DRAFT-DO NOT CITE OR QUOTE
EPA/635/R-10/002C
www.epa.gov/iris
&EPA
TOXICOLOGICAL REVIEW OF
FORMALDEHYDE
INHALATION TOXICITY
(CAS No. 50-00-0)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
VOLUME III of IV
Quantitative Assessment, Major Conclusions in
the Characterization of Hazard and Dose
Response, and References
March 17,2010
NOTICE
This document is an Inter-Agency Science Consultation 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.
U.S. Environmental Protection Agency
Washington, DC
-------
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.
IH-ii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS—TOXICOLOGICAL REVIEW OF FORMALDEHYDE
(CAS No. 50-00-0)
LIST OF TABLES xi
LIST OF FIGURES xx
LIST 01 ABBREVIATIONS AM) ACRONYMS xxv
FOREWORD xxxii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxxiii
VOLUME I
1. INTRODUCTION 1-1
2. BACKGROUND 2-1
2.1. PI IYSICOCIIHYIICAI. PROPERTIES OF FORMALDEHYDE 2-1
2.2. PRODUCTION, USES, AND SOURCES OF FORMALDEHYDE 2-1
2.3. ENVIRONMENTAL LEVELS AND HUMAN EXPOSURE 2-4
2.3.1. Inhalation 2-5
2.3.2. Ingestion 2-10
2.3.3. Dermal Contact 2-11
3. TOXICOKINETICS 3-1
3.1. CHEMICAL PROPERTIES AND REACTIVITY 3-1
3.1.1. Binding of Formaldehyde to Proteins 3-1
3.1.2. Endogenous Sources of Formaldehyde 3-3
3.1.2.1. Normal Cellular Metabolism (Enzymatic) 3-3
3.1.2.2. Normal Metabolism (Non-Enzymatic) 3-5
3.1.2.3. Exogenous Sources of Formaldehyde Production 3-5
3.1.2.4. FA-GSH Conjugate as a Method of Systemic Distribution 3-6
3.1.2.5. Metabolic Products of FA Metabolism (e.g., Formic Acid) 3-6
3.1.2.6. Levels of Endogenous Formaldehyde in Animal and Human
Tissues 3-6
3.2. ABSORPTION 3-9
3.2.1. Oral 3-9
3.2.2. Dermal 3-9
3.2.3. Inhalation 3-9
3.2.3.1. Formaldehyde Uptake Can be Affected by Effects at the
Portal of Entry 3-10
3.2.3.2. Variability in the Nasal Dosimetry of Formaldehyde in
Adults and Children 3-12
3.3. DISTRIBUTION 3-13
3.3.1. Levels in Blood 3-13
3.3.2. Levels in Various Tissues 3-15
3.3.2.1. Disposition of Formaldehyde: Differentiating covalent
Binding and Metabolic Incorporation 3-16
This document is a draft for review purposes only and does not constitute Agency policy.
III-iii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
3.4. METABOLISM 3-20
3.4.1. In Vitro and In Vivo Characterization of Formaldehyde Metabolism 3-20
3.4.2. Formaldehyde Exposure and Perturbation of Metabolic Pathways 3-23
3.4.3. Evidence for Susceptibility in Formaldehyde Metabolism 3-24
3.5. EXCRETION 3-25
3.5.1. Formaldehyde Excretion in Rodents 3-26
3.5.2. Formaldehyde Excretion in Exhaled Human Breath 3-27
3.5.3. Formaldehyde Excretion in Human Urine 3-31
3.6. MODELING THE TOXICOKINETICS OF FORMALDEHYDE AND DPX 3-32
3.6.1. Motivation 3-32
3.6.2. Species Differences in Anatomy: Consequences for Gas Transport and
Risk 3-34
3.6.3. Modeling Formaldehyde Uptake in Nasal Passages 3-40
3.6.3.1. Flux Bins 3-41
3.6.3.2. Flux Estimates 3-41
3.6.3.3. Mass Balance Errors 3-42
3.6.4. Modeling Formaldehyde Uptake in the Lower Respiratory Tract 3-42
3.6.5. Uncertainties in Formaldehyde Dosimetry Modeling 3-44
3.6.5.1. Verification of Predicted Flow Profiles 3-44
3.6.5.2. Level of Confidence in Formaldehyde Uptake Simulations 3-45
3.6.6. PBPK Modeling of DNA Protein Cross-Links (DPXs) Formed by
Formaldehyde 3-48
3.6.6.1. PBPK Models for DPXs 3-48
3.6.6.2. A PBPK Model for DPXs in the F344 Rat and Rhesus
Monkey that uses Local Tissue Dose of Formaldehyde 3-50
3.6.6.3. Uncertainties in Modeling the Rat and Rhesus DPX Data 3-51
3.6.7. Uncertainty in Prediction of Human DPX Concentrations 3-53
VOLUME II
4. HAZARD CHARACTERIZATION 4-1
4.1. HUMAN STUDIES 4-1
4.1.1. Noncancer Health Effects 4-1
4.1.1.1. Sensory Irritation (Eye, Nose, Throat Irritation) 4-1
4.1.1.2. Pulmonary Function 4-11
4.1.1.3. Asthma 4-19
4.1.1.4. Respiratory Tract Pathology 4-26
4.1.1.5. Immunologic Effects 4-30
4.1.1.6. Neurological/Behavioral 4-42
4.1.1.7. Developmental and Reproductive Toxicity 4-45
4.1.1.8. Oral Exposure Effects on the Gastrointestinal Tract 4-56
4.1.1.9. Summary: Noncarcinogenic Hazard in Humans 4-56
This document is a draft for review purposes only and does not constitute Agency policy.
