient chemicals in urea fertilizer factories included inorganic
3 gases (nitrogen dioxide and sulfur dioxide) and dust In one of these studies [Rahman etal.. 2007).
4 nitrogen dioxide was measured concurrently with ammonia and found to be below detection limits
5 for all areas (urea plant, ammonia plant, and administration area). The other urea fertilizer studies
6 (Mi etal.. 2001: Ballal etal.. 1998: Hamid and El-Gazzar. 1996) did not describe potential co-
7 exposures. [It appears from the exposure measurements that the plant in Alietal. (2001) is
8 "Factory A" in Ballal etal. (1998)]. In the fertilizer plant in Bhatand Ramaswamy (1993). co-
9 exposures are not discussed, but the workers are grouped based on different parts of the plant
10 (ammonia, urea, and diammonium phosphate); effects observed with respect to lung function tests
11 were similar in magnitude, albeit slightly stronger, in the ammonia plant workers compared with
12 the urea plant workers. One study was conducted in a soda ash production plant (Holness etal..
13 1989). No measurements of co-exposures were described in this study, but the authors note the
14 high level of control of exposures (resulting in low ammonia levels) in this facility. Because of the
15 lack of demonstration of co-exposures correlated with ammonia levels in these studies, and lack of
16 demonstration of stronger associations between potential co-exposures and respiratory outcomes,
17 EPA concluded that confounding by other workplace exposures, although a potential concern, was
18 unlikely to be a major limitation.
19 The analyses of respiratory symptoms and lung function may also be confounded by
20 smoking. In these five studies, analyses accounted for smoking as follows: the analysis included
21 either an adjustment for smoking (Rahman et al., 2007: Holness etal., 1989), the exclusion of
22 smokers (Bhatand Ramaswamy, 1993), or stratification of the results by smoking status (Alietal.,
23 2001: Ballal etal., 1998). EPA did not consider potential confounding by smoking to be a major
24 limitation of these studies. In reviewing the study of liver function by Hamid and El-Gazzar (1996).
25 however, EPA noted the lack of information on smoking habits or use of alcohol (another exposure
26 potentially affecting liver function tests) to be a major limitation.
27
28 Statistical analysis
29 EPA considered the statistical analysis in the epidemiological studies (Rahman et al., 2 007:
30 Alietal.. 2001: Ballal etal.. 1998: Hamid and El-Gazzar. 1996: Bhatand Ramaswamy. 1993: Holness
31 etal.. 1989) to be adequate and appropriate. Although the type of statistical testing was not
32 specified in Hamid and El-Gazzar (1996), the results were presented in sufficient detail to allow
33 interpretation of the data and analysis. Sample size, an important consideration with respect to
34 statistical power, was also considered. EPA noted the small number of exposed workers and low
35 levels of exposure in the study by Holness etal. (1989) as limitations that could result in "false
36 negative" results (i.e., a test result indicating a lack of association, whereas, in fact, a positive
37 association between exposure and a health effect exists).
38
This document is a draft for review purposes only and does not constitute Agency policy.
xli DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Studies of Health Care and Cleaning Settings
2 EPA also evaluated the studies that examined exposure to ammonia when used as a
3 cleaning or disinfectant product EPA noted the potential for the "healthy worker" bias arising from
4 movement out of jobs by affected individuals in most of these studies [Le Moual et al., 2008]. This
5 issue was less of a concern in the study by Zocketal. [2007]. which was conducted in a general
6 (non-occupational] population sample, focusing on cleaning activities in the home.
7 None of these studies used a direct measure of ammonia exposure in the analysis,
8 precluding interpretation of the results in relation to an absolute level of exposure. The limited
9 data available concerning exposure levels in cleaning scenarios found median exposures of 0.6 to
10 5.4 ppm (0.4 to 3.8 mg/m3], with peaks exceeding 50 ppm (35 mg/m3], in a small study (n = 9]
11 using personal samples during a domestic cleaning session (Medina-Ramon etal., 2005]. Although
12 an absolute level of exposure is not available, the relative ranking of exposure used in these studies
13 does allow examination of relative risk in relation to relative levels of exposure. Key considerations
14 regarding the validity of the exposure measures are the specificity of the classification and the
15 extent to which classification could be influenced by knowledge of the disease or symptoms under
16 study. Methodological research has reported underestimation of self-reported exposure to specific
17 products by health care workers, and differential reporting by disease status (i.e., asthma] for self-
18 reported use of cleaning products in patient care, but not in instrument cleaning or building
19 materials (Donnay etal.. 2011: Delclos etal.. 2009: Kennedy etal.. 2000]. Two of these studies used
20 an exposure assessment protocol that incorporated an independent, expert review, blinded to
21 disease status (Dumas etal., 2012: Lemiere etal., 2012], and one study collected exposure
22 information using a 2-week daily diary (Medina-Ramon et al., 2006]. EPA considered these to be
23 the strongest of the exposure protocols used within this set of studies.
24 Five of the studies in this set of studies used standard protocols for the assessment of
25 asthma symptoms in epidemiological studies (Arif and Delclos, 2012: Dumas etal., 2012: Vizcaya et
26 al.. 2011: Zocketal.. 2007: Medina-Ramon et al.. 2005], and one study included a clinical
27 assessment protocol designed specifically for the assessment of occupational asthma (Lemiere et
28 al., 2012]. Details of the specific questions were provided, and EPA did not consider any of these
29 methods to be a limitation in terms of specificity of the outcome. The study by Medina-Ramon et al.
30 (2006] collected information on daily respiratory symptoms in a two-week diary, and also trained
31 the participants to measure peak expiratory flow three times daily. EPA considered the potential
32 for knowledge of use of cleaning products to influence perception of symptoms to be a possible
33 limitation of this study, and also noted a lack of information about the reliability of the pulmonary
34 function measures.
35 All of these studies addressed the potential for smoking to act as a confounder in the
36 analysis. Two of the studies reported relatively weak correlations between ammonia and other
37 products assessed (Zocketal.. 2007: Medina-Ramon etal.. 2005] and one study reported stronger
38 associations with ammonia than with bleach (Dumas etal.. 2012]. Based on this information, EPA
39 did not consider potential confounding to be a major limitation of this set of studies.
This document is a draft for review purposes only and does not constitute Agency policy.
xlii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 EPA considered the statistical analysis in this set of studies to be appropriate. One study,
2 however, was limited in terms of the level of detail provided pertaining to the results for ammonia
3 from multivariate models [Medina-Ramon etal.. 2005).
4
5 Studies of Livestock Farmers
6 EPA also evaluated a set of studies conducted among livestock farmers. As with the other
7 occupational studies discussed above, the selection of sensitive individuals out of the workforce
8 would be a potential bias in cross-sectional studies in this type of population.
9 Among the studies examining pulmonary function, two studies used area-based exposure
10 sampling in animal confinement buildings [Monso etal.. 2004: Zejdaetal.. 1994). one study used
11 area samples taken in conjunction with specific tasks and calculated a personal exposure measure
12 taking into account duration spent in specific locations and tasks [Heederiketal., 1990], and four
13 studies collected personal samples over a workshift [Donham etal., 2000: Reynolds etal., 1996:
14 Preller etal.. 1995). or an unspecified time period [Donham etal.. 1995). EPA considered the use of
15 the area-based samples without consideration of duration to be limitations of the studies by Zejda
16 etal. [1994] and Monso etal. [2004].
17 All of the studies reported using a standard spirometric technique; five studies compared
18 two measures per individual (i.e., pre- and post-shift] [Monso etal., 2004: Donham etal., 2000:
19 Reynolds etal., 1996: Heederik et al., 1990] and two studies used a single pulmonary function
20 measure, adjusted for height, age, and smoking variables [Preller etal., 1995: Zejdaetal., 1994].
21 EPA did not consider any of these outcome measures to be limitations in these studies.
22 Five of these studies controlled for co-exposures (e.g., endotoxin, dust, disinfectants]
23 [Reynolds etal.. 1996: Donham etal.. 1995: Preller etal.. 1995]. noted only weak correlations (i.e.,
24 Spearman r < 0.20] between ammonia and dust or endotoxin [Donham et al.. 2000]. or observed
25 associations with ammonia but not with endotoxin or dust measures (Heederiketal., 1990]. The
26 two studies that did not address confounding were those that also used the more limited exposure
27 measure [Monso etal.. 2004: Zejdaetal.. 1994].
28 Based on these considerations, EPA considered the studies by Reynolds etal. [1996]. Preller
29 etal. [1995], Donham etal. [2000], Donham etal. [1995], and Heederiketal. [1990] to be the
30 methodologically strongest studies of this set Because of the variety of exposures in the type of
31 environment examined in these studies (including dust, endotoxin, mold, and disinfectant
32 products] and the availability of sets of studies in settings with a lesser degree of co-exposures, this
33 set of studies is considered to be supporting material.
34
35 Considerations for evaluation of animal studies
36 Relatively few repeat-dose toxicity studies of ammonia in experimental animals are
37 available. Many of the available animal studies come from the older toxicological literature and are
38 limited in terms of study design (e.g., small group sizes] and reporting of results. These studies
39 were evaluated consistent with EPA principles and practices for evaluating study quality [U.S. EPA.
40 2005a, 1998b, 1996,1994,1991]: however, detailed documentation of the methodological features
41 of the available animal studies was not necessary to convey the limitations of this body of ammonia
This document is a draft for review purposes only and does not constitute Agency policy.
xliii DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 literature. The animal studies are summarized in detail in Appendix E, Section E.3. Essentially all
2 the animal toxicology studies were included in this assessment. Any studies excluded from the
3 hazard identification as uninformative are identified in Section 1.1, along with the basis for
4 exclusion.
5 The references considered and cited in this document, including bibliographic information
6 and abstracts, can be found on the Health and Environmental Research On-line (HERO) website3
7 [http://hero.epa.gov/ammonia].
3HERO (Health and Environmental Research On-line) is a database of scientific studies and other references
used to develop EPA's risk assessments aimed at understanding the health and environmental effects of
pollutants and chemicals. It is developed and managed in EPA's Office of Research and Development (ORD)
by the National Center for Environmental Assessment (NCEA). The database includes more than 300,000
scientific articles from the peer-reviewed literature. New studies are added continuously to HERO.
This document is a draft for review purposes only and does not constitute Agency policy.
xliv DRAFT—DO NOT CITE OR QUOTE
-------
1. HAZARD IDENTIFICATION
5 1.1. SYNTHESIS OF EVIDENCE
6 1.1.1. Respiratory Effects
7 The respiratory system is the primary target of toxicity of inhaled ammonia in humans and
8 experimental animals. Five cross-sectional occupational epidemiology studies in industrial settings
9 [Rahman etal.. 2007: Alietal.. 2001: Ballaletal.. 1998: Bhatand Ramaswamy. 1993: Holnessetal..
10 1989] examined the association between inhaled ammonia and prevalence of respiratory
11 symptoms or changes in lung function (Table 1-1). Another set of studies examined pulmonary
12 function or asthma symptoms in relation to ammonia exposure in health care workers and
13 domestic cleaners [ArifandDelclos, 2012: Dumas etal., 2012: Lemiere etal., 2012: Vizcaya etal.,
14 2011: Zock etal.. 2007: Medina-Ramon etal.. 2006: Medina-Ramon et al.. 2005] (Table 1-2]. The
15 association between ammonia exposure and respiratory effects indicated by these studies is also
16 informed by studies of pulmonary function in livestock farmers, volunteer studies involving acute
17 exposures to inhaled ammonia, and human case reports (see Supplemental Material, Appendix E,
18 Section E.2], and in subchronic inhalation toxicity studies in various experimental animal species
19 (Table 1-3]. The evidence of respiratory effects in humans and experimental animals exposed to
20 ammonia is summarized in an exposure-response array in Figure 1-1 at the end of this section.
21
22 Respiratory Symptoms
23 Respiratory symptoms (including cough, wheezing, and other asthma-related symptoms]
24 were reported in two cross-sectional studies of industrial worker populations exposed to ammonia
25 at levels greater than or equal to approximately 18 mg/m3 (Rahman etal., 2007: Ballal etal., 1998]
26 (Table 1-1]. One of these studies also examined frequency of respiratory symptoms by cumulative
27 ammonia concentration (CAC, mg/m3-years] and observed significantly higher relative risks (2.5-
28 5.3] with higher CAC (>50 mg/m3-years] compared to those with a lower CAC (<50 mg/m3-years]
29 (Ballal etal.. 1998]. In three studies examining lower exposures settings (Rahman etal.. 2007:
30 Ballal etal., 1998: Holness et al., 1989] (Table 1-1], no differences were observed in the prevalence
31 of respiratory symptoms between ammonia-exposed workers and controls. Ammonia
32 concentrations reported in these lower exposure settings included a mean ammonia concentration
33 of 6.5 mg/m3 and a high-exposure group defined as >8.8 mg/m3 in Holness etal. (1989], an
34 exposure range of 0.2—7 mg/m3 in "Factory B" of Ballaletal. (1998], and a mean concentration of
35 4.9 mg/m3 in Rahman et al. (2007]. The primary limitation noted in all of these studies was the
36 potential under-ascertainment of effects inherent in the study of a long-term worker population
37 (i.e., "healthy worker" effect] (see Literature Search Strategy | Study Selection and Evaluation
38 section and Table D-2 in the Supplemental Information]. Confounding by other workplace
This document is a draft for review purposes only and does not constitute Agency policy.
1-1 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 exposures, although a potential concern, was unlikely to be a major limitation affecting the
2 interpretation of the pattern of results seen in these studies, given the lack of nitrogen dioxide
3 measurements above the detection limit in one study [Rahman etal.. 2007] and the high level of
4 control of exposures in another study [Holness etal., 1989].
5 Studies of health care workers or hospital workers [Arif andDelclos. 2012: Dumas etal..
6 2012] (Table 1-2] provide evidence that exposure to ammonia as a cleaning or disinfectant product
7 is associated with increased risk of asthma or asthma symptoms. Use of ammonia as a cleaning
8 product in other settings has also been associated with asthma and respiratory symptoms [Vizcaya
9 etal.. 2011: Zock etal.. 2007: Medina-Ramon etal.. 2005] (Table 1-2]. Occupational exposure to
10 ammonia was associated with work-exacerbated asthma (compared to non-work related asthma]
11 in a study at two occupational asthma specialty clinics by Lemiere et al. (2012] (Table 1-2]. Each of
12 six studies, from Europe, Canada, and the United States, observed elevated odds ratios, generally
13 between 1.5 and 2.0, with varying degrees of precision. These studies were conducted using a
14 variety of designs, including a prospective study (Zock etal., 2007] and a nested case-control study
15 (Medina-Ramon et al., 2005]. Criteria used to define current asthma or asthma symptoms were
16 generally well defined and based on validated methods. A major limitation of this collection of
17 studies is the lack of direct measures of ammonia exposure. Two of the studies included expert
18 assessment of exposure (blinded to case status]; expert assessment, improves reliance on self-
19 reported exposure (Dumas etal.. 2012: Lemiere etal.. 2012]. Confounding by other cleaning
20 products is an unlikely explanation for these results, as two of the studies noted only weak
21 correlations between ammonia and other product use (Zock etal., 2007: Medina-Ramon et al.,
22 2005], and another study observed stronger associations with ammonia than with bleach (Dumas
23 etal., 2012]. All of the studies addressed smoking as a potential confounder.
24 Studies in swine and dairy farmers analyzing prevalence of respiratory symptoms
25 (including cough, phlegm, wheezing, chest tightness, and eye, nasal, and throat irritation] in relation
26 to ammonia exposure provided generally negative results (Melbostad and Eduard. 2001: Preller et
27 al., 1995: Zejdaetal., 1994] (Appendix E, Table E-7]. Two other studies that measured ammonia,
28 but did not present an analysis in relation to variability in ammonia levels, reported an increased
29 prevalence of respiratory symptoms in pig farmers exposed to ammonia from animal waste
30 (Choudatetal., 1994: Crook etal., 1991] (Appendix E, Table E-8). In addition to ammonia, these
31 studies also documented exposures to other compounds, such as airborne dust, endotoxin, mold,
32 and disinfectants.
33 Reports of irritation and hyperventilation in volunteers acutely exposed to ammonia at
34 concentrations ranging from 11 to 354 mg/m3 ammonia for durations up to 4 hours under
35 controlled exposure conditions (Petrovaetal., 2008: Smeets etal., 2007: Ihrigetal., 2006: Verberk,
36 1977: Silvermanetal., 1949] provide support for ammonia as a respiratory irritant (Appendix E,
37 Section E.2 and Table E-9]. Two controlled-exposure studies report habituation to eye, nose, and
38 throat irritation in volunteers after several weeks of ammonia exposure (Ihrigetal.. 2006:
39 Ferguson et al., 1977]. Numerous case reports document the acute respiratory effects of inhaled
40 ammonia, ranging from mild symptoms (including nasal and throat irritation and perceived
This document is a draft for review purposes only and does not constitute Agency policy.
1-2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 tightness in the throat) to moderate effects (including pharyngitis, tachycardia, dyspnea, rapid and
2 shallow breathing, cyanosis, transient bronchospasm, and rhonchi in the lungs) to severe effects
3 (including burns of the nasal passages, soft palate, posterior pharyngeal wall, and larynx, upper
4 airway obstruction, bronchospasm, persistent, productive cough, bilateral diffuse rales and rhonchi,
5 mucous production, pulmonary edema, marked hypoxemia, and necrosis of the lung) (Appendix E,
6 Section E.2).
7 Experimental studies in laboratory animals also provide consistent evidence that repeated
8 exposure to ammonia can affect the respiratory system (Table 1-3 and Appendix E, Section E.3).
9 The majority of available animal studies did not look at measures of respiratory irritation, in
10 contrast to the majority of human studies, but rather examined histopathological changes of
11 respiratory tract tissues. Histopathological changes in the nasal passages were observed in
12 Sherman rats after 75 days of exposure to 106 mg/m3 ammonia and in F344 rats after 35 days of
13 exposure to 177 mg/m3 ammonia, with respiratory and nasal epithelium thicknesses increased 3-4
14 times that of normal (Brodersonetal., 1976). Thickening of nasal and tracheal epithelium (50-
15 100%) was also observed in pigs exposed to 71 mg/m3 ammonia continuously for 1-6 weeks (Doig
16 and Willoughby, 1971). Nonspecific inflammatory changes (not further described) were reported
17 in the lungs of Sprague-Dawley and Long-Evans rats continuously exposed to 127 mg/m3 ammonia
18 for 90 days and rats and guinea pigs intermittently exposed to 770 mg/m3 ammonia for 6 weeks;
19 continuous exposure to 455 and 470 mg/m3 ammonia increased mortality in rats (Coonetal..
20 1970). Focal or diffuse interstitial pneumonitis was observed in all Princeton-derived guinea pigs,
21 New Zealand white rabbits, beagle dogs, and squirrel monkeys exposed to 470 mg/m3 ammonia
22 (Coonetal., 1970). Additionally, under these exposure conditions, dogs exhibited nasal discharge
23 and other signs of irritation (marked eye irritation, heavy lacrimation). Nasal discharge was
24 observed in 25% of rats exposed to 262 mg/m3 ammonia for 90 days (Coonetal.. 1970).
25 At lower concentrations, approximately 50 mg/m3 and below, the majority of studies of
26 inhaled ammonia did not identify respiratory effects in laboratory animals exposed to ammonia.
27 No increase in the incidence of respiratory or other diseases common to young pigs was observed
28 after continuous exposure to ammonia and inhalable dust at concentrations representative of those
29 found in commercial pig farms (<26 mg/m3 ammonia) for 5 weeks (Done etal., 2005). No gross or
30 histopathological changes in the turbinates, trachea, and lungs of pigs were observed after
31 continuous exposure to 35 or 53 mg/m3 ammonia for up to 109 days (Curtis etal.. 1975). No signs
32 of toxicity in rats or dogs were observed after continuous exposure to 40 mg/m3 ammonia for 114
33 days or after intermittent exposure (8 hours/day) to 155 mg/m3 ammonia for 6 weeks (Coonetal..
34 1970). Only one study reported respiratory effects at concentrations <50 mg/m3 (i.e., lung
35 congestion, edema, and hemorrhage in guinea pigs and mice exposed to 14 mg/m3 ammonia for up
36 to 42 days; Anderson et al. (1964)), but confidence in the findings from this study is limited by
37 inadequate reporting and small numbers of animals tested.
38
This document is a draft for review purposes only and does not constitute Agency policy.
1-3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Lung Function
1 Decreased lung function in ammonia-exposed workers has been reported in three of the
3 four studies examining this outcome measure [Rahman et al.. 2007: Ali etal.. 2001: Holness etal..
4 1989]: the exception is the study by Holness etal. [1989] (Table 1-1] in which no significant
5 changes in lung function were observed in workers exposed to ammonia in an industrial setting
6 with relatively low ammonia exposure levels (Table 1-1]. These effects were observed in short-
7 term scenarios (i.e., cross-work shift changes in lung function] in fertilizer factor workers (mean
8 ammonia concentration of 18.5 mg/m3] compared with administrative staff controls (Rahman etal.,
9 2007], and in longer-term scenarios, in workers with a cumulative exposure of >50 mg/m3-years
10 when compared with workers with a lower cumulative exposure of <50 mg/m3-years (Ali etal..
11 2001]. There were no decrements in the percent of predicted lung function values when comparing
12 the total exposed group to a control group of office workers in this study (Ali etal.. 2001]. in the
13 relatively low exposure scenario examined in Holness etal. (1989] (mean ammonia concentration
14 of 6.5 mg/m3 and high-exposure group defined as >8.8 mg/m3], or in the low-exposure group
15 (mean ammonia concentration of 4.9 mg/m3] in Rahman etal. (2007]. Another study of ammonia
16 plant fertilizer workers reported statistically significant decreases in forced expiratory volume
17 (FEVi] and peak expiratory flow rate (PEFR/minute] in workers compared to controls (Bhat and
18 Ramaswamy, 1993]: however, measurements of ammonia levels were not included in this study.
19 As discussed previously in the summary of respiratory symptoms studies, the primary limitation
20 within this set of studies is the potential under-ascertainment of effects in these studies of long-
21 term worker populations.
22 One of the studies of domestic cleaning workers described in Table 1-2 included a measure
23 of pulmonary function (Medina-Ramon etal., 2006]. Ammonia use was associated with a decrease
24 in peak expiratory flow (PEF) (-9.4 [95% CI,-17,-2.3]]. A limitation of this study was the use of
25 lung function measurements conducted by the participant; the reliability of this procedure has not
26 been established.
27 Impaired respiratory function (e.g., decreased FEVi and forced vital capacity [FVC]] in
28 livestock farmers was associated with ammonia exposure in five of the seven studies that included
29 pulmonary function measures (Monso etal., 2004: Donham etal., 2000: Reynolds etal., 1996:
30 Donham etal.. 1995: Preller etal.. 1995: Zejda etal.. 1994: Heederiketal.. 1990] (Appendix E, Table
31 E-7]. EPA considered these studies to be the strongest with respect to methodology, based on
32 considerations of exposure assessment and assessment of potential confounding (see Literature
33 Search Strategy | Study Selection and Evaluation section].
34 Changes in lung function following acute exposure to ammonia have been observed in some,
35 but not all, controlled exposure studies conducted in volunteers (Appendix E, Section E.2 and Table
36 E-9]. Cole etal. (1977] reported reduced lung function as measured by reduced expiratory minute
37 volume and changes in exercise tidal volume in volunteers exposed for a half-day in a chamber at
38 ammonia concentrations >106 mg/m3, but not at 71 mg/m3. Bronchoconstriction was reported in
39 volunteers exposed to ammonia through a mouthpiece for 10 inhaled breaths of ammonia gas at a
40 concentration of 60 mg/m3 (Douglas and Coe. 1987]: however, there were no bronchial symptoms
This document is a draft for review purposes only and does not constitute Agency policy.