in-iv DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.1.2. Cancer Health Effects 4-57
4.1.2.1. Respiratory Tract Cancer 4-57
4.1.2.2. Non-Respiratory Tract Cancer 4-84
4.1.2.3. Summary: Carcinogenic Hazard in Humans 4-107
4.2. ANIMAL STUDIES 4-109
4.2.1. Noncancer Health Effects 4-110
4.2.1.1. Reflex Bradypnea 4-110
4.2.1.2. Respiratory Tract Pathology 4-120
4.2.1.3. Gastrointestinal Tract and Systemic Toxicity 4-201
4.2.1.4. Immune Function 4-216
4.2.1.5. Hypersensitivity and Atopic Reactions 4-225
4.2.1.6. Neurological and Neurobehavioral Function 4-250
4.2.1.7. Reproductive and Developmental Toxicity 4-285
4.2.2. Carcinogenic Potential: Animal Bioassays 4-324
4.2.2.1. Respiratory Tract 4-324
4.2.2.2. Gastrointestinal Tract 4-326
4.2.2.3. Lymphohematopoietic Cancer 4-328
4.2.2.4. Summary 4-335
4.3. GENOTOXICITY 4-335
4.3.1. Formaldehyde-DNAReactions 4-335
4.3.1.1. DNA-Protein Cross-Links (DPXs) 4-336
4.3.1.2. DNA Adducts 4-341
4.3.1.3. DNA-DNA Cross-Links (DDXs) 4-343
4.3.1.4. Single Strand Breaks 4-344
4.3.1.5. Other Genetic Effects of Formaldehyde in Mammalian Cells 4-345
4.3.2. In Vitro Clastogenicity 4-345
4.3.3. In Vitro Mutagenicity 4-347
4.3.3.1. Mutagenicity in Bacterial Systems 4-347
4.3.3.2. Mutagenicity in Non-mammalian Cell Systems 4-353
4.3.3.3. Mutagenicity in Mammalian Cell Systems 4-353
4.3.4. In Vivo Mammalian Genotoxicity 4-360
4.3.4.1. Genotoxicity in Laboratory Animals 4-360
4.3.4.2. Genotoxicity in Humans 4-362
4.3.5. Summary of Genotoxicity 4-370
4.4. SYNTHESIS AND MAJOR EVALUATION OF NONCARCINOGENIC
EFFECTS 4-371
4.4.1. Sensory Irritation 4-376
4.4.2. Pulmonary Function 4-379
4.4.3. Hypersensitivity and Atopic Reactions 4-382
4.4.4. Upper Respiratory Tract Histopathology 4-383
4.4.5. Toxicogenomic and Molecular Data that May Inform MOAs 4-385
4.4.6. Noncancer Modes of Actions 4-387
4.4.7. Immunotoxicity 4-389
This document is a draft for review purposes only and does not constitute Agency policy.
III-v DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.4.8. Effects on the Nervous System 4-390
4.4.8.1. Irritant Threshold Detection 4-390
4.4.8.2. Behavioral Effects 4-391
4.4.8.3. Neurochemistry, Neuropathology, and Mechanistic Studies 4-392
4.4.8.4. Summary 4-392
4.4.8.5. Data Gaps 4-393
4.4.9. Reproductive and Developmental Toxicity 4-393
4.4.9.1. Spontaneous Abortion and Fetal Death 4-393
4.4.9.2. Congenital Malformations 4-396
4.4.9.3. Low Birth Weight and Growth Retardation 4-396
4.4.9.4. Functional Development Outcomes (Developmental
Neurotoxicity) 4-397
4.4.9.5. Male Reproductive Toxicity 4-398
4.4.9.6. Female Reproductive Toxicity 4-399
4.4.9.7. Mode of Action 4-400
4.4.9.8. Data Gaps 4-402
4.5. SYNTHESIS AND EVALUATION OF CARCINOGENICITY 4-402
4.5.1. Cancers of the Respiratory Tract 4-402
4.5.2. Lymphohematopoietic Malignancies 4-408
4.5.2.1. Background 4-408
4.5.2.2. All LHP Malignancies 4-410
4.5.2.3. All Leukemia 4-414
4.5.2.4. Subtype Analysis 4-418
4.5.2.5. Myeloid Leukemia 4-419
4.5.2.6. Solid Tumors of Lymphoid Origin 4-421
4.5.2.7. Supporting Evidence from Animal Bio-Assays for
Formaldehyde-Induced Lymphohematopoietic Malignancies 4-423
4.5.3. Carcinogenic Mode(s) of Action 4-427
4.5.3.1. Mechanistic Data for Formaldehyde 4-428
4.5.3.2. Mode of Action Evaluation for Upper Respiratory Tract
Cancer (Nasopharyngeal Cancer, Sino-Nasal) 4-439
4.5.3.3. Mode(s) of Action for Lymphohematpoietic Malignancies 4-446
4.5.4. Hazard Characterization for Formaldlehyde Carcinogenicity 4-453
4.6. SUSCEPTIBLE POPULATIONS 4-454
4.6.1. Life Stages 4-454
4.6.1.1. Early Life Stages 4-455
4.6.1.2. Later Life Stages 4-459
4.6.1.3. Conclusions on Life-Stage Susceptibility 4-459
4.6.2. Health/Disease Status 4-460
4.6.3. Nutritional Status 4-461
4.6.4. Gender Differences 4-462
4.6.5. Genetic Differences 4-462
This document is a draft for review purposes only and does not constitute Agency policy.