1-4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 reported in volunteers exposed to ammonia in an exposure chamber at concentrations of up to 35
2 mg/m3 for 10 minutes [MacEwen et al., 1970]. Similarly, no changes in bronchial responsiveness or
3 lung function (as measured by FVC and FEVi) were reported in healthy volunteers exposed to
4 ammonia at concentrations up to 18 mg/m3 for 1.5 hours during exercise [Sundblad et al., 2004].
5 There were no changes in lung function as measured by FEVi in 25 healthy volunteers and 15
6 mild/moderate persistent asthmatic volunteers exposed to ammonia concentrations up to 354
7 mg/m3 ammonia for up to 2.5 hours [Petrovaetal., 2008], or in 6 healthy volunteers and 8 mildly
8 asthmatic volunteers exposed to 11-18 mg/m3 ammonia for 30-minute sessions [Sigurdarsonetal.,
9 2004].
10 Lung function effects following ammonia exposure were not evaluated in the available
11 animal studies.
12
13
This document is a draft for review purposes only and does not constitute Agency policy.
1-5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-1. Evidence pertaining to respiratory effects in humans following
inhalation exposure in industrial settings
Study design and reference
Results
Respiratory symptoms
Rahman et al. (2007) (Bangladesh)
Urea fertilizer factory worker (all men); 24 ammonia
plant workers, 64 urea plant workers, and 25
controls (staff from administration building). Mean
employment duration: 16 years
Exposure: Personal samples (2 methods3;
correlation = 0.80)
Low-exposure group (ammonia plant), mean: 6.9
ppm (4.9 mg/m3); range: 2.8-11.1 ppm (2-8
mg/m3)
High-exposure group (urea plant), mean: 26.1 ppm
(18.5 mg/m3); range: 13.4-43.5 ppm (9-31 mg/m3)
Outcome: Respiratory symptoms (5 point scale for
severity over last shift), based on Optimal Symptom
Score Questionnaire
Percentage of workers reporting symptoms (p-value):
Low exposed High exposed
Controls (n = 24) (n = 64)
(n = 25) (p-value)1 (p-value)2 (p-value)3
Cough
Chest tightness
Stuffy nose
Runny nose
Sneeze
8
8
4
4
8
17 (0.42)
17 (0.42)
12 (0.35)
4 (1.0)
0 (0.49)
28 (0.05)
33 (0.02)
16(0.17)
16 (0.17)
22 (0.22)
(0.41)
(0.19)
(1.0)
(0.28)
(Q.oi)
Vvalue for ammonia plant compared to control
2p-value for urea plant compared to control
3p-value for urea plant compared to ammonia plant
Ballal et al. (1998) (Saudi Arabia)
Urea fertilizer factory workers (two factories) (all
men); 161 exposed workers and 355 unexposed
controls'5. Mean employment duration: 51.8 months
(exposed workers) and 73.1 months (controls)
Exposure: Area monitors (3 sets in each work
section taken at least 3 months apart, mean 16
measures per set).
Factory A (high-exposure factory): 2-1301 mg/m3
(mid-point = 66 mg/m3); geometric mean <18
mg/m3, except for urea packaging and store areas
(geometric means = 18.6 and 115 mg/m3,
respectively)
Factory B (low-exposure factory): 0.02-7 mg/m3;
geometric mean <18 mg/m3
Cumulative exposure calculated based on exposure
and duration; dichotomized to high and low at 50
mg/m3-years
Outcome: Respiratory symptoms based on British
Medical Research Council questionnaire
Relative risk (95% Cl), compared with controls
Factory B2 Factory A2
(0.02-7 mg/m3; n = 77) (2-27.1 mg/m3; n = 78)1
Cough
Phlegm
Wheezing
Dyspnea
No cases
No cases
0.97(0.21,4.5)
0.45(0.11, 1.9)
2.0 (0.38, 10.4)
2.0 (0.38, 10.4)
3.4(1.2,9.5)
1.8 (0.81, 4.2)
Relative risk (95% Cl), compared with lower exposure setting
(<18 mg/m3 [n = 138] or <50 mg/m3-years [n = 130])
Cumulative
mg/m3 >50 mg/m3-years
(n = 17) (n = 30)
Cough
Phlegm
Wheezing
Dyspnea
Asthma
Chronic
bronchitis
3.5(1.8,6.6)
3.8(2.0.7.1)
5.0 (2.4, 10.6)
4.6 (2.4, 8.8)
4.3(2.1,9.0)
2.3 (0.31, 17)
2.8 (1.6, 5.0)
3.0 (1.7, 5.5)
5.2 (2.9, 9.5)
2.6 (1.3, 5.4)
2.4 (1.1, 5.4)
5.3 (1.7, 16)
ammonia concentration range in Factory A is
better represented as 2-27.1 mg/m3. This range
excludes the employees in the urea store (n = 6;
range of ammonia concentrations = 90-130.4
mg/m3) who were required to wear full protective
clothing, thus minimizing potential exposure.
Number of workers in Factory A excluding urea
store workers = 78.
Factory-specific analyses stratified by smoking status;
results presented here are for non-smokers. Similar patterns
seen in other smoking categories.
Approximate 1.3-1.5 relative risk (p < 0.05) per unit increase
in ammonia concentration for cough, phlegm, wheezing, and
asthma, adjusting for duration of work, cumulative exposure,
smoking, and age.
This document is a draft for review purposes only and does not constitute Agency policy.
1-6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-1. Evidence pertaining to respiratory effects in humans following
itinn evnnsiire in industrial spttinps
ictuic; xx. uv iuc;iii,c; pc;i iciiiiiiig lu i c;3pu cii
inhalation exposure in industrial settings
Study design and reference
Holness et al. (1989) (Canada)
Soda ash plant workers (all men); 58 exposed
workers and 31 controls (from stores and office
areas of plant)0. Average exposure: 12.2 years
Exposure: Personal samples, one work-shift per
person, mean 8.4 hours
Low: <6.25 ppm (<4.4 mg/m3); n = 34
Medium: 6.25-12.5 ppm (4.4-8.8 mg/m3); n = 12
High: >12.5 ppm (>8.8 mg/m3); n = 12
All exposed workers (mean): 6.5 mg/m3
Outcome: Respiratory symptoms based on
American Thoracic Society questionnaire
Lung function
Rahman et al. (2007) (Bangladesh)
Urea fertilizer factory worker (all men); 24 ammonia
plant workers, 64 urea plant workers, and 25
controls (staff from administration building). Mean
employment duration: 16 years
Exposure: Personal samples (2 methods3;
correlation = 0.80)
Low-exposure group (ammonia plant), mean: 6.9
ppm (4.9 mg/m3); range: 2.8-11.1 ppm (2-8
mg/m3)
High-exposure group (urea plant), mean: 26.1 ppm
(18.5 mg/m3); range: 13.4-43.5 ppm (9-31 mg/m3)
Outcome: Lung function (standard spirometry)
All et al. (2001) (Saudi Arabia)
Urea fertilizer factory workers (all men)— (additional
study of "Factory A" in Ballal et al. (1998)); 73
exposed workers and 348 unexposed controls.
Mean employment duration: not reported
Exposure: 4-hour measurements. Cumulative
exposure calculated based on exposure and
duration; dichotomized to high and low at 50
mg/m3-years
Outcome: Lung function (standard spirometry;
morning measurement)
Results
Percentage of workers reporting symptoms (%):
Control Exposed
(n = 31) (n = 58) p-value
Cough 10 16 0.53
Sputum 16 22 0.98
Bronchitis 19 22 0.69
Wheeze 10 10 0.91
Chest tightness 6 3 0.62
Dyspnea 13 7 0.05
(shortness of
breath)
Chest pain 6 2 0.16
Rhinitis (nasal 19 10 0.12
complaints)
Throat irritation 3 7 0.53
No increased risk seen in analyses stratified by exposure
group.
Pre-shift Post-shift p-value
Ammonia plant (low-exposure group, 4.9 mg/m ); n = 24
ammonia plant workers
FVC 3.308 3.332 0.67
FEVi 2.627 2.705 0.24
PEFR 8.081 8.313 0.22
Urea plant (high-exposure group, 18.5 mg/m3); n = 64 urea
plant workers
FVC 3.362 3.258 0.01
FEVi 2.701 2.646 0.05
PEFR 7.805 7.810 0.97
p-value reflects the comparison of pre- and post-shift values.
Control Exposed
(n = 348) (n = 73) p-value
FEV^/o predicted 96.6 98.1 NS
FVC% predicted 101.0 105.6 0.002
FEV!/FVC°/o 83.0 84.2 NS
<50 mg/m3-y >50 mg/m3-y
(n = 45) (n = 28) p-value
FVCi% 100.7 93.4 0.006
predicted
FVC% 105.6 100.2 0.03
predicted
FEV!/FVC°/o 84.7 83.4 NS
NS = not significant (p-values not provided by study authors)
This document is a draft for review purposes only and does not constitute Agency policy.
1-7 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-1. Evidence pertaining to respiratory effects in humans following
itinn evnnsiire in industrial spttinps
ictuic; xx. uv iuc;iii,c; pc;i iciiiiiiig lu i c;3pu cii
inhalation exposure in industrial settings
Study design and reference
Bhat and Ramaswamv (1993) (India)
Fertilizer chemical plant workers; 30 diammonium
phospate (DAP) plant workers, 30 urea plant
workers, 31 ammonia plant workers, and 68
controls (people with comparable body surface area
chosen from the same socio-economic status and
sex as exposed workers)
Exposure: Measurements not reported; duration
dichotomized as <10 and >10 years
Outcome: Lung function (standard spirometry)
Holness et al. (1989) (Canada)
Soda ash plant workers (all men); 58 exposed
workers and 31 controls (from stores and office
areas of plant)0. Average exposure: 12.2 years
Exposure: Personal samples, one work-shift per
person, mean 8.4 hours
Low: <6.25 ppm (<4.4 mg/m3); n = 34
Medium: 6.25-12.5 ppm (4.4-8.8 mg/m3); n = 12
High: >12.5 ppm (>8.8 mg/m3); n = 12
All exposed workers (mean): 6.5 mg/m
Outcome: Lung function (standard spirometry;
beginning and end of shift, at least two test days per
worker)
Results
Controls
(n = 68)
FVC 3.4 ±0.21
FEVi 2.8 ±0.10
PEFR 383 ±7.6
DAP plant
(n = 30)
2.5 ±0.06*
2.1 ±0.08*
228 ± 18*
Urea plant
(n = 30)
3.3 ±0.11
2.7 ±0.10
307 ± 19*
Ammonia
plant
(n = 31)
3.2 ±0.07
2.5 ±0.1*
314 ± 20*
*p<0.05
Lung function (%
FVC
FEVi
FEVi/FVC
Control
(n = 31)
predicted values)
98.6
95.1
96.5
Change in lung function over work
FVC dayl
day 2
FEVi day 1
day 2
-0.9
+0.1
-0.2
+0.5
Exposed
(n = 58)
96.8
94.1
97.1
shift:
-0.8
-0.0
-0.2
+0.7
p- value
0.094
0.35
0.48
0.99
0.84
0.94
0.86
FEVi = forced expiratory volume in 1 second; FVC = forced vital capacity; PEFR = peak expiratory flow rate.
aExposure concentrations were determined by both the Dra'ger tube and Dra'ger PAC III methods. Using the Dra'ger
tube method, concentrations of ammonia in the ammonia and urea plants were 17.7 and 88.1 mg/m3, respectively;
using the Dra'ger PAC III method, ammonia concentrations were 4.9 and 18.5 mg/m3, respectively (Rahman et al.
(2007). The study authors observed that their measurements indicated only relative differences in exposures
between workers and production areas, and that the validity of the exposure measures could not be evaluated
based on their results. Based on communication with technical support at Dra'ger Safety Inc (telephone
conversations and e-mails dated June 22, 2010, from Michael Yanosky, Dra'ger Safety Inc., Technical Support
Detection Products to Amber Bacom, SRC, Inc., contractor to NCEA, ORD, U.S. EPA), EPA considered the PAC III
instrument to be a more sensitive monitoring technology than the Dra'ger tubes. Therefore, higher confidence is
attributed to the PAC III air measurements of ammonia for the Rahman et al. (2007) study.
bThe process of fertilizer production involved synthesis of ammonia from natural gas, followed by reaction of the
ammonia and carbon dioxide to form ammonium carbamide, which was then converted to urea.
°At this plant, ammonia, carbon dioxide, and water were the reactants used to form ammonium bicarbonate, which
in turn was reacted with salt to produce sodium bicarbonate and subsequently processed to form sodium
carbonate. Ammonia and carbon dioxide were recovered in the process and reused.
This document is a draft for review purposes only and does not constitute Agency policy.
1-8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-2. Evidence pertaining to respiratory effect in humans following
inhalation exposure in cleaning settings
Study design and reference
Results
Asthma or asthma symptoms
Dumas et al. (2012) (France)
Hybrid design, hospital workers, drawn from
population-based case-control study; 179 hospital
workers (136 women), 333 other workers (545 women).
Exposure: Asthma-specific job exposure matrix plus +
expert review (blinded), ever exposed, 18 specific
products, based on all jobs held at least 3 months;
ammonia prevalence 23% in female hospital workers
Outcome: Current asthma: Asthma attack, respiratory
symptoms or asthma treatment in the last 12 months
(based on standardized questionnaire)
Odds ratio (95% Cl), current asthma
Women: 3.05 (1.19, 7.82)
Men: no associations with any specific products
(prevalence low)
Adjusted for age and smoking, and accounting for
familial dependence (due to sampling of cases and first
degree relatives)
Arif and Delclos (2012) (United States, Texas)
Population survey of 3,650 health care workers
(physicians, nurses, respiratory therapists, occupational
therapists), (total n = 5,600, response rate 66%)
Exposure: Structured questionnaire—frequency of use
of products for longest job held; ever contact with list of
28 products; ammonia prevalence 23%
Outcome: Structured questionnaire
• Work-related asthma symptoms: wheezing/whistling
at work or shortness of breath at works that gets
better away from work or worse at work
• Work-exacerbated asthma: onset before began work
• Occupational asthma: onset after began work)
Odds ratio (95% Cl) [n cases]
Work-related asthma symptoms [n = 132]
2.45(1.28,4.69)
Work-exacerbated asthma [n = 41]
1.58 (0.56, 4.43)
Occupational asthma [n = 33]
1.86 (0.49, 7.13)
Adjusted for age, sex, race/ethnicity, body mass index,
seniority, atopy, and smoking status
Lemiere et al. (2012) (Quebec, Canada)
Case-control study, workers seen at two tertiary care
centers specializing in occupational asthma. Asthma
(defined below) based on reversible airflow limitation
or airway hyper-responsiveness tests; referent group =
non-work related asthma (NWRA) seen at same clinics
but symptoms did not worsen at work (n = 33).
Exposure: Structured interview focusing on last/current
job, combined with expert review (blinded); ammonia
prevalence 19/153 = 12%
Outcome: Diagnoses made based on reference tests
• Occupational asthma if specific inhalation challenge
test was positive
• Work-exacerbated asthma if specific inhalation test
was negative but symptoms worsened at work
Odds ratio (95% Cl) [n cases]
Work exacerbation [n = 53]
8.4(1.1,371.7)
Occupational asthma [n = 67]
3.7 (0.4,173.4)
Age, smoking, occupational exposure to heat, cold,
humidity, dryness, and physical strain assessed as
confounders.
[Wide confidence intervals reflect sparseness in
referent group, with only 1 of the 33 classified as
exposed to ammonia]
This document is a draft for review purposes only and does not constitute Agency policy.
1-9 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-2. Evidence pertaining to respiratory effect in humans following
inhalation exposure in cleaning settings
Study design and reference
Results
Vizcava et al. (2011) (Spain)
Survey of cleaning service workers (n = 917) from 37
businesses (19% response rate to questionnaire
distributed through the employers); 761 current
cleaners, 86 former cleaners, 70 never cleaners;
referent group = never cleaners and current cleaners
who have not used any of the specified cleaning
products in last year (n = 161)
Exposure: Structured questionnaire, use of cleaning
tasks and 12 products; ammonia prevalence 66%
Outcome: Structured questionnaire
• Current asthma: in past 12 months, woken by an
attack of shortness of breath, had an attack of
asthma or currently taking any asthma medications
(including inhalers, aerosols or tablets)
• Asthma score: Sum of "yes" answers to 5 symptoms
in last 12 months (wheeze with breathlessness,
woken up with chest tightness, attack of shortness
of breath at rest, attack of shortness of breath after
exercise, woken by attack of shortness of breath)
Odds ratio (95% Cl) (among current cleaners) [n]
Current asthma 1.4 (0.6, 3.2) [81]
Wheeze without having a cold 2.1 (0.9, 4.7) [83]
Chronic cough 1.6 (0.8, 3.3) [95]
Asthma score 1.6 (1.0, 2.5)
[mean 0.59, SD 1.12]
Adjusted for age, country of birth (Spanish versus non-
Spanish), sex, and smoking status
Zock et al. (2007) (Europe, 22 sites)
Longitudinal study, n = 3,503, 9-year follow-up of
European Community Respiratory Health Survey,
population-based sample, ages 20-44 years. Excluded
764 individuals with asthma at baseline; limited to
individuals reporting doing the cleaning or washing in
their home.
Exposure: Structured interview at follow-up; frequency
of use of 15 products
Outcome: Structured interview at follow-up
• New onset (since baseline survey) current asthma,
defined by asthma attack or nocturnal shortness of
breath in the past 12 months or current use of
medication for asthma
• Current wheeze defined as wheezing or whistling in
the chest in last 12 months when not having a cold
• New onset physician-diagnosed asthma, asthma
defined as above with confirmation by a physician
and information on age or date of first attack
Odds ratio (95% Cl) [n]
Current asthma
Current wheeze
Physician-diagnosed asthma
1.4 (0.87, 2.23) [199]
1.3(0.81, 2.13) [226]
0.92 (0.33, 2.59) [71]
Adjusted for sex, age, smoking, employment in a
cleaning job during follow-up, and study center;
heterogeneity by center also assessed. Correlations
among products generally weak (Spearman rho < 0.3)
This document is a draft for review purposes only and does not constitute Agency policy.
1-10 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-2. Evidence pertaining to respiratory effect in humans following
inhalation exposure in cleaning settings
Study design and reference
Results
Medina-Ramon et al. (2005) (Spain)
Nested case-control, cleaning workers; case (n = 40;
74% participation rate) based on asthma and/or
bronchitis at both assessments. Controls (n = 155, 69%
participation rate)—no history of respiratory symptoms
in preceding year and no asthma at either assessment.
Exposure: Structured interview; frequency of use of 22
products; ammonia prevalence 16% undiluted, 56%
diluted
Outcome: Asthma: asthma attack or being woken by
attack or shortness of breath in past 12 months;
Chronic bronchitis: regular cough or regular bringing up
phlegm for at least 3 months each year
Odds ratio (95% Cl) (unadjusted), >12 compared with
<12 times per year
Undiluted 3.1(1.2,8.0)
Diluted 0.8 (0.4,1.7)
Pulmonary function and respiratory symptoms
Medina-Ramon et al. (2006) (Spain)
Panel study, sample selected from participants in
nested case-control study by Medina-Ramon et al.
(2005). Current asthma symptoms or chronic bronchitis
in 2000-2001 survey; n = 51 of 80 (64%); 8 excluded for
possible recording errors, outliers, learning effects
Exposure: Daily diary of use of products
Outcome: Respiratory symptoms based on 2-week daily
diary (7 symptoms, 5 point intensity scale); summed
score for upper respiratory symptoms (blocked nose,
throat irritation, watery eyes) and lower respiratory
symptoms (chest tightness, wheezing, shortness of
breath, and cough); PEF measured with mini-Wright
peak flow meter (with training and written
instructions); measured morning, lunchtime, night (3
measurements each; highest recorded)
Diluted and Diluted
undiluted only
OR (95% Cl)
Upper
respiratory
symptoms
Lower
respiratory
symptoms
1.8 (0.7, 4.9)
1.6(0.6,4.4)
1.3 (0.3, 5.0)
3.0(1.0,9.1)
Beta (95% Cl)
PEF at night -9.4 (-17, -2.3) -10.3 (-18, -2.7)
PEF,
following -1.2 (-8.5, 6.2) -2.9 (-11, 6.2)
morning
Adjusted for respiratory infection, use of maintenance
medication, and age; daily number of
cigarettes smoked, years of employment in domestic
cleaning, and/or weekly working hours in domestic
cleaning also assessed as potential confounders
This document is a draft for review purposes only and does not constitute Agency policy.
1-11 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-3. Evidence pertaining to respiratory effects in animals
Study design and reference
Results
Effects on the lungs
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (Saimiri sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
Coon et al. (1970)
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (5. sciureus); male; 3/group
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Sprague-Dawley or Long-Evans rat; male and female; 15-51/group
0 or 40 mg/m3 for 114 d or 127, 262 or 470 mg/m3 for 90 d or 455
mg/m3 for 65 d
Anderson et al. (1964)
Swiss albino mouse; male and female; 4/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d
Anderson et al. (1964)
Guinea pig (strain not specified); male and female; 2/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d or 50 ppm (35 mg/m3) for
42 d
Done et al. (2005)
Pig (several breeds); sex not specified; 24/group
0, 0.6, 10, 18.8, or 37 ppm (0, 0.4, 7, 13.3, or 26 mg/m3) and 1.2,
2.7, 5.1, or 9.9 mg/m3 inhalable dust for 5 wks
(Exposure to ammonia and inhalable dust at concentrations
commonly found at pig farms)
Curtis et al. (1975)
Pig (crossbred); sex not specified; 4-8/group
0, 50, or 75 ppm (0, 35, or 53 mg/m3 for 109 d)
Gross necropsies were normal; focal
pneumonitis in one of three monkeys at
155 mg/m3.
Nonspecific lung inflammation observed in
guinea pigs and rats, but not in other
species, at 770 mg/m3.3
Focal or diffuse interstitial pneumonitis in all
animals. Calcification of bronchial
epithelium observed in several animals.
Hemorrhagic lung lesion in one of two dogs;
moderate lung congestion in two of three
rabbits.3
Dyspnea (mild) at 455 mg/m3. Focal or
diffuse interstitial pneumonitis in all
animals, and calcification of bronchial
epithelium observed in several animals at
470 mg/m3.3'b
Lung congestion, edema, and hemorrhage
observed at 14 mg/m3 after 42 d.3
Lung congestion, edema, and hemorrhage
observed at 14 and 35 mg/m3 after 42 d.3
No increase in the incidence of respiratory
or other diseases.
Turbinates, trachea, and lungs of all pigs
were classified as normal.
Effects on the upper respiratory tract
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
Broderson et al. (1976)°
Sherman rat; 5/sex/group
10 or 150 ppm (7 or 106 mg/m3) from bedding for 75 d
Dyspnea in rats and dogs exposed to
770 mg/m3 during wk 1 only; no indication
of irritation after wk 1; nasal tissues not
examined for gross or histopathologic
changes.
4 times) and nasal lesions at 106 mg/m3.3
This document is a draft for review purposes only and does not constitute Agency policy.
1-12 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-3. Evidence pertaining to respiratory effects in animals
Study design and reference
Broderson et al. (1976)°
F344 rat; 6/sex/group
0 or 250 ppm (0 or 177 mg/m3) in an inhalation chamber for 35 d
Coon et al. (1970)
Sprague-Dawley or Long-Evans rat; male and female; 15-51/group
0 or 40 mg/m3 for 114 d or 127, 262, or 470 mg/m3 for 90 d or
455 mg/m3 for 65 d
Gaafaretal. (1992)
White albino mouse; male; 50
Ammonia vapor of 0 or 12% ammonia solution for 15 min/d,
6 d/wk, for 8 wks
Doig and Willoughbv (1971)
Yorkshire-Landrace pig; sex not specified; 6/group
0 or 100 ppm (0 or 71 mg/m3) for 6 wks
Stombaugh et al. (1969)
Duroc pig; both sexes; 9/group
12, 61, 103, 145 ppm (8, 43, 73, or 103 mg/m3) for 5 wks
Coon et al. (1970)
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Results
t\ tk,' i f tk, i 'tk, r /o
4 times) and nasal lesions at 177 mg/m3.3
Nasal irritation in all animals at
455 mg/m3.a'b
Histological changes in the nasal mucosa.3
'f thickness of nasal and tracheal
epithelium (50-100% increase).3
Excessive nasal, lacrimal, and mouth
secretions and /T" frequency of cough at
73 and 103 mg/m3.3
Nasal discharge at 470 mg/m3.3
Incidence data not provided.
bExposure to 455 and 470 mg/m3 ammonia increased mortality in rats.
°The Broderson et al. (1976) paper includes a number of experiments in rats designed to examine whether
ammonia at concentrations commonly encountered in laboratory cage environments plays a role in the
pathogenesis of murine respiratory mycoplasmosis caused by the bacterium Mycoplasma pulmonis. The
experiments conducted without co-exposure to M. pulmonis are summarized in this table; the results of
experiments involving co-exposure to M. pulmonis are discussed in Section 1.1.4, Immune System Effects.
This document is a draft for review purposes only and does not constitute Agency policy.
1-13 DRAFT—DO NOT CITE OR QUOTE
-------
1000
100 -
t>o
o
10
c
0)
u
c
o
u
(U
I/)
o
Q.
X
1 -
0.1
•
spiratory symptoms &
ig function (occupational);
Rahmanetal. (2007)
ratory symptoms &
:tion (male occupational);
Iness etal. (1989)
& = % 1 5
QJ "^
-> <£ DO
c
_D
HUMAN STUDIES
A • A
• •
A
• A 0 •
A • • ±
. 1 t
<> L
|
L
<
<
k
1 |
^
»
1 ' I ]
|
k
• LOAEL ANOAEL •Additional concentrations
Vertical lines show range of concentrations in study.
t Highest concentration (470 mg/m3) and LOAEL(455
mg/m3) cannot be distinguished on this figure
*
.trJ5 c j§ ££; £ ~ js =r js S '5 .1 & |f
S S .2 S »a £ i _ g £ t S t 2 .a ~ ™ — - 1
nj Cl ro'~l ' — ' • ro £_ Ol o01 O ~— - CU"i E — : **" =
-^- E • .i2 c ro iti -° ^ E ~— ' E — : >•§ Ero -^ ^
,-nj P~ro '^^•'-'^'i1^- QJ — : QJ^ rM ra 1^ CC* £
1 « | « H c a ° ™ 15 i « - r i .S sig 'i
IJ Si It! Hi 1 i! H ja 1*5 !
i| is IP 115 5| 53 .1 *l gii \
OJ3 ^ra tn >" nTT(i- '^ '^ • — • • — • o °fl O ^
Isf a g s£^S^f 3 ~ 3 f f i „- 5 |
— ^ w •— -ti •" c 5_ -H E MM cu3 • — • ^a S
Oc g ^ .= *^ O E7 -^ _D on nS "° u
i- "S f a g 1- 3 "^ D^ ^ "™
°ll &01
u oa
Exposures were intermittent: 8 hrs/d, 5 d/wk
t); Broderson etal. (1976)
ining of nasal epithelium
t); Broderson etal. (1976)
Nasal irritation (rat);
ro cy ro
c ^^
ro -c <*
c 1— m
C LJ_
^
i 1.1 ill 1
— < .SP >. rog— on
-S ™ — ^ -- '1 ra ™
g III =| f !°
3 |iS ||| |
Ji: .SP u ^ O
.y 5 QJ "c i!n
£ ° S
Coon etal. (1970)
Effects on the lung EXPERIMENTAL ANIMAL STUDIES Effects on the upper respiratory tract
Figure 1-1. Exposure-response array of respiratory effects following inhalation exposure to ammonia.
This document is a draft for review purposes only and does not constitute Agency policy,
1-14
DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Mode-of-Action Analysis—Respiratory Effects
1 Data on the potential mode of action for respiratory effects associated with chronic
3 exposure to ammonia are limited. However, acute exposure data demonstrate that injury to
4 respiratory tissues is primarily due to ammonia's alkaline (i.e., caustic) properties from the
5 formation of hydroxide ion when it comes in contact with water and is solubilized. Ammonia
6 readily dissolves in the moisture on the mucous membranes, forming ammonium hydroxide, which
7 causes liquefactive necrosis of the tissues. Specifically, ammonia directly denatures tissue proteins
8 and causes saponification of cell membrane lipids, which leads to cell disruption and death
9 (necrosis). In addition, the cellular breakdown of proteins results in an inflammatory response,
10 which further damages the surrounding tissues (Amshel etal.. 2000: Milleaetal.. 1989: Tarudi and
11 Golden. 1973).
12
13 Summary of Respiratory Effects
14 Evidence for respiratory toxicity associated with exposure to ammonia comes from studies
15 in humans and animals. Multiple occupational studies involving chronic exposure to ammonia in
16 industrial settings provide evidence of an increased prevalence of respiratory symptoms (Rahman
17 etal.. 2007: Ballal etal.. 1998) and decreased lung function (Rahman etal.. 2007: Mi etal.. 2001:
18 Bhat and Ramaswamy, 1993) (Table 1-1 and Appendix E, Section E.2). An increase in respiratory
19 effects was reported both with higher workplace ammonia concentrations (Rahman etal.. 2007:
20 Ballal etal., 1998) and with greater cumulative ammonia concentration (expressed in mg/m3-
21 years) (Alietal., 2001: Ballal etal., 1998). Additional evidence is provided by studies of asthma,
22 asthma symptoms, and pulmonary function in health care and cleaning workers, in a variety of
23 study designs and populations (Arif and Delclos, 2012: Dumas etal., 2012: Lemiere etal., 2012:
24 Vizcayaetal..2011: Zock etal.. 2007: Medina-Ramon etal.. 2006: Medina-Ramon etal.. 20051
25 (Table 1-2) and in studies of pulmonary function in livestock workers, specifically in the studies
26 that accounted for effects of co-exposures such as endotoxin and dust (Donham etal.. 2000:
27 Reynolds etal.. 1996: Donham etal.. 1995: Preller etal.. 1995: Heederik etal.. 1990) (Appendix E,
28 Table E-7). The livestock farmer studies, however, do not provide evidence of associations between
29 ammonia and respiratory symptoms. Controlled volunteer studies of ammonia inhalation and case
30 reports of injury in humans with inhalation exposure to ammonia provide additional support for
31 the respiratory system as a target of ammonia toxicity when inhaled (Appendix E, Section E.2).
32 Evidence from animal studies supports an association between inhaled ammonia and
33 respiratory effects. Short-term and subchronic animal studies show histopathological changes of
34 respiratory tissues in several animal species (lung inflammation in guinea pigs and rats; focal or
35 interstitial pneumonitis in monkeys, dogs, rabbits, and guinea pigs; pulmonary congestion in mice;
36 thickening of nasal epithelium in rats and pigs; nasal inflammation or lesions in rats and mice)
37 across different dosing regimens (Gaafar etal.. 1992: Brodersonetal.. 1976: Doig and Willoughby.
38 1971: Coon etal.. 1970: Anderson etal.. 19641 (Table 1-3 and Appendix E, Section E.3). In general,
39 responses in respiratory tissues increased with increasing ammonia exposure concentration.
This document is a draft for review purposes only and does not constitute Agency policy.
1-15 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Based on evidence of respiratory effects in multiple human and animal studies (including
2 epidemiological studies in different settings and populations), respiratory system effects are
3 identified as a hazard associated with inhalation exposure to ammonia.
4
5 1.1.2. Gastrointestinal Effects
6 Reports of gastrointestinal effects of ammonia in humans are limited to case reports
7 involving intentional or accidental ingestion of household cleaning solutions or ammonia inhalant
8 capsules (Dworkinetal.. 2004: Rosenbaum etal.. 1998: Christesen. 1995: Wasonetal.. 1990: Lopez
9 etal.. 1988: Klein etal.. 1985: Klendshoi and Reient. 1966] (Appendix E, Section E.2). Clinical signs
10 of gastrointestinal effects reported in these case studies include stomachache, nausea, diarrhea,
11 drooling, erythematous and edematous lips, reddened and blistered tongues, dysphagia, vomiting,
12 oropharyngeal burns, laryngeal and epiglottal edema, erythmatous esophagus with severe
13 corrosive injury, and hemorrhagic esophago-gastro-duodeno-enteritis. These effects appear to
14 reflect the corrosive properties of ammonia, and their relevance to effects associated with chronic
15 low-level exposure to ammonia is unclear.
16 The experimental animal toxicity database for ammonia lacks standard toxicity studies that
17 evaluate a range of tissues/organs and endpoints. Exposure to ammonia in drinking water has,
18 however, been associated with effects on the gastric mucosa. Evidence for this association comes
19 from animal studies (Hataetal.. 1994] designed to investigate the mechanisms by which the
20 bacterium Helicobacter pylori, which produces a potent urease that increases ammonia production,
21 may have a significant role in the etiology of chronic atrophic gastritis (Appendix E, Section E.3].
22 Statistically significant decreases of 40-60% in the thickness of the antral gastric mucosa were
23 reported in Sprague-Dawley rats administered ammonia in drinking water at concentrations
24 >0.01% for durations of 2-8 weeks (Tsujiietal., 1993: Kawano etal., 1991]: estimated doses in two
25 studies by the same group of investigators were 22 mg/kg-day (Kawano etal., 1991] and 33 mg/kg-
26 day (Tsujii etal.. 1993]. The magnitude of the decrease in gastric mucosal thickness increased with
27 dose and duration of ammonia exposure (Tsujii etal.. 1993: Kawano etal.. 1991]. Further, the
28 effect was more prominent in the mucosa of the antrum region of the stomach than in the body
29 region of the stomach.4 Antral gastric mucosal thickness decreased significantly (by 56-59% of the
30 tap water control] at 4 and 8 weeks of exposure to 0.01% ammonia in drinking water, but there
31 was no significant effect on the thickness of the body gastric mucosa. Similarly, the height of fundic
32 and pyloric glands in the gastric mucosa was decreased by approximately 30% in Donryu rats
33 exposed to ammonia in drinking water for up to 24 weeks at concentrations of 0.02 and 0.1%
34 (estimated doses of 28 and 140 mg/kg-day, respectively] (Hataetal.. 1994].
35 Mucosal cell proliferation and migration (as measured by 5-bromo-2'-deoxyuridine
36 labeling] were also significantly increased in rats exposed to ammonia (Tsujiietal., 1993]. The
37 authors observed that it was not clear whether mucosal cell proliferation was primarily stimulated
4The body is the main, central region of the stomach. The antrum is the distal part of the stomach near the
pyloric sphincter and adjacent to the body.
This document is a draft for review purposes only and does not constitute Agency policy.
1-16 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 directly by ammonia or indirectly by increased cell loss followed by compensatory cell
2 proliferation. Cell proliferation in the gastric mucosa was also affected in the 24-week drinking
3 water study in Donryu rats [Hataetal.. 1994). although the pattern differed from that reported by
4 Tsujii etal. [1993]. The labeling index in gastric mucosal glands was increased at earlier time
5 points (up to week 1 for fundic glands and up to week 4 for pyloric glands), suggesting enhanced
6 cell cycling subsequent to repeated erosion and repair. At later time points (up to 24 weeks of
7 exposure), however, the labeling index was decreased, a finding that the authors' attributed to
8 reduced capability of the generative cell zone of the mucosal region.
9 The gastric changes observed by Kawano etal. (1991), Tsujii etal. (1993), and Hata etal.
10 (1994) were characterized by the study authors as consistent with changes observed in human
11 atrophic gastritis; however, Kawano etal. (1991) and Tsujii etal. (1993) observed that no mucosal
12 lesions were found macroscopically or microscopically in the stomachs of rats after exposure to
13 ammonia in drinking water for 4-8 weeks, and Hataetal. (1994) reported that there was no
14 evidence of ammonia-induced gastritis or ulceration in rats following 24 weeks of exposure to 0.1%
15 ammonia in drinking water.
16 A relationship between ammonia ingestion and gastrointestinal effects is supported by
17 findings from three acute oral studies in rats following gavage administration of ammonium
18 hydroxide (Nagy etal.. 1996: Takeuchi etal.. 1995: Murakami etal.. 1990). Takeuchi etal. (1995)
19 reported hemorrhagic necrosis of the gastric mucosa in male Sprague-Dawley rats that received a
20 single gavage dose of ammonium hydroxide (concentration >1%). Nagy etal. (1996) observed
21 severe hemorrhagic mucosal lesions in female Sprague-Dawley rats 15 minutes after exposure to an
22 estimated dose of 48 mg/kg ammonium hydroxide via gavage. Lesions of the gastric mucosa,
23 including necrosis, were observed in male Sprague-Dawley rats 15 minutes after being given 1 mL
24 of ammonia by intubation at concentrations of 0.5-1%, but not at concentrations of 0.025-0.1%
25 (Murakami etal.. 1990).
26 The evidence of gastrointestinal effects in experimental animals following oral exposure to
27 ammonia is summarized in Table 1-4 and as an exposure-response array in Figure 1-2.
28
This document is a draft for review purposes only and does not constitute Agency policy.
1-17 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-4. Evidence pertaining to gastrointestinal effects in animals
Study design and references
Results3
Histopathologic changes of the gastric mucosa
Kawano et al. (1991)
Sprague-Dawley rat; male; 6/group
0, 0.01, or 0.1% in drinking water (0, 22, or
220 mg/kg-d)b for 2 or 4 wks
% change in thickness of mucosa compared to control:
Antrum Body
Wk2: 0,-1, 3%
Wk4: 0,-22,-30*%
Wk2: 0,-5,-20*%
Wk4: 0,-38*,-61*%
Tsujii et al. (1993)
Sprague-Dawley rat; male; 36/group
0 or 0.01% in drinking water (0 or 33 mg/kg-
d)c for 3 d or 1, 2, 4, or 8 wks; tap water
provided for the balance of the 8-wk study
% change in thickness of mucosa compared to control (at d 3, wks 1,
2, 4, and 8):
Antrum Body
D3: 0,8% D3: 0,5%
Wk 1: 0, -4% Wk 1: 0,1%
Wk2: 0,6% Wk2: 0,4%
Wk4: 0,-44%* Wk4: 0,-1%
Wk8: 0,-41%* Wk8: 0,-5%
(extracted from Figure 3 of Tsujii et al., 1993)
Hata et al. (1994)
Donryu rat; male; 6/group and time point
0, 0.02, or 0.1% in drinking water (0, 28, or
140 mg/kg-d)c for 1, 3, or 5 d and 1, 4, 8,12,
or 24 wks
% change in gland height compared to control (week 24):
Fundic region: 0, -18*, -34*%
Pyloric region: 0,-17*,-26*%
(estimated from Figure 3 of Hata et al., 1994)
% change in labeling index compared to control (week 24):
Fundic region: 0,-35*,-27*%
Pyloric region: 0,-17*,-11*%
aPercent change compared to control calculated as: (treated value - control value)/control value x 100.
bDoses were estimated based on a body weight of 230 g for male rats and an estimated drinking water intake of
50 mL/day (as reported by study authors).
cDoses were estimated based on an initial body weight of 150 g and an estimated drinking water intake of
50 mL/day (as reported by study authors).
dBody weights and drinking water intakes were not provided by the authors. Doses were estimated assuming a
body weight of 267 g [subchronic value for a male Sprague-Dawley rat, Table 1-2, (U.S. EPA, 1988)1 and a drinking
water intake of 37 mL/d [subchronic value for a male Sprague-Dawley rat, Table 1-5 (U.S. EPA, 1988)1.
*Statistically significantly different from the control (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy,
1-18 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1000
1
2
o
J
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
oo
^
00
.§100
0)
1/1
o
0
10
T
• LOAEL Vertical lines show
range of doses in
ANOAEL study.
• Additional doses
1
thickness of gastric mucosa -i, thickness of gastric mucosa; -i, height of gastric mucosal
(rat); increased cell migration (rat); glands; suppressed cell cycle
Kawanoetal. (1991) Tsujii et al. (1993) (rat);
Hataetal. (1994)
Gastric mucosa
Figure 1-2. Exposure-response array of gastrointestinal effects following oral
exposure to ammonia.
Mode-of-Action Analysis—Gastrointestinal Effects
The alkalinity of the ammonia solution does not seem to play a direct role in the gastric
effects associated with ammonia. An ammonia solution (pH 10.3) produced dose-related acute
macroscopic mucosal lesions, whereas a glycine-sodium hydroxide buffer (pH 10.3) or ammonium
chloride (pH 4.5) did not [Tsujii etal., 1992a). Rather, the available evidence suggests that the
ability of ammonia to damage the gastric mucosa is related to its ionization state. Ammonia (NHs)
(in its non-ionized state) can easily penetrate cell membranes, whereas the ionized form (NH4+) is
less permeable to cell membranes [Tsujii etal.. 1992a). The finding that antral and body regions of
the rat stomach mucosa responded differently following administration of 3 3 mg/kg-day ammonia
in drinking water for 8 weeks [Tsujii etal., 1993) is consistent with the influence of ionization. The
hydrogen chloride secreted by the mucosa in the body of the stomach resulted in a lower pH in the
body mucosa and a corresponding decrease in the ratio of ammonia to NH4+. In contrast, in the
antral mucosa (a nonacid-secreting area), the pH was higher, the ratio of ammonia to NH4+ was
increased, and measures of gastric mucosal changes were increased compared to those observed in
the stomach body where there was relatively higher exposure to NH4+.
This document is a draft for review purposes only and does not constitute Agency policy,
1-19 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Several specific events that may contribute to the induction of gastric mucosal changes by
2 ammonia have been proposed. Increased cell vacuolation and decreased viability of cells were
3 associated with increasing ammonia concentration in an in vitro system [Megraudetal.. 1992): the
4 effect was not linked to pH change because of the high buffering properties of the medium. Using
5 an in situ rat stomach model, hemorrhagic mucosal lesions induced by ammonia were associated
6 with the rapid release and activation of cathepsins, which are mammalian cysteine proteases that
7 are released from lysosomes or activated in the cytosol and can be damaging to cells, tissues, or
8 organs [Nagy etal., 1996]. Ammonia also appears to inhibit cellular and mitochondrial respiration,
9 possibly by elevating intracellular or intraorganelle pH or by impairing adenosine triphosphate
10 synthesis [Tsujiietal.. 1992a]. Mori etal. [1998] proposed a role for increased release of
11 endothelin-1 and thyrotropin-releasing hormone from the gastric mucosa in ammonia-induced
12 gastric mucosal injury based on findings in rats given ammonia intragastrically. Tsujiietal.
13 [1992b] suggested that ammonia may accelerate mucosal cell desquamation and stimulate cell
14 proliferation by a compensatory mechanism. Overall, although hypotheses have been proposed, a
15 specific mechanism(s) by which ammonia may induce cellular toxicity has not been established,
16
17 Summary of Gastrointestinal Effects
18 Evidence that oral exposure to ammonia causes gastrointestinal effects is based on human
19 case reports and studies in rats that focused on mechanistic understandings of effects of ammonia
20 on the gastric mucosa. Acute gastric toxicity observed in case reports involving intentional or
21 accidental ingestion of cleaning solutions or ammonia inhalant capsules appears to reflect the
22 corrosive properties of ammonia. Whether these acute effects are relevant to toxicity following
23 chronic low-level ammonia exposure is not known. Indirect evidence for the biological plausibility
24 of gastric tissue as a target of ammonia toxicity is provided by the association between the
25 bacterium H. pylori, which produces urease that catalyzes urea into ammonia, and human diseases
26 of the upper gastrointestinal tract (including chronic gastritis, gastric ulcers, and stomach cancer).
27 Three mechanistic studies in male rats [Hataetal., 1994: Tsujiietal., 1993: Kawano etal.,
28 1991] provide consistent evidence of changes in the gastric mucosa associated with exposure to
29 ammonia in drinking water, including decreased thickness or gland height These gastric changes
30 did not correlate, however, with other lesions in the stomach. No evidence of other microscopic
31 lesions, gastritis, or ulceration was found in the stomachs of these rats. It is also interesting to note
32 that chronic toxicity studies of other ammonia compounds have not identified the gastrointestinal
33 tract as a target of ammonia toxicity. For example, no treatment-related changes in the stomach or
34 other parts of the gastrointestinal tract were observed in Wistar rats exposed to ammonium
35 chloride in the diet for 130 weeks at doses up to 1,200 mg/kg-day [Lina and Kuijpers, 2004] or in
36 F344 rats exposed to ammonium sulfate for 104 weeks at a dose up to 1,371 mg/kg-day [Ota etal.,
37 2006] (Appendix C, Table C-l]. Therefore, while drinking water studies with a mechanistic focus
38 provide evidence for ammonia-related changes in rat gastric mucosa, adverse changes of the
39 gastrointestinal tract were not identified in standard toxicity bioassays of ammonia compounds.
This document is a draft for review purposes only and does not constitute Agency policy.
1-20 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Mechanistic studies in rodent models support the biological plausibility that ammonia
2 exposure may be associated with gastric effects in humans. Conditions that favor the un-ionized
3 form of ammonia (pH > 9.25) facilitate penetration of the cell membrane and are associated with
4 greater gastric cytotoxicity. In summary, the evidence primarily from human case reports as
5 supported by mechanistic studies in experimental animals suggests that gastric effects are a
6 potential hazard associated with oral exposure to ammonia.
7
8 1.1.3. Immune System Effects
9 A limited number of studies have evaluated the immunotoxicity of ammonia in human
10 populations and in experimental animal models. Immunological function was evaluated in two
11 independent investigations of livestock farmers exposed to ammonia via inhalation.
12 Immunoglobulin G- (IgG) and E-specific (IgE) antibodies for pig skin and urine [Crook etal., 1991],
13 elevated neutrophils from nasal washes, and increased white blood cell counts [Cormier etal..
14 2000] were reported. These data on immunological function are suggestive of immunostimulatory
15 effects; however, the test subjects were also exposed to a number of other respirable agents in
16 addition to ammonia, such as endotoxin, bacteria, fungi, and mold, that are known to stimulate
17 immune responses. Data in humans following exposure to ammonia only are not available.
18 Animal studies that examined ammonia immunotoxicity were conducted using short-term
19 inhalation exposures and were measured by three general types of immune assays: host resistance,
20 T cell proliferation, and delayed-type hypersensitivity. Immunotoxicity studies of ammonia using
21 measures of host resistance provide the most relevant data for assessing immune function since
22 they directly measure ability of the immune system to control microorganism growth. Other
23 available studies of ammonia employed assays that evaluated immune function. Changes in
24 immune cell populations without corresponding functional data are considered to be the least
25 predictive, and studies that looked only at these endpoints [Gustinetal.. 1994: Neumann etal..
26 1987] were excluded from the hazard identification for ammonia.
27 Several host resistance studies utilized lung pathogens to assess bacterial clearance
28 following ammonia exposure; however, these studies were not designed to discriminate between
29 direct immunosuppression associated with ammonia exposure or immune effects secondary to
30 damage to the protective mucosal epithelium of the respiratory tract. The available studies also do
31 not correlate increased bacterial colonization with reduced immune function. Lung lesions, both
32 gross and microscopic, were positively correlated with ammonia concentration in F344 rats
33 continuously exposed to ammonia in an inhalation chamber for 7 days prior to inoculation with 108
34 colony forming units [CPU] of Mycop/asmapu/mon/s followed by up to 42 days of ammonia
35 exposure post inoculation [Brodersonetal., 1976]. (Inoculation with the respiratory pathogen
36 M. pulmonis causes murine respiratory mycoplasmosis [MRM] characterized by lung lesions.] The
37 incidence of lung lesions was significantly increased at ammonia concentrations >35 mg/m3,
38 suggesting that ammonia exposure decreased bacterial clearance resulting in the development of M.
39 pulmonis-induced MRM. However, increasing ammonia concentration was not associated with
40 increased CPU of M. pulmonis isolated from the respiratory tract The high number of inoculating
This document is a draft for review purposes only and does not constitute Agency policy.
1-21 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 CPU could have overwhelmed the innate immune response and elicited a maximal response that
2 could not be further increased in immunocompromised animals.
3 Conversely, significantly increased CPU of M. pulmonis bacteria isolated in the trachea, nasal
4 passages, lungs, and larynx were observed in F344 rats continuously exposed to 71 mg/m3
5 ammonia for 7 days prior to M. pulmonis (104-106 CPU) inoculation and continued for 28 days post
6 inoculation [Schoeb etal.. 1982). This increase in bacterial colonization indicates a reduction in
7 bacterial clearance following exposure to ammonia. Lesions were not assessed in this study.
8 OF1 mice exposed to 354 mg/m3 ammonia for 7 days prior to inoculation with a 50% lethal
9 dose [LDso] ofPasteurella multocida exhibited significantly increased mortality compared to
10 controls (86 versus 50%, respectively); however, an 8-hour exposure was insufficient to affect
11 mortality [Richard etal., 1978a]. The authors suggested that the irritating action of ammonia
12 destroyed the tracheobronchial mucosa and caused inflammatory lesions thereby increasing
13 sensitivity to respiratory infection with prolonged ammonia exposure.
14 Pig studies support the findings observed in the rodent studies that ammonia exposure
15 increases the colonization of respiratory pathogens. Andreasen et al. [2000] demonstrated that
16 63 days of ammonia exposure increased the number of bacterial positive nasal swabs following
17 inoculation with P. multocida and Mycoplasma hyopneumoniae; however, the effect was not dose
18 responsive and did not result in an increase in lung lesions. Additional data obtained from pigs
19 suggest that ammonia exposure eliminates the commensal flora of the nasal cavities, which allows
20 for increased colonization of P. multocida; however, this effect abates following cessation of
21 ammonia exposure [Hamilton etal., 1999: Hamilton etal., 1998].
22 Suppressed cell-mediated immunity and decreased T cell proliferation was observed
23 following ammonia exposure. Using a delayed-type hypersensitivity test to evaluate cell-mediated
24 immunity, Hartley guinea pigs were vaccinated with Mycobacterium bovis bacillus Calmette-Guerin
25 (BCG] and exposed to ammonia followed by intradermal challenge with a purified protein
26 derivative (PPD]. Dermal lesion size was reduced in animals exposed to 64 mg/m3 ammonia,
27 indicating immunosuppression [Targowski et al.. 1984]. Blood and bronchial lymphocytes
28 harvested from naive guinea pigs treated with the same 3-week ammonia exposure and stimulated
29 with phytohaemagglutinin or concanavalin A demonstrated reduced T cell proliferation [Targowski
30 etal., 1984]. Bactericidal activity in alveolar macrophages isolated from ammonia-exposed guinea
31 pigs was not affected. Lymphocytes and macrophages isolated from unexposed guinea pigs and
32 treated with ammonia in vitro showed reduced proliferation and bactericidal capacity only at
33 concentrations that reduced viability, indicating nonspecific effects of ammonia-induced
34 immunosuppression [Targowski etal.. 1984]. These data suggest that T cells may be the target of
35 ammonia since specific macrophage effects were not observed.
36 The evidence of immune system effects in experimental animals exposed to ammonia is
37 summarized in Table 1-5 and as an exposure-response array in Figure 1-3.
38
This document is a draft for review purposes only and does not constitute Agency policy.
1-22 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-5. Evidence pertaining to immune system effects in animals
Study design and reference
Results
Host resistance
Broderson et al. (1976)
F344 rat; male and female; 11-12/sex/ group
<5 (control), 25, 50, 100, or 250 ppm (<3.5 [control], 18, 35,
71, or 177 mg/m3), 7 d (continuous exposure) pre-
inoculation/28-42 d post-inoculation with M. pulmonis
Schoeb et al. (1982)
F344 rat; 5-15/group (sex unknown)
<2 or 100 ppm (<1.4 [control] or 71 mg/m3), 7 d
(continuous exposure) pre-inoculation/28 d post-
inoculation with M. pulmonis
Richard et al. (1978a)
OF1 mouse; male; 99/group
0 or 500 ppm (0 or 354 mg/m3), 8 hrs or 7 d (continuous
exposure), prior to infection with P. multocida
Andreasen et al. (2000)
Landrace X large white pigs; 10/group (sex unknown)
<5 (control), 50, or 100 ppm (3.5, 35, or 71 mg/m3), 63 d
(continuous exposure) inoculated with M. hyopneumoniae
on day 9 and P. multocida on d 28, 42, and 56
Hamilton et al. (1998)
Large white pigs; 4-7/group (sex unknown)
0 or 20 ppm (0 or 14 mg/m3), 14 d (continuous exposure),
inoculated with P. multocida on d 0
Hamilton et al. (1999)
Large white pigs; 5/group (sex unknown)
0 or 50 ppm (0 or 35 mg/m3), 1 wk pre-inoculation with P.
multocida, 3 wks post-inoculation
% of animals with gross lung lesions: 16, 46, 66*, 33,
and 83%
No effect on CPU.
/T" bacterial colonization (as a result of reduced
bacterial clearance).
% Mortality: 50 and 86%*
% of animals with positive day 49 nasal swab:
24, 100*, and 90%*
1" bacterial colonization
1" bacterial colonization
Bacteria isolated from nasal cavities: 3.18 and 4.30*
CPU
T cell proliferation
Targowski et al. (1984)
Hartley guinea pig; 8/group (sex unknown)
<15, 50, or 90 ppm (<11 [control], 35, or 64 mg/m3), 3 wks
(continuous exposure)
•^ proliferation in blood and bronchial T cells.
Delayed-type hypersensitivity
Targowski et al. (1984)
Hartley guinea pig, BCG immunized; 8/group (sex unknown)
<15, 50, or 90 ppm (<11 [control], 35, or 64 mg/m3), 3 wks
(continuous exposure) followed by PPD challenge
Mean diameter of dermal lesion (mm): 12, 12.6, and
8.7*
*Statistically significantly different from the control (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy,
1-23 DRAFT—DO NOT CITE OR QUOTE
-------
1000
E
"SB 100
O
C
(U
u
C
O
u
^ 10 -
O
Q.
X
I
I I
ILOAEL
ANOAEL
• Additional concentrations
Vertical lines show range of
concentrations in study.
T" incidence of f" bacterial
gross lung lesions colonization (rat);
(rat); Broderson et Schoebetal.
al. (1976) (1982)
1s mortality 1s bacterial 1s bacterial 1s bacterial
(mouse); Richard colonization (pig); colonization (pig); colonization (pig);
etal. (1978b) Andreasen etal. Hamilton et al. Hamilton et al.
(2000) (1998) (1999)
Host resistance
\|/ proliferation in
blood and
bronchial
lymphocytes
(guinea pig);
Targowski et al.
(1984)
Tcell proliferation
\|/ dermal lesion
size (guinea pig);
Targowski et al.
(1984)
Delayed-type
hypersensitivity
Figure 1-3. Exposure-response array of immune system effects following inhalation exposure to ammonia.
This document is a draft for review purposes only and does not constitute Agency policy.
1-24 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Summary of Immune System Effects
1 The evidence for ammonia immunotoxicity is based on epidemiological and animal studies.
3 Available epidemiological studies that addressed immunological function are confounded by
4 exposures to a number of other respirable agents that have been demonstrated to be
5 immunostimulatory. Single-exposure human studies of ammonia evaluating immune endpoints are
6 not available. Therefore, human studies are not particularly informative for evaluating whether
7 ammonia has immunotoxic properties.
8 Animal studies provide consistent evidence of elevated bacterial growth following ammonia
9 exposure. This is supported by observations of lung lesions [Brodersonetal., 1976], elevated CPU
10 [Schoeb etal.. 1982). and increased mortality [Richard etal.. 1978a] in rats or mice exposed to
11 ammonia; however, the findings from the Brodersonetal. [1976] study (which described the
12 percent of animals with gross lesions) were not dose-responsive, and the other studies used single
13 concentrations of ammonia and therefore did not provide information on dose-response. A single
14 study suggested that T cells are inhibited by ammonia [Targowskietal., 1984], but the data were
15 not dose responsive.
16 Overall, the evidence in humans and animals indicates that ammonia exposure may be
17 associated with immunotoxicity, but it is unclear if elevated bacterial colonization is the result of
18 damage to the protective mucosal epithelium of the respiratory tract or the result of suppressed
19 immunity. Therefore, the evidence does not support the immune system as a potential hazard of
20 ammonia exposure.
21
22 1.1.4. Other Systemic Effects
23 Although the majority of information suggests that ammonia induces effects in and around
24 the portal of entry, there is limited evidence that ammonia can produce effects on organs distal
25 from the portal of entry, including the liver, adrenal gland, kidney, spleen, and heart Alterations in
26 liver function, based on elevated mean levels of aspartate aminotransferase (AST], alanine
27 aminotransferase (ALT], and blood urea, decreased hemoglobin, and inhibition of catalase and
28 monoamine oxidase (MAO] activities, were reported in workers in an Egyptian urea fertilizer
29 production plant (Hamid and El-Gazzar, 1996]: there were no direct measurements of workplace
30 exposure to ammonia and information on control for potentially confounding exposures was not
31 provided (Table 1-6].
32 Evidence of liver toxicity in animals comes from observations of histopathological
33 alterations in the liver. Fatty changes in liver plate cells were consistently reported at exposure
34 concentrations >470 mg/m3 ammonia in rats, guinea pigs, rabbits, dogs, and monkeys following
35 identical subchronic inhalation exposure regimens (Coon etal., 1970]. Congestion of the liver was
36 observed in guinea pigs following subchronic and short-term inhalation exposure to 35 and
37 120 mg/m3 (Anderson etal.. 1964: Weatherby. 1952]: no liver effects were observed in similarly
38 exposed mice at 14 mg/m3 (Anderson etal.. 1964: Weatherby. 1952].
39 No histopathological or hematological effects were observed in rats, guinea pigs, rabbits,
40 dogs, or monkeys when these animals were repeatedly, but not continuously, exposed to ammonia
This document is a draft for review purposes only and does not constitute Agency policy.
1-25 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 even at high concentrations (e.g., 770 mg/m3 for 8 hours/day, 5 days/week; Table 1-8 ), suggesting
2 that animals can recover from intermittent exposure to elevated ammonia levels [Coonetal., 1970].
3 In addition, no effects on nonrespiratory system organs were observed in mice exposed to 14
4 mg/m3 for up to 6 weeks [Anderson etal., 1964].
5 Adrenal effects were observed in animals following subchronic and short-term exposure to
6 ammonia. Increased mean adrenal weights and fat content of the adrenal gland, as well as
7 histological changes in the adrenal gland (enlarged cells of the zona fasiculata of the adrenal cortex
8 that were rich in lipid], were observed in rabbits exposed via gavage to ammonium hydroxide for
9 durations ranging from 5.5 days to 17 months (Fazekas, 1939]. The strength of these findings is
10 limited by inadequate reporting and study design. A separate study identified early degenerative
11 changes in the adrenal glands of guinea pigs exposed to 120 mg/m3 ammonia by inhalation for
12 18 weeks (Weatherby. 1952]. providing additional limited evidence for effects on the adrenal gland.
13 Evidence that inhaled ammonia can affect the kidney and spleen is limited to studies in
14 experimental animals. Nonspecific degenerative changes in the kidneys (not further described] in
15 rats exposed to 262 mg/m3 ammonia for 90 days were reported (Coonetal., 1970].
16 Histopathological evaluation of other animal species in the same study exposed to 470 mg/m3, an
17 ammonia concentration that induced a high rate of mortality in rats, consistently showed
18 alterations in the kidneys (calcification and proliferation of tubular epithelium; incidence not
19 reported]. Exposure of guinea pigs to inhaled ammonia at a concentration of 120 mg/m3 for 18
20 weeks (but not 6 or 12 weeks] resulted in histopathological alterations (congestion] of the kidneys
21 and spleen, although incidence was not reported (Weatherby, 1952]. Enlarged and congested
22 spleens were reported in guinea pigs exposed to 35 mg/m3 ammonia for 6 weeks in a separate
23 study (Anderson etal., 1964].
24 Myocardial fibrosis was observed in monkeys, dogs, rabbits, guinea pigs, and rats following
25 subchronic inhalation exposure to 470 mg/m3 ammonia; no changes were observed at lower
26 concentrations (Coonetal.. 1970]. At the same concentration, ocular irritation (characterized as
27 heavy lacrimation, erythema, discharge, and ocular opacity of the cornea] was also reported by
28 Coonetal. (1970] in dogs and rabbits, but was not observed in similarly exposed monkeys or rats.
29 Additionally, there is limited evidence of biochemical or metabolic effects of acute or short-
30 term ammonia exposure. Evidence of slight acidosis, as indicated by a decrease in blood pH, was
31 reported in rats exposed to 18 or 212 mg/m3 ammonia for 5 days; the study authors stated that
32 differences in pH leveled off at 10 and 15 days (Manninenetal., 1988]. In another study, blood pH
33 in rats was not affected by exposure to ammonia at concentrations up to 818 mg/m3 for up to
34 24 hours (Schaerdel et al.. 1983].
35 Encephalopathy related to ammonia may occur in humans following disruption of the
36 body's normal homeostatic regulation of the glutamine and urea cycles, e.g., due to severe liver or
37 kidney disease resulting in elevated ammonia levels in blood (Minanaetal.. 1995: Souba. 1987].
38 Acute inhalation exposure studies have identified alterations in amino acid levels and
39 neurotransmitter metabolism (including glutamine concentrations] in the brain of rats and mice
40 (Manninen and Savolainen. 1989: Manninen et al.. 1988: Sadasivudu etal.. 1979: Sadasivudu and
This document is a draft for review purposes only and does not constitute Agency policy.
1-26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review of Ammonia
Radha Krishna Murthy, 1978]. It has been suggested that glutamate and y-aniino butyric acid play a
role in ammonia-induced neurotoxicity [Tones, 2002]. There is no evidence, however, that
ammonia is neurotoxic in humans or animals following chronic inhalation exposures.
In the only study of the reproductive and developmental toxicity of ammonia, no changes in
reproductive or developmental endpoints were found between two groups of female pigs
(crossbred gilts] exposed to ammonia via inhalation for 6 weeks at mean concentrations of 5 or
25 mg/m3 and then mated [Diekman et al., 1993]. A control group without ammonia exposure was
not evaluated. Age at puberty did not differ significantly between the two groups. Gilts exposed to
25 mg/m3 ammonia weighed 7% less (p < 0.05] atpuberty than those exposed to 5 mg/m3;
however, body weights of the two groups were similar at gestation day 30. Conception rates in the
mated females were similar between the two groups (94.1 versus 100% in low- versus high-
exposure groups]. At sacrifice on day 30 of gestation, there were no significant differences between
the two exposed groups in body weights of the pregnant gilts, number of corpora lutea, number of
live fetuses, or weight and length of the fetuses. The strength of the findings from this study are
limited by the absence of a control group and possible confounding by exposures to bacterial and
mycoplasm pathogens.
The evidence of systemic toxicity in humans and experimental animals exposed to ammonia
is summarized in Tables 1-6 and 1-7 and as an exposure-response array in Figure 1-4.
Table 1-6. Evidence pertaining to other systemic effects in humans
Study design and reference
Results
Hamid and EI-Gazzar (1996) (Egypt)
Urea fertilizer plant workers (all men); 30 exposed and
30 control subjects (from administrative departments).
Average employment duration: 12 yrs
Exposure: No direct measurement of ammonia
concentrations; blood urea used as surrogate measure
Outcome: Blood sample measurements of AST, ALT,
hemoglobin, and catalase and monoamine oxidase
enzyme activities
AST, ALT, and blood urea in exposed workers;
hemoglobin and inhibition of catalase and MAO.
20
21
22
This document is a draft for review purposes only and does not constitute Agency policy.
1-27 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-7. Evidence pertaining to other systemic effects in animals
Study design and reference
Results
Liver effects
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (5. sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Sprague-Dawley or Long-Evans rat; male and female; 15-
51/group
0 or 40 mg/m3 for 114 d or 127, 262, or 470 mg/m3 for 90 d
Anderson et al. (1964)
Swiss albino mouse; male and female; 4/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d
Weatherbv(1952)
Guinea pig (strain not specified); male; 6-12/group
0 or 170 ppm (0 or 120 mg/m3) for 6 hrs/d, 5 d/wk for 6, 12 or
18 wks
Anderson et al. (1964)
Guinea pig (strain not specified); male and female; 2/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d or 50 ppm (35 mg/m3)
for 42 d
No histopathologic changes observed.
Fatty liver changes in plate cells at 470 mg/m3.3
Fatty liver changes in plate cells at
470 mg/m3.a'b
No visible signs of liver toxicity.
Congestion of the liver at 18 wks, not observed
at earlier times.3
Congestion of the liver at 35 mg/m3 for 42 d.a
Adrenal gland effects
Weatherbv(1952)
Guinea pig (strain not specified); male; 6-12/group
0 and 170 ppm (0 and 120 mg/m3) 6 hrs/d, 5 d/wk for 6, 12, or
18 wks
"Early" degenerative changes in the adrenal
gland (swelling of cells, degeneration of the
cytoplasm with loss of normal granular
structure) at 18 wks, not observed at earlier
times.3
This document is a draft for review purposes only and does not constitute Agency policy.
1-28 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-7. Evidence pertaining to other systemic effects in animals
Study design and reference
Fazekas (1939)
Rabbit (strain and sex not specified); 16-33/group
50-80 mL of a 0.5 or 1.0% ammonium hydroxide solution by
gavage; initially every other day, later daily; duration ranged
from 5.5 d to 17 mo; estimated dose: 61-110 and 120-230
mg/kg-d, respectively0
Results
Mean adrenal weight compared to control: 95%
Fat content of adrenal gland compared to
control: 4.5-fold 1\
Note: results by dose level were not provided.
Kidney and spleen effects
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (5. sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
Coon et al. (1970)
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Sprague-Dawley or Long-Evans rat; male and female; 15-
51/group
0 or 40 mg/m3 for 114 d or 127, 262, or 470 mg/m3 for 90 d
Anderson et al. (1964)
Swiss albino mouse; male and female; 4/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d
Weatherbv(1952)
Guinea pig (strain not specified); male; 6-12/group
0 or 170 ppm (0 or 120 mg/m3) 6 hrs/d, 5 d/wk for 6, 12, or
18 wks
Anderson et al. (1964)
Guinea pig (strain not specified); male and female; 2/group
0 or 20 ppm (0 or 14 mg/m3) for 7-42 d or 50 ppm (35 mg/m3)
for 42 d
No histopathologic changes observed.
Calcification and proliferation of renal tubular
epithelium at 470 mg/m3.3
Calcification and proliferation of renal tubular
epithelium at 470 mg/m3.a'b
No visible signs of toxicity.
Congestion of the spleen and kidneys.3
Enlarged and congested spleens at 35 mg/m3.3
Myocardial effects
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
No histopathologic changes observed.
This document is a draft for review purposes only and does not constitute Agency policy.
1-29 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-7. Evidence pertaining to other systemic effects in animals
Study design and reference
Coon et al. (1970)
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (5. sciureus); male; 3/group
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Sprague-Dawley or Long-Evans rat; male and female; 15-
51/group
0 or 40 mg/m3 for 114 d or 127, 262, or 470 mg/m3 for 90 d
Results
Myocardial fibrosis at 470 mg/m3.a'b
Myocardial fibrosis at 470 mg/m3.3
Ocular effects
Coon et al. (1970)
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
New Zealand albino rabbit; male; 3/group
Princeton-derived guinea pig; male and female; 15/group
Squirrel monkey (S. sciureus); male; 3/group
Beagle dog; male; 2/group
0, 155, or 770 mg/m3 8 hrs/d, 5 d/wk for 6 wks
Coon et al. (1970)
Sprague-Dawley and Long-Evans rat; male and female; 15-
51/group
0 or 40 mg/m3 for 114 d or 127, 262, or 470 mg/m3 for 90 d
Coon et al. (1970)
New Zealand albino rabbit; male; 3/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
Coon et al. (1970)
Beagle dog; male; 2/group
0 or 40 mg/m3 for 114 d or 470 mg/m3 for 90 d
No ocular irritation observed.
No ocular irritation observed.
No ocular irritation observed.
Erythema, discharge, and ocular opacity over
%-Vi of cornea at 470 mg/m3.3
Heavy lacrimation at 470 mg/m3.3
Blood pH changes
Manninen etal. (1988)
Wistar rat; female; 5/group
0, 25 or 300 ppm (0, 18, or 212 mg/m3) 6 hrs/d for 5, 10 or 15 d
Schaerdel et al. (1983)
CrhCOBS CD(SD) rat; male; 8/group [blood pO2 based on n = 5]
15, 32, 310, or 1,157 ppm (11, 23, 219, or 818 mg/m3) for
0 (control), 8, 12, or 24 hrs
•^ blood pH at 5 days; pH differences "leveled
off at later time points (data not shown)".
Blood pH (day 5): 7 A3, 7.34*, 7.36*
T* blood pO2 at 11 and 23 mg/m3 at 8-, 12-, and
24-hr time points; no change at higher
concentrations; no change in blood pH.
Percent change in pO2from time 0 (at 24 hours
of exposure*: 20*, 17*, 1, -2%
This document is a draft for review purposes only and does not constitute Agency policy.
1-30 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-7. Evidence pertaining to other systemic effects in animals
Study design and reference
Results
Amino acid levels and neurotransmitter metabolism in the brain
Manninen and Savolainen (1989)
Wistar rat; female; 5/group
0, 25, or 300 ppm (0, 18, or 212 mg/m3) 6 hrs/d for 5 d
Manninen et al. (1988)
Wistar rat; female; 5/group
0, 25, or 300 ppm (0, 18, or 212 mg/m3) 6 hrs/d for 5, 10, or
15 d
% change compared to control?
Brain glutamine: 42*, 40*%
% change compared to control at 212 mg/m3?
Blood glutamine (5, 10, 15 d): 44*, 13, 14%
Brain glutamine (5, 10, 15 d): 40*, 4, 2%
Reproductive and developmental effects
Diekman et al. (1993)
Crossbred gilt (female pig); 4.5 mo old; 40/group
7 ppm (5 mg/m3), range 4-12 ppm (3-8.5 mg/m3) or 35 ppm
(25 mg/m3), range 26-45 (18-32 mg/m3) for 6 wksf
No change in any of the reproductive or
developmental parameters measured (age at
puberty, conception rates, body weight of
pregnant gilts, number of corpora lutea,
number of live fetuses, and weight or length of
fetuses).
Incidence data not provided.
bExposure to 470 mg/m3 ammonia increased mortality in rats.
cAmmonia doses estimated using assumed average default body weight of 3.5-4.1 kg for adult rabbits (U.S. EPA,
1988).
Measurements at time zero were used as a control; the study did not include an unexposed control group.
ePercent change compared to control calculated as: (treated value - control value)/control value x 100.
fA control group was not included. Prior to exposure to ammonia, pigs were also exposed naturally in
conventional grower units to Mycoplasma hypopneumoniae and Pasteurella multocida, which cause pneumonia
and atrophic rhinitis, respectively.
*Statistically significantly different from the control (p < 0.05).
This document is a draft for review purposes only and does not constitute Agency policy.
1-31 DRAFT—DO NOT CITE OR QUOTE
-------
1000
,£
bo
O
1100
c
01
o
c
o
o
Ol
in
O
Q.
X
10
i
i
\
¥
•
A
/
, ^
(
> • •
•
^
'
i
i
':
i
,
A
' '
4
. .
':
i
1 i
i
• 1
TBLOAEL ANOAEL
A A
/
^
(
^
I
1
1
1
l i
* '
»
»
> /
• \
\ i
• Additional concentrations
Vertical lines show range of concentrations in study
• I
A A
1
\
* Exposures were intermittent : 6 or 8 hrs/d, 5 d/wk
:opathological changes (rat, rabbit, i
to
£
O
"2.
g, monkey, dog); Coon et al. (1970) *i
ver changes in plate cells (rat, rabbit,!
pig, monkey, dog); Coon et al. (1970)i
Q_ in TO
ro 3" c
C TO '5
'5 LL. 00
00
:atty 1 iver changes i n pi ate cells (rat);!
Coon etal. (1970) |
No livereffects(mouse);;:
Anderson etal. (1964) i
Liver congestion (guinea pig);i
Weatherby (1952)* |
Liver congestion (guinea pig);i
Anderson etal. (1964) ]
Liver effects
degenerative changes in the adrenali
id (guinea pig); Weatherby (1952)* i
..^ i_
ro oo
LLJ
Adrenal
gland
effects
:opathological changes (rat, rabbit, i
to
£
O
^
g, monkey, dog); Coon et al. (1970) *|
Droliferation renal tubular epithelium;
Q_ ^
TO ^
C ^
^ ro
00 _<_>
<_)
ro
U
pig, monkey,dog); Coonetal. (1970);
Icification & proliferation of renal i
ro ro
QJ U
C
CUD
~
_Q
TO
• — '
ar epithelium (rat); Coon etal. (1970)i
No kidney or spleen effects (mouse);;
Anderson etal. (1964) |
ofthespleen & kidneys (guinea pig);|
Weatherby (1952)* |
ged & congested spleen (guinea pig);i
Anderson etal. (1964) i
H i_
_Q '° "E
^ to LLJ
*-> OJ
00
c
o
u
Kidney and spleen effects
:opathological changes (rat, rabbit, i
g, monkey, dog); Coon et al. (1970) *]
lyocardial fibrosis (rabbit, guinea pig,i
to Q. ,>
o ^
'5
00
1 1
"re -°
tj ^
c ~
o P
0 ro
U LJ
"oo ^
o S
OJ
c
o
E
Coon etal. (1970) ;
Myocardial effects
ation observed (guinea pig, monkey);:
^±
t
o
z
Coon etal. (1970) |
tion observed (rat, rabbit, guinea pig,:
ro
~
.b=
O
'nkey, dog); Coon etal. (1970)* i
No irritation observed (rat);:
o
E
Coon etal. (1970) i
discharge & comeal opacity (rabbit);!
Coon etal. (1970) |
Heavy lacrimation(dog);;:
nT
E
01
J-.
LLJ
Coon etal. (1970) |
Ocular Irritation
Figure 1-4. Exposure-response array of systemic effects following inhalation exposure to ammonia.
This document is a draft for review purposes only and does not constitute Agency policy,
1-32
DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Summary of Other Systemic Effects
1 Effects of ammonia exposure on organs distal from the portal of entry are based largely on
3 evidence in animals and, to a more limited extent, in humans. Effects on various organs, including
4 liver, adrenal gland, kidney, spleen, and heart, were observed in several studies that examined
5 responses to ammonia exposure in a number of laboratory animal species. While effects on many
6 of these organs were observed in multiple species, including monkey, dog, rabbit, guinea pig, and
7 rat, effects were not consistent across exposure protocols. Evidence of ocular irritation in
8 experimental animals was inconsistently observed, and then only at high ammonia concentrations
9 (470 mg/m3).
10 Studies of ammonia toxicity that examined other systemic effects were all published in the
11 older toxicological literature. The only oral study of ammonium hydroxide was published in 1939
12 [Fazekas. 1939). and three subchronic inhalation studies were published between 1952 and 1970
13 [Coonetal., 1970: Anderson etal., 1964: Weatherby, 1952]. In general, the information from these
14 studies is limited by small group sizes, minimal characterization of some of the reported responses
15 (e.g., "congestion," "enlarged," "fatty liver"), insufficiently detailed reporting of study results, and
16 incomplete, if any, incidence data. In addition, Weatherby (1952], Anderson etal. (1964], and some
17 of the experiments reported by Coonetal. (1970) used only one ammonia concentration in addition
18 to the control, so no dose-response information is available from the majority of experimental
19 studies to inform the evidence for systemic effects of ammonia.
20 Ammonia is produced endogenously in all human and animal tissues during fetal and adult
21 life, and concentrations of free ammonia in physiological fluids are homeostatically regulated to
22 remain at low levels (Souba, 1987]. Thus, tissues are normally exposed to ammonia, and external
23 concentrations that do not alter homeostasis would not be expected to pose a hazard for systemic
24 effects. Experimental animal data suggest that ammonia exposures below 18 mg/m3 will not
25 increase blood ammonia levels (Manninenetal., 1988: Schaerdeletal., 1983]. See Appendix E,
26 Section E. 1, Metabolism, for a more detailed summary of the available literature that describes the
27 relationship between environmental ammonia concentrations and changes in ammonia
28 homeostasis.
29 Overall, the evidence in humans and animals indicates that ammonia exposure may be
30 associated with effects on organs distal from the portal of entry, but does not support the liver,
31 adrenal gland, kidney, spleen, or heart as sensitive targets of ammonia toxicity.
32
33 1.1.5. Carcinogenicity
34 No information is available regarding the carcinogenic effects of ammonia in humans
35 following oral or inhalation exposure. The carcinogenic potential of ammonia by the inhalation
36 route has not been assessed in animals, and animal carcinogenicity data by the oral route of
37 exposure are limited. Toth (1972] concluded that tumor incidence was not increased in Swiss mice
38 exposed for their lifetime (exact exposure duration not specified] to ammonium hydroxide in
39 drinking water at concentrations up to 0.3% (equivalent to 410 and 520 mg/kg-day in female and
40 male mice, respectively] or in C3H mice exposed to ammonium hydroxide in drinking water at a
This document is a draft for review purposes only and does not constitute Agency policy.
1-33 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 concentration of 0.1% (equivalentto 214 and 191 mg/kg-day in female and male mice,
2 respectively). With the exception of mammary gland tumors in female C3H mice, concurrent
3 control tumor incidence data were not reported and, therefore, comparison of tumor incidence in
4 exposed and control mice could not be performed. The general lack of concurrent control data
5 limits the ability to interpret the findings of this study.
6 The incidence of gastric cancer and the number of gastric tumors per tumor-bearing rat
7 were statistically significantly higher in rats exposed to 0.01% ammonia solution in drinking water
8 (equivalent to 10 mg/kg-day) for 24 weeks following pretreatment (for 24 weeks) with the
9 initiator, N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), compared with rats receiving only MNNG
10 and tap water (Tsujiietal.. 1992b). An ammonia-only exposure group was not included in this
11 study. In another study with the same study design, Tsujiietal. (1995) reported similar increases
12 in the incidence of gastric tumors in rats following exposure to MNNG and 10 mg/kg-day ammonia.
13 Additionally, the size and penetration to deeper tissue layers of the MNNG-initiated gastric tumors
14 were enhanced in the rats treated with ammonia (Tsujiietal., 1995). The investigators suggested
15 that ammonia administered in drinking water may act as a cancer promoter (Tsujiietal., 1995:
16 Tsujiietal.,1992b).
17 The evidence of carcinogenicity in experimental animals exposed to ammonia is
18 summarized in Table 1-8.
19
This document is a draft for review purposes only and does not constitute Agency policy.
1-34 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
Table 1-8. Evidence pertaining to cancer in animals
Study design and reference
Results
Carcinogenesis studies
Toth (1972)
Swiss mouse; 50/sex/group
0.1, 0.2, and 0.3% ammonium hydroxide in
drinking water for their lifetime [250, 440, and
520 mg/kg-d (males); 240, 370, and 410 mg/kg-d
(females)]3
Toth (1972)
C3H mouse; 40/sex/group
0.1% ammonium hydroxide in drinking water for
their lifetime [191 (males) and 214 mg/kg-d
(females)]"
Tumor incidence was not increasec
mice; however, concurrent control
were not reported.
Tumor incidence was not increased
mice; however, with the exception
tumors in female mice, concurrent
data were not reported.
Mammary gland adenocarcinoma:
in ammonia-exposed
tumor incidence data
in ammonia-exposed
of mammary gland
control tumor incidence
76, 60%
Initiation-promotion studies
Tsuiiietal. (1992b)
Sprague Dawley rat; male; 40/group
0 or 0.01% ammonia in drinking water (0 or
10 mg/kg-d)c for 24 wks; both groups pretreated
for 24 wks with the tumor initiator, MNNG; no
ammonia-only group
Tsuiiietal. (1995)
Sprague-Dawley rat; male; 43^44/group
0 or 0.01% ammonia in drinking water (0 or
10 mg/kg-d)c for 24 wks; both groups pretreated
for 24 wks with the tumor initiator, MNNG; no
ammonia-only group
Gastric tumor incidence: 31, 70*%
# of gastric tumors/tumor-bearing rat: 1.3, 2.1*
Gastric tumor incidence: 30, 66*%
Penetrated muscle layer or deeper:
Size (mm): 4.4, 5.3*
12, 22*%
1
2
3
4
5
6
7
aAmmonium hydroxide doses estimated based on reported average daily drinking water intakes of 9.2, 8.2, and
6.5 mL/day for males and 8.3, 6.5, and 4.8 mL/day for females in the 0.1, 0.2, and 0.3% groups, respectively,
and assumed average default body weights of 37.3 and 35.3 g for males and females, respectively (U.S. EPA,
1988).
bAmmonium hydroxide doses estimated based on reported average daily drinking water intakes of 7.9 and
8.4 mL/day for males and females, respectively, and assumed average default body weights of 37.3 and 35.3 g
for males and females, respectively (U.S. EPA, 1988).
cAmmonia doses estimated based on reported drinking water intake of 50 mL/day and assumed average default
body weight of 523 g for male Sprague-Dawley rats during chronic exposure (U.S. EPA, 1988).
*Statistically significantly different from the control (p < 0.05).
A limited number of genotoxicity studies are available for ammonia vapor, including one
study in exposed fertilizer factory workers in India that reported chromosomal aberrations and
sister chromatid exchanges in lymphocytes [Yadav and Kaushik, 1997], two studies that found no
evidence of DNA damage in rabbit gastric mucosal or epithelial cell lines [Suzuki etal.. 1998: Suzuki
etal.. 1997}. mutation assays in Salmonella typhimurium (not positive) and Escherichia coli
(positive) (Shimizu etal., 1985: Demerec et al., 1951], a micronucleus assay in mice (positive)
(Yadav and Kaushik, 1997], one positive and one negative study in Drosophila melanogaster
This document is a draft for review purposes only and does not constitute Agency policy.
1-35 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 [Auerbach and Robson, 1947: Lobasov and Smirnov, 1934], and a positive chromosomal aberration
2 test in chick fibroblast cells in vitro [Rosenfeld. 1932] (see Appendix E, Section E.4, Tables E-14 and
3 E-15]. The finding of chromosomal aberrations and sister chromatid exchanges in human
4 lymphocytes [Yadav and Kaushik, 1997] was difficult to interpret because of the small number of
5 samples and confounding in the worker population by smoking and alcohol consumption. In
6 addition, the levels of ammonia in the plant were low compared to other fertilizer plant studies,
7 raising questions about the study's exposure assessment. Positive findings in in vitro studies with
8 nonhuman cell lines were difficult to interpret because of the presence of a high degree of toxicity
9 [Demerec etal.. 1951: Lobasov and Smirnov. 1934] or inadequate reporting [Rosenfeld. 1932]. It is
10 noteworthy that four of the eight available genotoxicity studies were published between 1932 and
11 1951. In two of the more recent studies, ammonia exposure did not induce DNA damage in rabbit
12 gastric mucosal or epithelial cell lines in vitro [Suzuki etal.. 1998: Suzuki etal.. 1997]. Overall, the
13 available genotoxicity literature is inadequate to characterize the genotoxic potential of ammonia.
14
15 1.2. SUMMARY AND EVALUATION
16 1.2.1. Weight of Evidence for Effects Other than Cancer
17 The respiratory system is the primary and most sensitive target of inhaled ammonia toxicity
18 in humans and experimental animals. Evidence for respiratory system toxicity in humans comes
19 from cross-sectional occupational studies in industrial settings that reported changes in lung
20 function and an increased prevalence of respiratory symptoms. The findings of respiratory effects
21 in workers exposed to ammonia as a disinfectant or cleaning product (primarily studies of asthma
22 or asthma symptoms], studies of livestock farmers (i.e., lung function studies], controlled exposures
23 in volunteers, and case reports of injury following acute exposure provide additional evidence that
24 the respiratory system is a target of inhaled ammonia. Short-term and subchronic animal studies
25 show respiratory effects in several animal species across different dose regimens. Thus, the weight
26 of evidence of observed respiratory effects observed across multiple human and animal studies
27 identifies respiratory system effects as a hazard from ammonia exposure.
28 Evidence for an association between inhaled ammonia exposure and effects on other organ
29 systems distal from the portal of entry, including the immune system, liver, adrenal gland, kidney,
30 spleen, and heart, is less compelling than for the respiratory system. The two epidemiological
31 studies that addressed immunological function are confounded by exposures to a number of other
32 respirable agents that have been demonstrated to be immunostimulatory and provide little support
33 for ammonia immunotoxicity. Animal studies provide consistent evidence of elevated bacterial
34 growth following ammonia exposure. It is unclear, however, whether elevated bacterial
35 colonization is the result of suppressed immunity or damage to the barrier provided by the mucosal
36 epithelium of the respiratory tract Overall, the weight of evidence does not support the immune
37 system as a target of ammonia toxicity. Findings from animal studies indicate that ammonia
38 exposure may be associated with effects in the liver, adrenal gland, kidney, spleen, and heart;
This document is a draft for review purposes only and does not constitute Agency policy.
1-36 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 however, the weight of evidence indicates that these organs are not sensitive targets of ammonia
2 toxicity.
3 A limited experimental toxicity database indicates that oral exposure to ammonia may be
4 associated with effects on the stomach mucosa. Increased epithelial cell migration in the antral
5 gastric mucosa leading to a statistically significant decrease in mucosal thickness was reported in
6 male Sprague-Dawley rats exposed to ammonia in drinking water for durations up to 8 weeks
7 [Tsujiietal., 1993: Kawano etal., 1991]. Similarly, decreases in the height and labeling index of
8 gastric mucosa glands were reported in Donryu rats exposed to ammonia in drinking water for up
9 to 24 weeks [Hataetal., 1994]. The gastric mucosal effects observed in rats were reported to
10 resemble mucosal changes in human atrophic gastritis [Tsujiietal.. 1993: Kawano etal.. 1991]:
11 however, the investigators also reported an absence of microscopic lesions, gastritis, or ulceration
12 in the stomach of these rats. Evidence that oral exposure to ammonia is associated with
13 gastrointestinal effects in humans is limited to case reports of individuals suffering from
14 gastrointestinal effects (e.g., stomach ache, nausea, diarrhea, distress, and burns along the digestive
15 tract] from intentionally or accidentally ingesting household cleaning solutions containing
16 ammonia or biting into capsules of ammonia smelling salts. Mechanistic studies in rodent models
17 support the biological plausibility that ammonia exposure may be associated with gastric effects.
18 Given the weight of evidence from human, animal, and mechanistic studies, gastric effects may be a
19 hazard from ammonia exposure.
20 Studies of the potential reproductive or developmental toxicity of ammonia in humans are
21 not available. Reproductive effects were not associated with inhaled ammonia in the only animal
22 study that examined the reproductive effects of ammonia (i.e., a limited-design inhalation study in
23 the pig]. Further, ammonia is produced endogenously in human and animal tissues during fetal and
24 adult life, and concentrations of free ammonia in physiological fluids are homeostatically regulated
25 to remain at low levels (Souba, 1987]. Thus, exposures to ammonia at levels that do not alter
26 homeostasis (i.e., that do not alter normal blood or tissue ammonia levels] would not be expected to
27 pose a hazard for systemic effects, including effects on the developing fetus or reproductive tissues.
28
29 1.2.2. Weight of Evidence for Carcinogenicity
30 The available information on carcinogenicity following exposure to ammonia is limited to
31 oral animal studies. There was inadequate reporting in studies in Swiss or C3H mice administered
32 ammonium hydroxide in drinking water for a lifetime (Toth, 1972]. There is limited evidence that
33 ammonia administered in drinking water may act as a cancer promoter (Tsujii etal.. 1995: Tsujii et
34 al., 1992b]. The genotoxic potential cannot be characterized based on the available genotoxicity
35 information. Thus, under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a], there is
36 "inadequate information to assess carcinogenic potential" of ammonia.
37
38 1.2.3. Susceptible Populations and Lifestages
39 Studies of the toxicity of ammonia in children or young animals compared to other
40 lifestages that would support an evaluation of childhood susceptibility have not been conducted.
This document is a draft for review purposes only and does not constitute Agency policy.
1-37 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Hyperammonemia is a condition of elevated levels of circulating ammonia that can occur in
2 individuals with severe diseases of the liver or kidney, organs that biotransform and excrete
3 ammonia, or with hereditary urea cycle disorders [Cordoba etal.. 1998: Schubiger etal.. 1991:
4 Gilbert, 1988: Teffers etal., 1988: Souba, 1987]. The elevated ammonia levels that accompany
5 human diseases such as acute liver or renal failure can predispose an individual to encephalopathy
6 due to the ability of ammonia to cross the blood-brain barrier; these effects are especially marked
7 in newborn infants [Minanaetal., 1995: Souba, 1987]. Thus, individuals with disease conditions
8 that lead to hyperammonemia may be more susceptible to the effects of ammonia from external
9 sources, but there are no studies that specifically support this hypothesized susceptibility.
10 Because the respiratory system is a target of ammonia toxicity, individuals with respiratory
11 disease (e.g., asthmatics] might be expected to be a susceptible population. Controlled human
12 studies that examined both healthy volunteers and volunteers with asthma [Petrovaetal.. 2008:
13 Sigurdarson et al., 2004] did not demonstrate greater respiratory sensitivity in asthmatics than
14 healthy volunteers after acute exposure to ammonia. Under longer-term exposure conditions,
15 however, as seen among livestock farmers, one study observed associations between ammonia
16 exposure and decreased lung function among workers with chronic respiratory symptoms, but not
17 among the asymptomatic workers [Preller etal.. 1995]. Additional research focusing on the
18 question of variability in response to ammonia exposure is needed.
19
This document is a draft for review purposes only and does not constitute Agency policy.
1-38 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
2. DOSE-RESPONSE ANALYSIS
5 2.1. ORAL REFERENCE DOSE FOR EFFECTS OTHER THAN CANCER
6 The RfD (expressed in units of mg/kg-day) is defined as an estimate (with uncertainty
7 spanning perhaps an order of magnitude) of a daily oral exposure to the human population
8 (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects
9 during a lifetime. It can be derived from a no-observed-adverse-effect level (NOAEL), lowest-
10 observed-adverse-effect level (LOAEL), or the 95% lower bound on the benchmark dose (BMDL),
11 with uncertainty factors (UFs) generally applied to these points of departure (PODs) to reflect
12 limitations of the data used.
13 The available human and animal data are inadequate to derive an oral RfD for ammonia.
14 Human data involving oral exposure to ammonia are limited to case reports of gastrointestinal
15 effects following intentional or accidental ingestion of household cleaning solutions containing
16 ammonia or ammonia inhalant capsules. Case reports can indicate the nature of acute effects of
17 ammonia exposure and thus inform hazard identification. Because of short exposure durations and
18 incomplete or missing quantitative exposure information, data from case reports are inadequate for
19 dose-response analysis and subsequent derivation of a chronic reference value.
20 The experimental animal database for ammonia lacks standard toxicity studies that
21 systematically evaluate a range of tissues/organs and endpoints. Repeat-exposure animal studies
22 of the noncancer effects of ingested ammonia are limited to three studies designed to investigate
23 the mechanisms by which ammonia can induce effects on rat gastric mucosa (Hataetal., 1994:
24 Tsujiietal.. 1993: Kawano etal.. 1991). While these studies provide consistent evidence of changes
25 in the gastric mucosa associated with exposure to ammonia in drinking water (see Section 1.1.2),
26 the investigators reported no evidence of microscopic lesions, gastritis, or ulceration in the
27 stomachs of these rats. In addition, the gastrointestinal tract has not been identified as a target of
28 ammonia toxicity in chronic toxicity studies of ammonium compounds, including ammonium
29 chloride and sulfate (see Section 1.1.2).
30 Given the limited amount of toxicity testing that has been conducted on ingested ammonia
31 and questions concerning the adversity of the observed gastric mucosal findings in rats, the
32 available oral database for ammonia was considered insufficient to adequately characterize toxicity
33 outcomes and dose-response relationships. Accordingly, an RfD for ammonia was not derived.
34
35 Previous IRIS Assessment
36 No RfD was derived in the previous IRIS assessment for ammonia.
37
This document is a draft for review purposes only and does not constitute Agency policy.
2-1 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
i 2.2. INHALATION REFERENCE CONCENTRATION FOR EFFECTS OTHER
2 THAN CANCER
3 The RfC (expressed in units of mg/m3) is defined as an estimate (with uncertainty spanning
4 perhaps an order of magnitude) of a continuous inhalation exposure to the human population
5 (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects
6 during a lifetime. It can be derived from a NOAEL, LOAEL, or the 95% lower bound on the
7 benchmark concentration (BMCL), with UFs generally applied to these PODs to reflect limitations of
8 the data used.
9
10 2.2.1. Identification of Studies and Effects for Dose-Response Analysis
11 As discussed in Section 1.2, the respiratory system is the primary and most sensitive target
12 of inhaled ammonia in humans and experimental animals, and respiratory effects have been
13 identified as a hazard following inhalation exposure to ammonia. The experimental toxicology
14 literature for ammonia provides evidence that inhaled ammonia may be associated with toxicity to
15 target organs other than the respiratory system, including the liver, adrenal gland, kidney, spleen,
16 heart, and immune system. Effects in these other (nonrespiratory) target organs were not
17 considered as the basis for RfC derivation because the evidence for these associations is weak
18 relative to that for respiratory effects.
19 Respiratory effects, characterized as increased prevalence of respiratory symptoms or
20 decreased lung function, have been observed in worker populations exposed to ammonia
21 concentrations >18.5 mg/m3 (Rahman etal.. 2007: Alietal.. 2001: Ballaletal.. 1998). Decrements
22 in lung function parameters and increased prevalence of respiratory symptoms such as wheezing,
23 chest tightness, and cough/phlegm, have been identified as adverse respiratory health effects by
24 the American Thoracic Society (ATS. 2000) and are similarly noted as adverse in the EPA's Methods
25 for Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry (U.S.
26 EPA. 1994). Respiratory effects have also been observed in animals, but at ammonia
27 concentrations higher than those associated with respiratory effects in humans and in studies
28 involving exposure durations (up to 114 days) shorter than those in occupational studies.
29 In general human data are preferred over animal data for deriving reference values
30 because these data are more relevant for assessing human health effects than animal studies and
31 avoid the uncertainty associated with interspecies extrapolation when animal data serve as the
32 basis for the RfC. In the case of ammonia, the available occupational studies provide adequate data
33 for the quantitative analysis of health outcomes considered relevant to potential general population
34 exposures. In addition, ammonia concentrations associated with respiratory effects in human
35 studies were generally lower than effect levels identified in animal studies (Section 1.1.1).
36 Therefore, data on respiratory effects in humans were used for the derivation of the RfC and
37 respiratory effects in animals were not further considered.
38 Of the available human data, associations between ammonia exposure and respiratory
39 effects have been examined in epidemiology studies of industrial worker populations (Table 1-1),
40 workers using ammonia as a cleaning product (Table 1-2), and livestock farmers. Studies of
This document is a draft for review purposes only and does not constitute Agency policy.
2-2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 workers using ammonia as a cleaning product provide evidence of an association between
2 ammonia exposure and increased risk of asthma; however, these studies did not measure ammonia
3 concentrations in workplace air and thus are not useful for dose-response analysis. Studies in
4 livestock farmers also support an association between ammonia exposure and decreased
5 pulmonary function; however, because of co-exposures to other agents in these studies (including
6 dust, endotoxin, mold, and disinfectant products) and the availability of studies with fewer co-
7 exposures, studies of livestock farmers were considered to be supportive of the association
8 between ammonia exposure and respiratory effects but were not carried forward for dose-
9 response analysis.
10 Of the available studies of ammonia exposure in industrial settings, four cross-sectional
11 epidemiology studies of industrial worker populations—three studies in urea fertilizer plants by
12 Rahman etal. (2007). Ballaletal. (1998). and Alietal. (2001). and a study in a soda ash plant by
13 Holness etal. (1989)—provide information useful for examining the relationship between chronic
14 ammonia exposure and increased prevalence of respiratory symptoms and/or decreased lung
15 function. Bhatand Ramaswamy (1993) evaluated lung function in ammonia plant workers, but did
16 not measure ammonia concentrations in workplace air. Therefore, this study was not considered
17 useful for RfC derivation.
18 In general, the four cross-sectional occupational studies provide a coherent set of estimated
19 NOAELs (i.e., workplace exposures up to 8.8 mg/m3) and effect levels, and are considered candidate
20 principal studies for RfC derivation. Rahman etal. (2007) observed an increased prevalence of
21 respiratory symptoms and decreased lung function in fertilizer plant workers exposed to a mean
22 ammonia concentration of 18.5 mg/m3, but not in workers in a second plant exposed to a mean
23 ammonia concentration of 4.9 mg/m3. Ballal etal. (1998) observed an increased prevalence of
24 respiratory symptoms among workers in one factory (Factory A) with ammonia exposures ranging
25 from 2-27.1 mg/m3,5 but no increase in symptoms in another factory (Factory B) with exposures
26 ranging from 0.02-7 mg/m3. A companion study by Alietal. (2001) observed decreased lung
27 function among workers in the factory with the higher ammonia exposures (Factory A); the factory
28 with the lower ammonia exposures, also studied by Ballal etal. (1998), was not included in this
29 companion study by Ali etal. (2001). Holness etal. (1989) found no differences in the prevalence
30 of respiratory symptoms or lung function between workers (mean exposure 6.5 mg/m3) and the
31 control group, and also no differences in respiratory symptoms or lung function when workers
32 were stratified by ammonia exposure level (lowest exposure group, <4.4 mg/m3; middle exposure
33 group, 4.4-8.8 mg/m3; highest exposure group, >8.8 mg/m3).
34 The NOAEL of 8.8 mg/m3 from the Holness etal. (1989) study represents the low end of the
35 high-exposure group (defined as those exposed to >8.8 mg/m3) from this study. The authors state
36 that 3 of the 12 workers in the high-exposure group were exposed to concentrations >17.7 mg/m3;
37 therefore, the majority of workers in the high-exposure group (9 of 12) would have been exposed to
5This concentration range does not include exposures in the urea store (number of employees = 6; range of
ammonia concentrations = 90-130.4 mg/m3) because employees in this area were required to wear full
protective clothing, thus minimizing potential exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
2-3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 ammonia concentrations in the range of 8.8-17.7 mg/m3. In the absence of more detailed exposure
2 information, the low-end of the range was considered a reasonable estimate of the NOAEL from the
3 Holnessetal. [1989] study.
4 Of the four candidate principal studies, higher confidence is associated with the exposure
5 measures from Holness etal. [1989]. Both Holness et al. [1989] and Rahman et al. [2007] collected
6 personal air samples, but confidence in the analytical method used by Holness etal. [1989] is
7 higher than that used by Rahman etal. [2007]. Rahman etal. [2007] used two analytical methods
8 for measuring ammonia concentrations in workplace air (i.e., Drager PAC III and Drager tube];
9 concentrations measured by the two methods differed by four- to fivefold, indicating some
10 uncertainty across the two measurement methods, although ammonia concentrations measured by
11 the two methods were strongly correlated (correlation coefficient of 0.8]. In contrast, the Holness
12 etal. [1989] study used an established analytical method for measuring exposure to ammonia
13 recommended by the National Institute for Occupational Safety and Health (NIOSH] that involved
14 the collection of air samples on acid-treated silica gel absorption tubes. Ballal etal. [1998] used
15 area monitors rather than personal air sampling methods; the latter method provides a better
16 estimate of an individual's exposure. Both Holnessetal. [1989] and Rahman etal. [2007] examined
17 both respiratory symptoms and lung function, which provides stronger evidence of respiratory
18 effects than symptom data alone. Ballal etal. [1998] evaluated only respiratory symptoms. Ali et
19 al. [2001]. the companion study to Ballal etal. [1998]. examined pulmonary function; however,
20 because Ali etal. [2001] evaluated only workers in the higher exposure setting, the data cannot be
21 used to estimate a NOAEL.
22 Considerations in selecting the principal study for RfC derivation include the higher
23 confidence placed in the measures of ammonia exposure in Holness etal. [1989] as compared to
24 the other candidate studies, evaluation of both respiratory symptoms and lung function parameters
25 in the Holness etal. [1989] study, and the fact that the estimate of the NOAEL for respiratory effects
26 of 8.8 mg/m3 from Holnessetal. [1989] was the highest of the NOAELs estimated from the
27 candidate principal studies. The Holness etal. [1989] study does not demonstrate a relationship
28 between ammonia exposure and respiratory effects probably because of the relatively low levels of
29 ammonia in the workplace that reflect the controlled nature of the operations at the plant. The
30 Holness etal. [1989] study is identified as the principal study for derivation of the RfC, but only
31 with support from the collection of occupational epidemiology studies that includes studies with
32 higher workplace ammonia concentrations.
33 In summary, the occupational study of ammonia exposure in workers in a soda ash plant by
34 Holness etal. (1989) was identified as the principal study for RfC derivation, with support
35 from Rahman etal. f20071. Ballal etal. f 19981. and Ali etal. fZOOll. and respiratory effects
36 were identified as the critical effect.
37
38 2.2.2. Methods of Analysis
39 A NOAEL of 8.8 mg/m3, identified from the Holness etal. (1989) study, was used as
40 the point of departure (POD) for RfC derivation.
This document is a draft for review purposes only and does not constitute Agency policy.
2-4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Because the RfC assumes continuous human exposure over a lifetime, the POD was adjusted
2 to account for the noncontinuous exposure associated with occupational exposure (i.e., 8-hour
3 workday and 5-day workweek). The duration-adjusted POD was calculated as follows:
4
5 NOAELADj = NOAEL x VEho/VEh x 5 days/7 days
6 = 8.8 mg/m3 x 10 m3/20 m3 x 5 days/7 days
7 = 3.1 mg/m3
8 Where:
9 VEho = human occupational default minute volume (10 m3 breathed during the 8-hour
10 workday, corresponding to a light to moderate activity level) (U.S. EPA. 2011a)
11 VEh = human ambient default minute volume (20 m3 breathed during the entire day).
12
13 2.2.3. Derivation of the Reference Concentration
14 Consistent with EPA's A Review of the Reference Dose and Reference Concentration Processes
15 (U.S. EPA. 2002: Section 4.4.5). also described in the Preamble, five possible areas of uncertainty
16 and variability were considered when deriving the RfC. A composite UF of 10 was applied to the
17 selected duration-adjusted POD of 3.1 mg/m3 to derive the RfC of 0.3 mg/m3. An explanation of the
18 five possible areas of uncertainty and variability follows:
19
20 • An intraspecies uncertainty factor, UFn, of 10 was applied to account for potentially
21 susceptible individuals in the absence of data evaluating variability of response to inhaled
22 ammonia in the human population;
23
24 • An interspecies uncertainty factor, UFA, of 1 was applied to account for uncertainty in
25 extrapolating from laboratory animals to humans because the POD was based on human
26 data from an occupational study;
27
28 • A subchronic to chronic uncertainty factor, UFS, of 1 was applied because the occupational
29 exposure period in the principal study (Holness etal.. 1989). defined as the mean number of
30 years at the present job for exposed workers, of approximately 12 years was considered to
31 be of chronic duration;
32
33 • An uncertainty factor for extrapolation from a LOAEL to a NOAEL, UFi, of 1 was applied
34 because a NOAEL was used as the POD; and
35
36 • A database uncertainty factor, UFo, of 1 was applied to account for deficiencies in the
37 database. The ammonia inhalation database consists of epidemiological studies and
38 experimental animal studies. The epidemiological studies include industrial worker
39 populations, populations exposed to ammonia through the use of cleaning products, studies
40 in livestock farmers exposed to inhaled ammonia and other airborne agents, controlled
41 exposure studies involving volunteers exposed to ammonia vapors for short periods of time,
42 and a large number of case reports of acute exposure to high ammonia concentrations (e.g.,
43 accidental spills/releases) that examined irritation effects, respiratory symptoms, and
44 effects on lung function. Studies of the toxicity of inhaled ammonia in experimental animals
45 include subchronic studies in a number of species, including rats, guinea pigs, and pigs, that
46 examined respiratory and other systemic effects of ammonia, several immunotoxicity
This document is a draft for review purposes only and does not constitute Agency policy.
2-5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 studies, and one limited reproductive toxicity study in young female pigs. (See Chapter 1
2 for more details regarding available studies.) The database lacks developmental and
3 multigeneration reproductive toxicity studies.
4
5 As noted in EPA's A Review of the Reference Dose and Reference Concentration Processes [U.S.
6 EPA. 2002]. "the size of the database factor to be applied will depend on other information
7 in the database and on how much impact the missing data may have on determining the
8 toxicity of a chemical and, consequently, the POD." While the database lacks
9 multigeneration reproductive and developmental toxicity studies, these studies would not
10 be expected to impact the determination of ammonia toxicity at the POD. Therefore, a
11 database UF to account for the lack of these studies is not considered necessary. This
12 determination was based on the observation that ammonia is endogenously produced and
13 homeostatically regulated in humans and animals during fetal and adult life. In vivo studies
14 in several animal species and in vitro studies of human placenta demonstrate that ammonia
15 is produced within the uteroplacenta and released into the fetal and maternal circulations
16 [Tozwiketal.. 2005: Tozwik etal.. 1999: Bell etal.. 1989: Johnson etal.. 1986: Hauguel etal..
17 1983: Meschiaetal.. 1980: Remesar etal.. 1980: Holzmanetal.. 1979: Holzmanetal.. 1977:
18 Rubaltelli and Formentin, 1968: Luschinsky, 19511. Ammonia concentrations in human
19 umbilical vein and artery blood (at term) of healthy individuals have been shown to be
20 higher than concentrations in maternal blood (i.e., 1.0-1.4 [ig/mL in umbilical arterial and
21 venous blood compared to 0.5 [ig/mL in the mothers' venous blood) (Jozwiketal.. 2005).
22 Human fetal umbilical blood levels of ammonia at birth were not influenced by gestational
23 age based on deliveries ranging from gestation week 25 to 43 (DeSanto etal.. 1993). This
24 evidence provides some assurance that endogenous ammonia concentrations in the fetus
25 are similar to other lifestages, and that baseline ammonia concentrations would not be
26 associated with developmental toxicity. Additionally, evidence in animals (Manninen etal..
27 1988: Schaerdel etal., 1983) suggests that exposure to ammonia at concentrations up to
28 18 mg/m3 does not alter blood ammonia levels (see Appendix E, Section E. 1, for a more
29 detailed discussion of ammonia distribution and elimination). Accordingly, exposure at the
30 duration-adjusted POD (3.1 mg/m3) would not be expected to alter ammonia homeostasis
31 nor result in measureable increases in blood ammonia concentrations. Thus, exposure to
32 ammonia at the POD for the RfC would not be expected to result in systemic toxicity,
33 including reproductive or developmental toxicity.
34
35 The RfC for ammonia6 was calculated as follows:
36
37 RfC = NOAELADj + UF
38 =3.1 mg/m3 -H 10
39 = 0.31 mg/m3 or 0.3 mg/m3 (rounded to one significant figure)
40
41 2.2.4. Uncertainties in the Derivation of the Reference Concentration
42 As presented earlier in this section and in the Preamble, EPA standard practices and RfC
43 guidance (U.S. EPA. 2002.1995.1994) were followed in applying an UF approach to a POD (from a
44 NOAEL) to derive the RfC. Specific uncertainties were accounted for by the application of UFs (i.e.,
6Due to uncertainty concerning the possible influence of anions on the toxicity of ammonium, information on
ammonium salts was not used to characterize the effects for ammonia and ammonium hydroxide. Therefore,
the RfC derived in this assessment is applicable to ammonia and ammonium hydroxide, but not ammonium
salts.
This document is a draft for review purposes only and does not constitute Agency policy.
2-6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 in the case of the ammonia RfC, a factor to address the absence of data to evaluate the variability in
2 response to inhaled ammonia in the human population). The following discussion identifies
3 additional uncertainties associated with the quantification of the RfC for ammonia.
4
5 Use of a NOAEL as a POD
6 Data sets that support benchmark dose modeling are generally preferred for reference
7 value derivation because the shape of the dose-response curve can be taken into account in
8 establishing the POD. For the ammonia RfC, no decreases in lung function or increases in the
9 prevalence of respiratory symptoms were observed in the worker population studied by Holness et
10 al. [1989]. i.e., the principal study used to derive the RfC, and as such, the data from this study did
11 not support dose-response modeling. Rather, a NOAEL from the Holness etal. [1989] study was
12 used to estimate the POD. The availability of dose-response data from a study of ammonia,
13 especially in humans, would increase the confidence in the estimation of the POD.
14
15 Endogenous Ammonia
16 Ammonia, which is produced endogenously, has been detected in breath exhaled from the
17 nose and trachea of humans (range: 0.0092-0.1 mg/m3] [Schmidt etal.. 2013: Smith etal.. 2008:
18 Larson et al., 1977]. Higher and more variable ammonia concentrations are reported in human
19 breath exhaled from the mouth or oral cavity, with the majority of ammonia concentrations from
20 these sources ranging from 0.085 to 2.1 mg/m3 [Schmidt etal.. 2013: Smith etal.. 2008: Spanel et
21 al.. 2007a. b: Turner etal.. 2006: Diskinetal.. 2003: Smith etal.. 1999: Norwood etal.. 1992: Larson
22 etal., 1977]. Ammonia in exhaled breath from the mouth or oral cavity is largely attributed to the
23 production of ammonia via bacterial degradation of food protein in the oral cavity or
24 gastrointestinal tract (Turner etal.. 2006: Smith etal.. 1999: Vollmuth and Schlesinger. 19841 and
25 can be influenced by factors such as diet, oral hygiene, and age. In contrast, ammonia
26 concentrations measured in breath exhaled from the nose and trachea are lower (range: 0.0092-0.1
27 mg/m3] [Schmidt etal., 2013: Smith etal., 2008: Larson etal., 1977] and appear to better represent
28 levels at the alveolar interface of the lung or in the tracheo-bronchial region and are thought to be
29 more relevant to understanding systemic levels of ammonia than ammonia in breath exhaled from
30 the mouth [Schmidt etal.. 2013: Smith etal.. 2008] (Appendix E, Section E.I and Table E-l].
31 It is important to recognize that ammonia in ambient air is the source of some of the
32 ammonia in exhaled breath. Studies of ammonia in exhaled breath (Appendix E, Table E-l] were
33 conducted in environments with measureable levels of ambient (exogenous] ammonia rather than
34 in ammonia-free environments, and it has been established that concentrations of certain trace
35 compounds in exhaled breath are correlated with their ambient concentrations [Spanel etal.,
36 2013]. Spanel etal. [2013] found that 70% (± 13%] of inhaled ammonia is retained in exhaled
37 breath. It is likely that ammonia concentrations in breath exhaled from the nose would be lower if
38 the inspired air were free of ammonia. Therefore, levels of ammonia in exhaled breath reported in
39 the literature would need to be adjusted if they are to be used as a measure of systemic ammonia.
This document is a draft for review purposes only and does not constitute Agency policy.
2-7 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Ammonia concentrations measured in breath exhaled from the nose and trachea,
2 considered to be more representative of systemic levels of ammonia than breath exhaled from the
3 mouth, are lower than the ammonia RfC of 0.3 mg/m3 by a factor of threefold or more. The range of
4 ammonia breath concentrations measured in samples collected from the mouth (0.085 to
5 2.1 mg/m3), i.e., concentrations that are largely influenced by such factors as ammonia production
6 via bacterial degradation of food protein, includes the value of the ammonia RfC. Ammonia exhaled
7 by an individual, whether through the nose or mouth, is rapidly diluted in the larger volume of
8 ambient air and would not contribute significantly to overall ammonia exposure. Further, such
9 endogenous exposures existed in the occupational epidemiology studies that served as the basis for
10 the ammonia RfC.
11
12 2.2.5. Confidence Statement
13 A confidence level of high, medium, or low is assigned to the study used to derive the RfC,
14 the overall database, and the RfC itself, as described in Section 4.3.9.2 of EPA's Methods for
15 Derivation of Inhalation Reference Concentrations and Application of Inhalation Dosimetry [U.S. EPA,
16 1994]. Confidence in the principal study [Holness etal., 1989] is medium. The design, conduct, and
17 reporting of this occupational exposure study were adequate, but the study was limited by a small
18 sample size and by the fact that workplace ammonia concentrations to which the study population
19 was exposed were below those associated with ammonia-related effects (i.e., only a NOAEL was
20 identified]. However, the results from the principal study are supported by the results from other
21 cross-sectional studies of workers in industrial settings, studies of workers using ammonia as a
22 cleaning product, studies of livestock farmers, multiple studies of acute ammonia exposure in
23 volunteers, and the available inhalation data from animals.
24 Confidence in the database is medium. The inhalation ammonia database includes one
25 limited study of reproductive and developmental toxicity in pigs that did not examine a complete
26 set of reproductive or developmental endpoints. Normally, confidence in a database lacking these
27 types of studies is considered to be lower due to the uncertainty surrounding the use of any one or
28 several studies to adequately address all potential endpoints following chemical exposure at
29 various critical lifestages. Unless a comprehensive array of endpoints is addressed by the database,
30 there is uncertainty as to whether the critical effect chosen for the RfC derivation is the most
31 sensitive or appropriate. However, reproductive, developmental, and other systemic effects are not
32 expected at the RfC because it is well documented that ammonia is endogenously produced in
33 humans and animals, ammonia concentrations in blood are homeostatically regulated to remain at
34 low levels, and ammonia concentrations in air at the POD are not expected to alter homeostasis.
35 Thus, confidence in the database, in the absence of these types of studies, is medium.
36 Reflecting medium confidence in the principal study and medium confidence in the
37 database, the overall confidence in the RfC is medium.
38
This document is a draft for review purposes only and does not constitute Agency policy.
2-8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 2.2.6. Previous IRIS Assessment
2 The previous IRIS assessment for ammonia (posted to the database in 1991) presented an
3 RfC of 0.1 mg/m3 based on co-principal studies—the occupational exposure study of workers in a
4 soda ash plant by Holness etal. [1989] and the subchronic study by Brodersonetal. [1976] that
5 examined the effects of ammonia exposure in F344 rats inoculated on day 7 of the study with the
6 bacterium M. pulmonis. The NOAEL of 6.4 mg/m3 (estimated as the mean concentration of the
7 entire exposed group] from the Holness etal. [1989] study (duration adjusted: NOAELADj =
8 2.3 mg/m3] was used as the POD.7
9 The previous RfC was derived by dividing the exposure-adjusted POD of 2.3 mg/m3 (from a
10 NOAEL of 6.4 mg/m3] by a composite UFof 30: 10 to account for the protection of sensitive
11 individuals and 3 for database deficiencies to account for the lack of chronic data, the proximity of
12 the LOAEL from the subchronic inhalation study in the rat [Brodersonetal.. 1976] to the NOAEL,
13 and the lack of reproductive and developmental toxicity studies. A UFo of 3 (rather than 10] was
14 applied because studies in rats (Schaerdel etal.. 1983] showed no increase in blood ammonia levels
15 at an inhalation exposure up to 32 ppm (22.6 mg/m3] and only minimal increases at 300-
16 1,000 ppm (212-707 mg/m3], suggesting that no significant distribution is likely to occur at the
17 human equivalent concentration. In this document, a UFo of one was selected because a more
18 thorough investigation of the literature on ammonia homeostasis and literature published since
19 1991 on fetoplacental ammonia levels provides further support that exposure to ammonia at the
20 POD would not result in a measureable increase in blood ammonia, including fetal blood levels.
21
22 2.3. Cancer Risk Estimates
23 The carcinogenicity assessment provides information on the carcinogenic hazard potential
24 of the substance in question, and quantitative estimates of risk from oral and inhalation exposure
25 may be derived. Quantitative risk estimates may be derived from the application of a low-dose
26 extrapolation procedure. If derived, and unless otherwise stated, the oral slope factor is a plausible
27 upper bound on the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit
28 risk is a plausible upper bound on the estimate of risk per [J.g/m3 air breathed.
29 As discussed in Section 1.2, there is "inadequate information to assess carcinogenic
30 potential" of ammonia. Therefore, a quantitative cancer assessment was not conducted and cancer
31 risk estimates were not derived for ammonia.
32 The previous IRIS assessment of ammonia also did not include a carcinogenicity
33 assessment
34
7In this document, the lower bound of the high exposure category from the Holness et al. [1989] study
(8.8 mg/m3, adjusted for continuous exposure to 3.1 mg/m3) was identified as the POD because workers in
this high-exposure category, as well as those in the two lower-exposure categories, showed no statistically
significant increase in the prevalence of respiratory symptoms or decreases in lung function.
This document is a draft for review purposes only and does not constitute Agency policy.
2-9 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1
2
3 REFERENCES
4
5
6 Multiple references published in the same year by the same author(s) have been assigned a
7 letter (e.g., 1986a, 1986b) based on author(s), year, and title. Those same letters have been
8 retained for the appendices.
9
10 Ali. BA: Ahmed. HO: Ballal. SG: Albar. AA. [2001]. Pulmonary function of workers exposed to
11 ammonia: a study in the Eastern Province of Saudi Arabia. Int 1 Occup Environ Health 7: 19-
12 22,
13 Amshel, CE: Fealk, MH: Phillips, BT: Caruso, DM. [2000]. Anhydrous ammonia burns case report and
14 review of the literature [Review]. Burns 26: 493-497.
15 Anderson. DP: Beard. CW: Hanson. RP. [1964]. The adverse effects of ammonia on chickens
16 including resistance to infection with Newcastle disease virus. Avian Pis 8: 369-379.
17 Andreasen, M: Baekbo, P: Nielsen, IP. [2000]. Lack of effect of aerial ammonia on atrophic rhinitis
18 and pneumonia induced by Mycoplasma hyopneumoniae and toxigenic Pasteurella
19 multocida. 1 Vet Med B 47: 161-171.
20 Arif. AA: Delclos. GL. [2012]. Association between cleaning-related chemicals and work-related
21 asthma and asthma symptoms among healthcare professionals. Occup Environ Med 69: 35-
22 40. http://dx.doi.org/10.1136/oem.2011.064865
23 ATS [American Thoracic Society]. [2000]. What constitutes an adverse health effect of air pollution?
24 This official statement of the American Thoracic Society was adopted by the ATS Board of
25 Directors. Tuly 1999. Am 1 Respir Grit Care Med 161: 665-673.
26 http://dx.doi.0rg/10.1164/airccm.161.2.ats4-00
27 ATSDR [Agency for Toxic Substances and Disease Registry). [2004]. Toxicological profile for
28 ammonia [ATSDR Tox Profile]. Atlanta, GA: U.S. Department of Health and Human Services,
29 Public Health Service, http://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=ll&tid=2
30 Auerbach. C: Robson. TM. [1947]. Tests of chemical substances for mutagenic action. Proc Roy Soc
31 EdinbBBiol 62: 284-291.
32 Bacom. A: Yanosky. M. [20101. E-mails dated Tune 22. 2010. from Michael Yanosky. Drager Safety
33 Inc., Technical Support Detection Products to Amber Bacom, SRC, Inc., contractor to NCEA,
34 ORD. U.S. EPA. Available online
35 Ballal. SG: Ali. BA: Albar. AA: Ahmed. HO: al-Hasan. AY. [19981. Bronchial asthma in two chemical
36 fertilizer producing factories in eastern Saudi Arabia. Int 1 Tuberc Lung Pis 2: 330-335.
37 Bell. AW: Kennaugh. TM: Battaglia. FC: Meschia. G. [1989]. Uptake of amino acids and ammonia at
38 mid-gestation by the fetal lamb. 0 1 Exp Physiol 74: 635-643.
39 Bhat, MR: Ramaswamy, C. [1993]. Effect of ammonia, urea and diammonium phosphate [DAP] on
40 lung functions in fertilizer plant workers. Indian 1 Physiol Pharmacol 37: 221-224.
41 Broderson. JR.: Lindsey. JR.: Crawford. IE. [1976]. The role of environmental ammonia in respiratory
42 mvcoplasmosis of rats. Am 1 Pathol 85: 115-130.
This document is a draft for review purposes only and does not constitute Agency policy.
R-l DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 CDC [Centers for Disease Control and Prevention]. [2004]. The health consequences of smoking: A
2 report of the Surgeon General. Washington, DC: U.S. Department of Health and Human
3 Services, http://www.surgeongeneral.gov/library/smokingconsequences/
4 Choudat. D: Goehen. M: Korobaeff. M: Boulet. A: Dewitte. ID: Martin. MH. [19941. Respiratory
5 symptoms and bronchial reactivity among pig and dairy farmers. Scand 1 Work Environ
6 Health 20: 48-54.
7 Christesen. HB. [1995]. Prediction of complications following caustic ingestion in adults. Clin
8 Otolaryngol Allied Sci 20: 272-278.
9 Cole. TT: Cotes. IE: Tohnson. GR: Martin Hde. V: Reed. TW: Saunders. Ml. [1977]. Ventilation, cardiac
10 frequency and pattern of breathing during exercise in men exposed to 0-chlorobenzylidene
11 malononitrile [CS] and ammonia gas in low concentrations. 0 1 Exp Physiol 62: 341-351.
12 Coon, RA: Tones, RA: Tenkins, LT. Tr: Siegel, 1. [1970]. Animal inhalation studies on ammonia, ethylene
13 glycol, formaldehyde, dimethylamine, and ethanol. Toxicol Appl Pharmacol 16: 646-655.
14 http://dx.doi.org/10.1016/0041-008xr70190069-4
15 Cordoba, 1: Gottstein, 1: Blei, AT. [1998]. Chronic hyponatremia exacerbates ammonia-induced brain
16 edema in rats after portacaval anastomosis. 1 Hepatol 29: 589-594.
17 Cormier. Y: Israel-Assayag. E: Racine. G: Duchaine. C. [2000]. Farming practices and the respiratory
18 health risks of swine confinement buildings. Eur Respir 115: 560-565.
19 Crook. B: Robertson. IF: Glass. SA: Botheroyd. EM: Lacey. 1: Topping. MD. [19911. Airborne dust.
20 ammonia, microorganisms, and antigens in pig confinement houses and the respiratory
21 health of exposed farm workers. Am Ind Hyg Assoc 1 52: 271-279.
22 http://dx.doi.org/10.1080/15298669191364721
23 Curtis. SE: Anderson. CR: Simon. 1: Tensen. AH: Day. PL: Kelley. KW. [1975]. Effects of aerial
24 ammonia, hydrogen sulfide and swine-house dust on rate of gain and respiratory-tract
25 structure in swine. 1 Anim Sci 41: 735-739.
26 Dean, TA. [1985]. Lange's handbook of chemistry. New York, NY: McGraw-Hill.
27 Delclos, GL: Gimeno, D: Arif, AA: Benavides, FG: Zock, IP. [2009]. Occupational exposures and
28 asthma in health-care workers: Comparison of self-reports with a workplace-specific job
29 exposure matrix. Am 1 Epidemiol 169: 581-587. http://dx.doi.org/10.1093/aie/kwn387
30 Demerec. M: Bertani. G: Flint. 1. [1951]. A survey of chemicals for mutagenic action on E coli. Am Nat
31 85:119-135.
32 DeSanto. IT: Nagomi. W: Liechty. EA: Lemons. TA. [1993]. Blood ammonia concentration in cord
33 blood during pregnancy. Early Hum Dev 33: 1-8.
34 Diekman. MA: Scheldt. AB: Sutton. AL: Green. ML: Clapper. TA: Kelly. DT: Van Alstine. WG. [19931.
35 Growth and reproductive performance, during exposure to ammonia, of gilts afflicted with
36 pneumonia and atrophic rhinitis. Am 1 Vet Res 54: 2128-2131.
37 Diskin, AM: Spanel, P: Smith, D. [2003]. Time variation of ammonia, acetone, isoprene and ethanol
38 in breath: a quantitative SIFT-MS study over 30 days. Physiol Meas 24: 107-119.
39 Doig, PA: Willoughby, RA. [1971]. Response of swine to atmospheric ammonia and organic dust 1
40 Am VetMed Assoc 159: 1353-1361.
41 Done. SH: Chennells. DT: Gresham. AC: Williamson. S: Hunt. B: Taylor. LL: Bland. V: Tones. P:
42 Armstrong. D: White. RP: Demmers. TG: Teer. N: Wathes. CM. f20051. Clinical and
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 pathological responses of weaned pigs to atmospheric ammonia and dust. VetRec 157: 71-
2 80,
3 Donham, KT: Cumro, D: Reynolds, ST: Merchant, TA. [2000]. Dose-response relationships between
4 occupational aerosol exposures and cross-shift declines of lung function in poultry workers:
5 recommendations for exposure limits. T Occup Environ Med 42: 260-269.
6 Donham. KT: Reynolds. ST: Whitten. P: Merchant. TA: Burmeister. L: Popendorf. WT. (1995).
7 Respiratory dysfunction in swine production facility workers: dose-response relationships
8 of environmental exposures and pulmonary function. Am T Ind Med 27: 405-418.
9 Donnay, C: Denis, MA: Magis, R: Fevotte, T: Massin, N: Dumas, 0: Pin, I: Choudat, D: Kauffmann, F: Le
10 Moual. N. (2011). Under-estimation of self-reported occupational exposure by
11 questionnaire in hospital workers. Occup Environ Med 68: 611-617.
12 http://dx.doi.org/10.1136/oem.2010.061671
13 Douglas, RB: Coe, IE. (1987]. The relative sensitivity of the human eye and lung to irritant gases.
14 Ann Occup Hyg 31: 265-267.
15 Dumas. 0: Donnay. C: Heederik. DT: Hery. M: Choudat. D: Kauffmann. F: Le Moual. N. (20121.
16 Occupational exposure to cleaning products and asthma in hospital workers. Occup Environ
17 Med 69: 883-889. http://dx.doi.org/10.1136/oemed-2012-100826
18 Dworkin. MS: Patel. A: Fennell. M: Vollmer. M: Bailey. S: Bloom. T: Mudahar. K: Lucht. R. (2004). An
19 outbreak of ammonia poisoning from chicken tenders served in a school lunch. T Food Prot
20 67:1299-1302.
21 Eggeman. T. (2001). Ammonia. In Kirk-Othmer encyclopedia of chemical technology. New York. NY:
22 Tohn Wiley & Sons.
23 Fazekas, IG. (1939]. Die durch ammoniak hervorgerufene experimentelle
24 nebennierenhypertrophie. Endokrinologie 21: 315-337.
25 Ferguson, WS: Koch, WC: Webster, LB: Gould, JR. (1977]. Human physiological response and
26 adaption to ammonia. T Occup Med 19: 319-326.
27 Gaafar, H: Girgis, R: Hussein, M: el-Nemr, F. (1992]. The effect of ammonia on the respiratory nasal
28 mucosa of mice. A histological and histochemical study. Acta Otolaryngol 112: 339-342.
29 Gilbert. GT. (1988). Acute ammonia intoxication 37 years after ureterosigmoidostomy. South Med T
30 81: 1443-1445.
31 Giroux, M: Ferrieres, T. (1998]. Serum nitrates and creatinine in workers exposed to atmospheric
32 nitrogen oxides and ammonia. Sci Total Environ 217: 265-269.
33 http://dx.doi.org/10.1016/S0048-9697r98100190-9
34 Gustin, P: Urbain, B: Prouvost, IF: Ansay, M. (1994]. Effects of atmospheric ammonia on pulmonary
35 hemodynamics and vascular permeability in pigs: interaction with endotoxins. Toxicol Appl
36 Pharmacol 125: 17-26. http://dx.doi.org/10.1006/taap.1994.1044
37 Guyatt. GH: Oxman. AD: Kunz. R: Vist. GE: Falck-Ytter. Y: Schiinemann. HI. (2008a1. GRADE: What is
38 "quality of evidence" and why is it important to clinicians? [Review]. BMT 336: 995-998.
39 http://dx.doi.org/10.1136/bmi.39490.551019.BE
40 Guyatt. GH: Oxman. AD: Vist. GE: Kunz. R: Falck-Ytter. Y: Alonso-Coello. P: Schiinemann. HI. (2008b).
41 GRADE: An emerging consensus on rating quality of evidence and strength of
42 recommendations. BMT 336: 924-926. http://dx.doi.org/10.1136/bmi.39489.470347.AD
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Hamid, HA: El-Gazzar, RM. [1996]. Effect of occupational exposure to ammonia on enzymatic
2 activities of catalase and mono amine oxidase. T Egypt Public Health Assoc 71: 465-475.
3 Hamilton. TD: Roe. TM: Hayes. CM: Tones. P: Pearson. GR: Webster. AT. [1999]. Contributory and
4 exacerbating roles of gaseous ammonia and organic dust in the etiology of atrophic rhinitis.
5 Clin Diagn Lab Immunol 6: 199-203.
6 Hamilton. TD: Roe. TM: Hayes. CM: Webster. AT. [1998]. Effects of ammonia inhalation and acetic
7 acid pretreatment on colonization kinetics of toxigenic Pasteurella multocida within upper
8 respiratory tracts of swine. T Clin Microbiol 36: 1260-1265.
9 Hata, M: Yamazaki, Y: Ueda, T: Kato, T: Kohli, Y: Fujiki, N. [1994]. Influence of ammonia solution on
10 gastric mucosa and acetic acid induced ulcer in rats. Eur T Histochem 38: 41-52.
11 Hauguel. S: Challier. 1C: Cedard. L: Olive. G. [1983]. Metabolism of the human placenta perfused in
12 vitro: glucose transfer and utilization, 02 consumption, lactate and ammonia production.
13 Pediatr Res 17: 729-732. http://dx.doi.org/10.1203/00006450-198309000-00009
14 Heederik, D: van Zwieten, R: Brouwer, R. [1990]. Across-shift lung function changes among pig
15 farmers. Am T Ind Med 17: 57-58.
16 HEW [U.S. Department of Health. Education and Welfare]. [1964]. Smoking and health: Report of
17 the advisory committee to the surgeon general of the public health service. Washington. DC:
18 U.S. Department of Health. Education, and Welfare.
19 http://profiles.nlm.nih.gov/ps/retrieve/ResourceMetadata/NNBBMQ
20 Hill, AB. [1965]. The environment and disease: Association or causation? Proc R Soc Med 58: 295-
21 300.
22 Holness. PL: Purdham. IT: Nethercott. JR. [1989]. Acute and chronic respiratory effects of
23 occupational exposure to ammonia. AIHAT 50: 646-650.
24 http://dx.doi.org/10.1080/15298668991375308
25 Holzman, IR: Lemons, TA: Meschia, G: Battaglia, FC. [1977]. Ammonia production by the pregnant
26 uterus [39868]. Proc Soc Exp Biol Med 156: 27-30.
27 Holzman, IR: Philipps, AF: Battaglia, FC. [1979]. Glucose metabolism, lactate, and ammonia
28 production by the human placenta in vitro. Pediatr Res 13: 117-120.
29 HSDB [Hazardous Substances Data Bank). [2012]. Ammonia. Washington. DC: National Library of
30 Medicine.
31 IARC [International Agency for Research on Cancer]. [2006]. Preamble to the IARC monographs.
32 Lyon. France. http://monographs.iarc.fr/ENG/Preamble/
33 Ihrig. A: Hoffmann. T: Triebig. G. [2006]. Examination of the influence of personal traits and
34 habituation on the reporting of complaints at experimental exposure to ammonia. Int Arch
35 Occup Environ Health 79: 332-338. http://dx.doi.org/10.1007/s00420-005-0042-y
36 IOM [Institute of Medicine]. [2008]. Improving the presumptive disability decision-making process
37 for veterans. InTM Samet: CC Bodurow [Eds.]. Washington, DC: National Academies Press.
38 http://www.nap.edu/openbook.php7record id=11908
39 IPCS [International Programme on Chemical Safety]. [1986]. Environmental health criteria:
40 Ammonia. [EHC 54). Geneva. Switzerland: World Health Organization.
41 http://www.inchem.org/documents/ehc/ehc/ehc54.htm
42 Tarudi. NI: Golden. B. (19731 Ammonia eve injuries. T Iowa Med Soc 63: 260-263.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Teffers, LT: Dubow, RA: Zieve, L: Reddy, KR: Livingstone, AS: Neimark, S: Viamonte, M: Schiff, ER.
2 [1988]. Hepatic encephalopathy and orotic aciduria associated with hepatocellular
3 carcinoma in a noncirrhotic liver. Hepatology 8: 78-81.
4 Tohnson, PR: Bedick, CR: Clark, NN: Mckain, PL. [2009]. Design and testing of an independently
5 controlled urea SCR retrofit system for the reduction of NOx emissions from marine diesels.
6 Environ SciTechnol 43: 3959-3963. http://dx.doi.org/10.1021/es900269p
7 Tohnson. RL: Gilbert. M: Block. SM: Battaglia. FC. fl9861 Uterine metabolism of the pregnant rabbit
8 under chronic steady-state conditions. Am T Obstet Gynecol 154: 1146-1151.
9 Tones, EA. [2002]. Ammonia, the GABA neurotransmitter system, and hepatic encephalopathy
10 [Review]. Metab Brain Pis 17: 275-281.
11 Tozwik. M: Tozwik. M: Pietrzycki. B: Chojnowski. M: Teng. C: Tozwik. M: Battaglia. FC. [2005].
12 Maternal and fetal blood ammonia concentrations in normal term human pregnancies. Biol
13 Neonate 87: 38-43. http://dx.doi.org/10.1159/000081702
14 Tozwik, M: Teng, C: Meschia, G: Battaglia, FC. [1999]. Contribution of branched-chain amino acids to
15 uteroplacental ammonia production in sheep. Biol Reprod 61: 792-796.
16 Kawano, S: Tsujii, M: Fusamoto, H: Sato, N: Kamada, T. [1991]. Chronic effect of intragastric
17 ammonia on gastric mucosal structures in rats. Dig Pis Sci 36: 33-38.
18 Kennedy. SM: Le Moual. N: Choudat. D: Kauffmann. F. [2000]. Development of an asthma specific job
19 exposure matrix and its application in the epidemiological study of genetics and
20 environment in asthma [EGEA]. Occup Environ Med 57: 635-641.
21 Klein. T: Olson. KR: McKinney. HE. [1985]. Caustic injury from household ammonia. Am T Emerg Med
22 3:320.
23 Klendshoj, NC: Rejent, TA. [1966]. Tissue levels of some poisoning agents less frequently
24 encountered. T Forensic Sci 11: 75-80.
25 Larson, TV: Covert, PS: Frank, R: Charlson, RT. [1977]. Ammonia in the human airways:
26 Neutralization of inspired acid sulfate aerosols. Science 197: 161-163.
27 Le Moual. N: Kauffmann. F: Eisen. EA: Kennedy. SM. [20081. The healthy worker effect in asthma:
28 Work may cause asthma, but asthma may also influence work [Review]. Am T Respir Grit
29 Care Med 177: 4-10. http://dx.doi.org/10.1164/rccm.200703-415PP
30 Lemiere. C: Begin. D: Camus. M: Forget. A: Boulet. LP: Gerin. M. [2012]. Occupational risk factors
31 associated with work-exacerbated asthma in Quebec. Occup Environ Med 69: 901-907.
32 http://dx.doi.org/10.1136/oemed-2012-100663
33 Lide. DR. [2008]. CRC handbook of chemistry and physics. In PR Lide [Ed.]. [88th ed). Boca Raton.
34 FL: CRC Press.
35 Lina, BAR: Kuijpers, MHM. [2004]. Toxicity and carcinogenicity of acidogenic or alkalogenic diets in
36 rats: effects of feeding NHf4]Cl. KHCOm orKCl. Food Chem Toxicol 42: 135-153.
37 Lobasov, M: Smirnov, F. [1934]. Nature of the action of chemical agents on mutational process in
38 Drosophila melanogaster: II. The effect of ammonia on the occurrence of lethal
39 transgenations. C R Biol 3: 174-176.
40 Lopez. GP: Dean. BS: Krenzelok. EP. [1988]. Oral-exposure to ammonia inhalants: A report of 8
41 cases [Abstract!. Vet Hum Toxicol 30:350.
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Luschinsky, HL. [1951]. The activity of glutaminase in the human placenta. Arch Biochem Biophys
2 31:132-140.
3 MacEwen, ID: Theodore, T: Vernot, EH. [1970]. Human exposure to EEL concentrations of
4 monomethylhydrazine. In Proceedings of the first annual conference on environmental
5 toxicology. Wright-Patterson Air Force Base, OH: Aerospace Medical Research Laboratory.
6 Manninen. A: Anttila. S: Savolainen. H. [1988]. Rat metabolic adaptation to ammonia inhalation.
7 Proc Soc Exp Biol Med 187: 278-281.
8 Manninen, ATA: Savolainen, H. [1989]. Effect of short-term ammonia inhalation on selected amino
9 acids in rat brain. Pharmacol Toxicol 64: 244-246.
10 Medina-Ramon. M: Zock. IP: Kogevinas. M: Sunyer. 1: Basagana. X: Schwartz. 1: Burge. PS: Moore. V:
11 Anto. TM. [2006]. Short-term respiratory effects of cleaning exposures in female domestic
12 cleaners. Eur Respir 1 27: 1196-1203. http://dx.doi.org/10.1183/09031936.06.00085405
13 Medina-Ramon, M: Zock, IP: Kogevinas, M: Sunyer, 1: Torralba, Y: Borrell, A: Burgos, F: Anto, TM.
14 [2005]. Asthma, chronic bronchitis, and exposure to irritant agents in occupational
15 domestic cleaning: A nested case-control study. Occup Environ Med 62: 598-606.
16 http://dx.doi.org/10.1136/oem.2004.017640
17 Megraud. F: Neman-Simha. V: Briigmann. D. [1992]. Further evidence of the toxic effect of ammonia
18 produced by Helicobacter pylori urease on human epithelial cells. Infect Immun 60: 1858-
19 1863.
20 Melbostad, E: Eduard, W. [2001]. Organic dust-related respiratory and eye irritation in Norwegian
21 farmers. Am 1 Ind Med 39: 209-217.
22 Meschia. G: Battaglia. FC: Hay. WW: Sparks. TW. fl9801 Utilization of substrates by the ovine
23 placenta in vivo. Fed Proc 39: 245-249.
24 Millea, TP: Kucan, TO: Smoot, EC. [1989]. Anhydrous ammonia injuries. Journal of Burn Care and
25 Rehabilitation 10: 448-453.
26 Minana, MD: Marcaida, G: Grisolia, S: Felipo, V. [1995]. Prenatal exposure of rats to ammonia
27 impairs NMDA receptor function and affords delayed protection against ammonia toxicity
28 and glutamate neurotoxicity. 1 Neuropathol Exp Neurol 54: 644-650.
29 Monso. E: Riu. E: Radon. K: Magarolas. R: Danuser. B: Iversen. M: Morera. 1: Nowak. D. [2004].
30 Chronic obstructive pulmonary disease in never-smoking animal farmers working inside
31 confinement buildings. Ami Ind Med 46: 357-362. http://dx.doi.org/10.1002/ajim.20077
32 Mori. S: Kaneko. H: Mitsuma. T: Hayakawa. T: Yamaguchi. C: Uruma. M. [1998]. Implications of
33 gastric topical bioactive peptides in ammonia-induced acute gastric mucosal lesions in rats.
34 Scand T Gastroenterol 33: 386-393.
35 Murakami. M: Yoo. TK: Teramura. S: Yamamoto. K: Saita. H: Matuo. K: Asada. T: Kita. T. [19901.
36 Generation of ammonia and mucosal lesion formation following hydrolysis of urea by
37 urease in the rat stomach. T Clin Gastroenterol 12: S104-S109.
38 Nagy, L: Kusstatscher, S: Hauschka, PV: Szabo, S. [1996]. Role of cysteine proteases and protease
39 inhibitors in gastric mucosal damage induced by ethanol or ammonia in the rat. T Clin Invest
40 98: 1047-1054. http://dx.doi.org/10.1172/TCI118865
41 Nelson. PL: Cox. MM. [2008]. Amino acid oxidation and the production of urea. In PL Nelson: AL
42 Lehninger: MM Cox [Eds.], Lehninger principles of biochemistry [5th ed., pp. 680, 683]. New
43 York. NY: W.H. Freeman & Co.
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Neumann, R: Mehlhorn, G: Leonhardt, W: Kastner, P: Willig, R: Schimmel, D: Tohannsen, U. [1987].
2 [Experimental studies on the effect on unweaned pigs of aerogenous toxic gas stress, using
3 ammonia of differing concentrations. 1. Clinical picture of ammonia-exposed piglets under
4 the influence of Pasteurella multocida infection, with and without exposure to thermo-
5 motor stress]. T Vet Med B 34: 183-196. http://dx.doi.org/10.1111/i.l439-
6 0450.1987.tb00386.x
7 Norwood. DM: Wainman. T: Lioy. PT: Waldman. TM. [1992]. Breath ammonia depletion and its
8 relevance to acidic aerosol exposure studies. Arch Environ Occup Health 47: 309-313.
9 http://dx.doi.org/10.1080/00039896.1992.9938367
10 NRG [National Research Council). [1983]. Risk assessment in the federal government: Managing the
11 process. Washington. DC: National Academies Press.
12 http://www.nap.edu/openbook.php7record id=366&page=Rl
13 NRG [National Research Council]. [2009]. Science and decisions: Advancing risk assessment
14 Washington, DC: National Academies Press, http://www.nap.edu/catalog/12209.html
15 O'Neil. Ml: Heckelman. PE: Koch. CB: Roman. KT. [20061. Ammonia. In The Merck index: An
16 encyclopedia of chemicals, drugs, and biologicals [14th ed.]. Whitehouse Station, NT: Merck
17 & Co.. Inc.
18 Ota. Y: Hasumura. M: Okamura. M: Takahashi. A: Ueda. M: Onodera. H: Imai. T: Mitsumori. K: Hirose.
19 M. [2006]. Chronic toxicity and carcinogenicity of dietary administered ammonium sulfate
20 in F344 rats. Food Chem Toxicol 44: 17-27. http://dx.doi.Org/10.1016/i.fct.2005.06.001
21 Petrova. M: Diamond. T: Schuster. B: Dalton. P. [2008]. Evaluation of trigeminal sensitivity to
22 ammonia in asthmatics and healthy human volunteers. Inhal Toxicol 20: 1085-1092.
23 http://dx.doi.org/10.1080/08958370802120396
24 Preller, L: Heederik, D: Boleij, ISM: Vogelzang, PFT: Tielen, MTM. [1995]. Lung function and chronic
25 respiratory symptoms of pig farmers: Focus on exposure to endotoxins and ammonia and
26 use of disinfectants. Occup Environ Med 52: 654-660.
27 Rahman, MH: Bratveit, M: Moen, BE. [2007]. Exposure to ammonia and acute respiratory effects in a
28 urea fertilizer factory. Int T Occup Environ Health 13:153-159.
29 Remesar. X: Arola. L: Palou. A: Alemany. M. [1980]. Activities of enzymes involved in amino-acid
30 metabolism in developing rat placenta. Eur T Biochem 110: 289-293.
31 http://dx.doi.0rg/10.llll/i.1432-1033.1980.tb04867.x
32 Reynolds. SI: Donham. KT: Whitten. P: Merchant. TA: Burmeister. LF: Popendorf. WT. [1996].
33 Longitudinal evaluation of dose-response relationships for environmental exposures and
34 pulmonary function in swine production workers. Am T Ind Med 29: 33-40.
35 http://dx.doi.org/10.1002/[SICI]1097-0274[199601]29:l<:33::AID-ATIM5>:3.0.CO:2-#
36 Richard, D: Bouley, G: Boudene, C. [1978a]. [Effects of ammonia gas continuously inhaled by rats
37 and mice]. Bull Europ Physiol Resp 14: 573-582.
38 Rosenbaum, AM: Walner, PL: Dunham, ME: Holinger, LD. [1998]. Ammonia capsule ingestion
39 causing upper aerodigestive tract injury. Otolaryngol Head NeckSurg 119: 678-680.
40 Rosenfeld. M. [1932]. [Experimental modification of mitosis by ammonia]. Archiv fur
41 experimentelle Zellforschung. besonders Gewebeziichtung 14: 1-13.
42 Rosswall, T. [1981]. The biogeochemical nitrogen cycle. In GE Likens [Ed.], Some perspectives of the
43 major biogeochemical cycles fpp. 25-491. Chichester, England: Tohn Wiley & Sons.
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Rothman, KT: Greenland, S. [1998]. Modern epidemiology [2nd ed.]. Philadelphia, PA: Lippincott,
2 Williams. & Wilkins.
3 Rubaltelli, FF: Formentin, PA. [1968]. Ammonia nitrogen, urea and uric acid blood levels in the
4 mother and in both umbilical vessels at delivery. Biol Neonat 13: 147-154.
5 http://dx.doi.org/10.1159/000240142
6 Sadasivudu. B: Radha Krishna Murthy. C. [1978]. Effects of ammonia on monoamine oxidase and
7 enzymes of GABA metabolism in mouse brain. Arch Physiol Biochem 86: 67-82.
8 Sadasivudu, B: Rao, TI: Murthy, CR. [1979]. Chronic metabolic effects of ammonia in mouse brain.
9 Arch Physiol Biochem 87: 871-885.
10 Schaerdel. AD: White. WT: Lang. CM: Dvorchik. BH: Bohner. K. [1983]. Localized and systemic effects
11 of environmental ammonia in rats. T Am Assoc Lab Anim Sci 33: 40-45.
12 Schmidt. FM: Vaittinen. 0: Metsala. M: Lehto. M: Forsblom. C: Groop. PH: Halonen. L. [20131.
13 Ammonia in breath and emitted from skin. T Breath Res 7: 017109.
14 http://dx.doi.Org/10.1088/1752-7155/7/l/017109
15 Schoeb, TR: Davidson, MK: Lindsey, TR. [1982]. Intracage ammonia promotes growth of
16 Mycoplasma pulmonis in the respiratory tract of rats. Infect Immun 38: 212-217.
17 Schubiger. G: Bachmann. C: Barben. P: Colombo. IP: Tonz. 0: Schiipbach. D. [1991]. N-
18 acetylglutamate synthetase deficiency: diagnosis, management and follow-up of a rare
19 disorder of ammonia detoxication. Eur T Pediatr 150: 353-356.
20 Shimizu. H: Suzuki. Y: Takemura. N: Goto. S: Matsushita. H. [19851. The results of microbial
21 mutation test for forty-three industrial chemicals. Sangyo Igaku 27: 400-419.
22 Sigurdarson. ST: O'Shaughnessy. PT: Watt. TA: Kline. TN. f20041 Experimental human exposure to
23 inhaled grain dust and ammonia: towards a model of concentrated animal feeding
24 operations. Am T Ind Med 46: 345-348. http://dx.doi.org/10.1002/ajim.20055
25 Silverman, L: Whittenberger, TL: Muller, T. [1949]. Physiological response of man to ammonia in low
26 concentrations. T Ind Hyg Toxicol 31: 74-78.
27 Smeets, MA: Bulsing, PT: van Rooden, S: Steinmann, R: de Ru, TA: Ogink, NW: van Thriel, C: Dalton,
28 PH. [2007]. Odor and irritation thresholds for ammonia: a comparison between static and
29 dynamic olfactometry. Chem Senses 32: 11-20. http://dx.doi.org/10.1093/chemse/bil031
30 Smith. D: Spanel. P: Davies. S. [1999]. Trace gases in breath of healthy volunteers when fasting and
31 after a protein-calorie meal: a preliminary study. T Appl Physiol 87: 1584-1588.
32 Smith. D: Wang. T: Pysanenko. A: Spanel. P. [2008]. A selected ion flow tube mass spectrometry
33 study of ammonia in mouth- and nose-exhaled breath and in the oral cavity. Rapid Commun
34 Mass Spectrom 22: 783-789. http://dx.doi.org/10.1002/rcm.3434
35 Socolow, RH. [1999]. Nitrogen management and the future of food: lessons from the management of
36 energy and carbon. PNAS 96: 6001-6008.
37 Souba, WW. [1987]. Interorgan ammonia metabolism in health and disease: a surgeon's view
38 [Review]. TPEN T Parenter Enteral Nutr 11: 569-579.
39 Spanel, P: Dryahina, K: Smith, D. [2007a]. Acetone, ammonia and hydrogen cyanide in exhaled
40 breath of several volunteers aged 4-83 years. T Breath Res 1: 011001.
41 htto://dx.doi.org/10.1088/1752-7155/l/l/011001
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Spanel, P: Dryahina, K: Smith, D. [2007b]. The concentration distributions of some metabolites in
2 the exhaled breath of young adults. T Breath Res 1: 026001.
3 http://dx.doi.Org/10.1088/1752-7155/l/2/026001
4 Spanel, P: Dryahina, K: Smith, D. [2013]. A quantitative study of the influence of inhaled compounds
5 on their concentrations in exhaled breath. T Breath Res 7: 017106.
6 http://dx.doi.Org/10.1088/1752-7155/7/l/017106
7 Stombaugh. DP: Teague. HS: Roller. WL. [1969]. Effects of atmospheric ammonia on the pig. T Anim
8 Sci 28: 844-847.
9 Sundblad, BM: Larsson, BM: Acevedo, F: Ernstgard, L: Tohanson, G: Larsson, K: Palmberg, L. [2004].
10 Acute respiratory effects of exposure to ammonia on healthy persons. Scand T Work Environ
11 Health 30: 313-321.
12 Suzuki, H: Mori, M: Suzuki, M: Sakurai, K: Miura, S: Ishii, H. [1997]. Extensive DNA damage induced
13 by monochloramine in gastric cells. Cancer Lett 115: 243-248.
14 Suzuki. H: Seto. K: Mori. M: Suzuki. M: Miura. S: Ishii. H. [19981. Monochloramine induced DNA
15 fragmentation in gastric cell line MKN45. Am T Physiol 275: G712-G716.
16 Takeuchi, K: Ohuchi, T: Harada, H: Okabe, S. [1995]. Irritant and protective action of urea-urease
17 ammonia in rat gastric mucosa. Different effects of ammonia and ammonium ion. Dig Pis Sci
18 40:274-281.
19 Targowski, SP: Klucinski, W: Babiker, S: Nonnecke, BT. [1984]. Effect of ammonia on in vivo and in
20 vitro immune responses. Infect Immun 43: 289-293.
21 Toth. B. [1972]. Hydrazine. methylhydrazine and methylhydrazine sulfate carcinogenesis in Swiss
22 mice. Failure of ammonium hydroxide to interfere in the development of tumors. IntT
23 Cancer 9: 109-118.
24 Tsujii, M: Kawano, S: Tsuji, S: Fusamoto, H: Kamada, T: Sato, N. [1992a]. Mechanism of gastric
25 mucosal damage induced by ammonia. Gastroenterology 102: 1881-1888.
26 Tsujii, M: Kawano, S: Tsuji, S: Nagano, K: Ito, T: Hayashi, N: Fusamoto, H: Kamada, T: Tamura, K.
27 [1992b]. Ammonia: a possible promotor in Helicobacter pylori-related gastric
28 carcinogenesis. Cancer Lett 65: 15-18.
29 Tsuiii. M: Kawano. S: Tsuii. S: Ito. T: Nagano. K: Sasaki. Y: Hayashi. N: Fusamoto. H: Kamada. T.
30 [1993]. Cell kinetics of mucosal atrophy in rat stomach induced by long-term
31 administration of ammonia. Gastroenterology 104: 796-801.
32 Tsujii. M: Kawano. S: Tsuji. S: Takei. Y: Tamura. K: Fusamoto. H: Kamada. T. [1995]. Mechanism for
33 ammonia-induced promotion of gastric carcinogenesis in rats. Carcinogenesis 16: 563-566.
34 Turner, C: Spanel, P: Smith, D. [2006]. A longitudinal study of ammonia, acetone and propanol in the
35 exhaled breath of 30 subjects using selected ion flow tube mass spectrometry, SIFT-MS.
36 Physiol Meas 27: 321-337. http://dx.doi.Org/10.1088/0967-3334/27/4/001
37 U.S. Congress. [20111. Consolidated Appropriations Act. 2012. [Pub. L. No. 112-74: 125 STAT. 7861.
38 112th U.S. Congress. http://www.gpo.gov/fdsys/pkg/PLAW-112publ74/pdf/PLAW-
39 112publ74.pdf
40 U.S. EPA [U.S. Environmental Protection Agency). [1986a]. Guidelines for mutagenicity risk
41 assessment [EPA Report]. [EPA/630/R-98/003]. Washington. DC.
42 http://www.epa.gov/iris/backgrd.html
This document is a draft for review purposes only and does not constitute Agency policy.
R-9 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 U.S. EPA [U.S. Environmental Protection Agency]. [1986b]. Guidelines for the health risk
2 assessment of chemical mixtures [EPAReport]. (EPA/630/R-98/002J. Washington. DC.
3 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2 25 67
4 U.S. EPA [U.S. Environmental Protection Agency]. [1988]. Recommendations for and documentation
5 of biological values for use in risk assessment [EPA Report]. fEPA/600/6-87/0081.
6 Cincinnati. OH. ht±p://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=34855
7 U.S. EPA [U.S. Environmental Protection Agency). [1991]. Guidelines for developmental toxicity risk
8 assessment [EPAReport]. (EPA/600/FR-91/001J. Washington. DC: U.S. Environmental
9 Protection Agency, Risk Assessment Forum.
10 http://www.epa.gov/raf/publications/guidelines-dev-toxicity-risk-assessmenthtm
11 U.S. EPA [U.S. Environmental Protection Agency). [1994]. Methods for derivation of inhalation
12 reference concentrations and application of inhalation dosimetry [EPA Report].
13 fEPA/600/8-90/066Fj. Research Triangle Park. NC.
14 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=71993
15 U.S. EPA [U.S. Environmental Protection Agency]. [1995]. The use of the benchmark dose approach
16 in health risk assessment [EPA Report]. (EPA/630/R-94/007J. Washington. DC.
17 http://www.epa.gov/raf/publications/useof-bda-healthrisk.htm
18 U.S. EPA [U.S. Environmental Protection Agency). [1996]. Guidelines for reproductive toxicity risk
19 assessment [EPAReport]. fEPA/630/R-96/0091. Washington. DC.
20 http://www.epa.gov/raf/publications/pdfs/REPR051.PDF
21 U.S. EPA [U.S. Environmental Protection Agency). [1998a]. 1998 update of ambient water quality
22 criteria for ammonia [EPA Report]. fEPA822R98008).
23 http://nepis.epa. gov/Exe/ZyPURL.cgi?Dockey=P1005TLO.txt
24 U.S. EPA [U.S. Environmental Protection Agency]. [1998bj. Guidelines for neurotoxicity risk
25 assessment [EPAReport]. [EPA/630/R-95/001F]. Washington. DC.
26 http://www.epa.gov/raf/publications/pdfs/NEUROTOX.PDF
27 U.S. EPA [U.S. Environmental Protection Agency]. [2000]. Supplementary guidance for conducting
28 health risk assessment of chemical mixtures [EPA Report]. fEPA/630/R-00/0021.
29 Washington. DC. ht±p://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=20533
30 U.S. EPA [U.S. Environmental Protection Agency). [2002]. A review of the reference dose and
31 reference concentration processes [EPA Report]. [EPA/630/P-02/002F]. Washington, DC:
32 Risk Assessment Forum. U.S. Environmental Protection Agency.
33 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=51717
34 U.S. EPA [U.S. Environmental Protection Agency]. (2005aj. Guidelines for carcinogen risk
35 assessment [EPAReport]. rEPA/630/P-03/001F1. Washington. DC: Risk Assessment Forum.
36 http://www.epa.gov/cancerguidelines/
37 U.S. EPA [U.S. Environmental Protection Agency]. (2005bj. Supplemental guidance for assessing
38 susceptibility from early-life exposure to carcinogens [EPA Report] [pp. 1125-1133].
39 fEPA/630/R-03/003Fj. Washington. DC.
40 http://www.epa.gov/cancerguidelines/guidelines-carcinogen-supplement.htm
41 U.S. EPA [U.S. Environmental Protection Agency). [2006a]. Approaches for the application of
42 physiologically based pharmacokinetic [PBPK] models and supporting data in risk
43 assessment [Final Report] [EPA Report]. fEPA/600/R-05/043F1. Washington. DC.
44 htto://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=l 57668
This document is a draft for review purposes only and does not constitute Agency policy.
R-10 DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 U.S. EPA [U.S. Environmental Protection Agency]. [2006b]. A framework for assessing health risk of
2 environmental exposures to children [EPA Report]. (EPA/600/R-05/093FJ. Washington.
3 DC. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=l58363
4 U.S. EPA [U.S. Environmental Protection Agency]. [2007]. Integrated Risk Information System
5 (IRIS): Announcement of 2008 program. Fed Reg 72: 72715-72719.
6 U.S. EPA [U.S. Environmental Protection Agency). [2009a]. EPAs Integrated Risk Information
7 System: Assessment development process [EPA Report]. Washington. DC.
8 http://epa.gov/iris/process.htm
9 U.S. EPA [U.S. Environmental Protection Agency]. [2009b]. Integrated Risk Information System
10 [IRIS]: Announcement of Availability of Literature Searches for IRIS Assessments
11 [FRL89741: Docket ID No. EPAHOORD20070664]. Fed Reg 74: 56611-56612.
12 U.S. EPA [U.S. Environmental Protection Agency]. [2010]. Integrated science assessment for carbon
13 monoxide [EPA Report]. rEPA/600/R-09/019F1. Research Triangle Park. NC.
14 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686
15 U.S. EPA [U.S. Environmental Protection Agency]. [2011a]. Exposure factors handbook 2011 edition
16 [final] [EPA Report]. fEPA/600/R-09/052Fj.
17 http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2 36252
18 U.S. EPA [U.S. Environmental Protection Agency). [2011b]. Recommended use of body weight 3/4
19 as the default method in derivation of the oral reference dose [EPA Report].
20 fEPA/lOO/Rll/OOOlj. Washington. DC.
21 http://www.epa.gov/raf/publications/interspecies-extrapolation.htm
22 U.S. EPA [U.S. Environmental Protection Agency). (2012a). Advances in inhalation gas dosimetry for
23 derivation of a reference concentration [rfc] and use in risk assessment [EPA Report].
24 fEPA/600/R-12/044j. Washington. DC.
25 http://cfpub.epa.gov/ ncea/cfm/recordisplay.cfm?deid=2 44650
26 U.S. EPA [U.S. Environmental Protection Agency]. [2012bj. Benchmark dose technical guidance.
27 fEPA/lOO/R-12/0011. Washington. DC: Risk Assessment Forum.
28 http://www.epa.gov/raf/publications/pdfs/benchmark dose guidance.pdf
29 Verberk. MM. [1977]. Effects of ammonia in volunteers. IntArch Occup Environ Health 39: 73-81.
30 http://dx.doi.org/10.1007/BF00380887
31 Vizcaya, D: Mirabelli, MC: Anto. TM: Orriols, R: Burgos, F: Arjona, L: Zock, IP. [2011]. A workforce-
32 based study of occupational exposures and asthma symptoms in cleaning workers. Occup
33 Environ Med 68: 914-919. http://dx.doi.org/10.1136/oem.2010.063271
34 Vollmuth, TA: Schlesinger, RB. [1984]. Measurement of respiratory tract ammonia in the rabbit and
35 implications to sulfuric acid inhalation studies. Toxicol Sci 4: 455-464.
36 Wason, S: Stephan, M: Breide, C. [1990]. Ingestion of aromatic ammonia 'smelling salts' capsules
37 [Letter]. Am 1 Pis Child 144: 139-140.
38 http://dx.doi.org/10.1001/archpedi.1990.02150260017009
39 Weatherby, TH. [1952]. Chronic toxicity of ammonia fumes by inhalation. Proc Soc Exp Biol Med 81:
40 300-301.
41 Yadav. IS: Kaushik. VK. [1997]. Genotoxic effect of ammonia exposure on workers in a fertilizer
42 factory. Indian 1 Exp Biol 35: 487-492.
This document is a draft for review purposes only and does not constitute Agency policy.
R-ll DRAFT—DO NOT CITE OR QUOTE
-------
Toxicological Review of Ammonia
1 Zejda. IE: Barber. E: Dosman. TA: Olenchock. SA: McDuffie. HH: Rhodes. C: Hurst. T. [1994].
2 Respiratory health status in swine producers relates to endotoxin exposure in the presence
3 of low dust levels. TOccupMed 36: 49-56.
4 Zock. IP: Plana. E: Tarvis. D: Anto. TM: Kromhout. H: Kennedy. SM: Kiinzli. N: Villani. S: Olivieri. M:
5 Toren. K: Radon. K: Sunyer. 1: Dahlman-Hoglund. A: Norback. D: Kogevinas. M. [20071. The
6 use of household cleaning sprays and adult asthma: An international longitudinal study. Am
7 1 Respir Grit Care Med 176: 735-741. ht±D://dx.doi.org/10.1164/rccm.200612-17930C
This document is a draft for review purposes only and does not constitute Agency policy.
R-12 DRAFT—DO NOT CITE OR QUOTE
------- |