IH-vi DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
4.6.6. Co-Exposures 4-464
4.6.6.1. Cumulative Risk 4-464
4.6.6.2. Aggregate Exposure 4-465
4.6.7. Uncertainties of Database 4-465
4.6.7.1. Uncertainties of Exposure 4-465
4.6.7.2. Uncertainties of Effect 4-466
4.6.8. Summary of Potential Susceptibility 4-467
VOLUME III
5. QUANTITATIVE ASSESSMENT: INHALATION EXPOSURE 5-1
5.1. INHALATION REFERENCE CONCENTRATION (RfC) 5-2
5.1.1. Candidate Critical Effects by Health Effect Category 5-3
5.1.1.1. Sensory Irritation of the Eyes, Nose, and Throat 5-3
5.1.1.2. Upper Respiratory Tract Pathology 5-5
5.1.1.3. Pulmonary Function Effects 5-6
5.1.1.4. Asthma and Allergic Sensitization (Atopy) 5-10
5.1.1.5. Immune Function 5-16
5.1.1.6. Neurological and Behavioral Toxicity 5-17
5.1.1.7. Developmental and Reproductive Toxicity 5-25
5.1.2. Summary of Critical Effects and Candidate RfCs 5-33
5.1.2.1. Selection of Studies for Candidate RfC Derivation 5-33
5.1.2.2. Derivation of Candidate RfCs from Key Studies 5-40
5.1.2.3. Evaluation of the Study-Specific Candidate RfC 5-66
5.1.3. Database Uncertainties in the RfC Derivation 5-69
5.1.4. Uncertainties in the RfC Derivation 5-72
5.1.5. Previous Inhalation Assessment 5-74
5.2. QUANTITATIVE CANCER ASSESSMENT BASED ON THE NATIONAL
CANCER INSTITUTE COHORT STUDY 5-74
5.2.1. Choice of Epidemiology Study 5-75
5.2.2. Nasopharyngeal Cancer 5-76
5.2.2.1. Exposure-Response Modeling of the National Cancer
Institute Cohort 5-76
5.2.2.2. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
Mortality 5-79
5.2.2.3. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer
Incidence 5-81
5.2.2.4. Sources of Uncertainty 5-83
5.2.3. Lymphohematopoietic Cancer 5-88
5.2.3.1. Exposure-Response Modeling of the National Cancer
Institute Cohort 5-88
This document is a draft for review purposes only and does not constitute Agency policy.
IH-vii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
5.2.3.2. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma
and Leukemia Mortality 5-91
5.2.3.3. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma
and Leukemia Incidence 5-93
5.2.3.4. Sources of Uncertainty 5-95
5.2.4. Conclusions on Cancer Unit Risk Estimates Based on Human Data 5-99
5.3. DOSE-RESPONSE MODELING OF RISK OF SQUAMOUS CELL
CARCINOMA IN THE RESPIRATORY TRACT USING ANIMAL DATA 5-102
5.3.1. Long-Term Bioassays in Laboratory Animals 5-104
5.3.1.1. Nasal Tumor Incidence Data 5-104
5.3.1.2. Mechanistic Data 5-105
5.3.2. The CUT Biologically Based Dose-Response Modeling 5-106
5.3.2.1. Maj or Results of the CUT Modeling Effort 5-111
5.3.3. This Assessment's Conclusions from Evaluation of Dose-Response
Models of DPX Cell-Replication and Genomics Data, and of BBDR
Models for Risk Estimation 5-111
5.3.4. Benchmark Dose Approaches to Rat Nasal Tumor Data 5-118
5.3.4.1. Benchmark Dose Derived from BBDR Rat Model and Flux
as Dosimeter 5-118
5.3.4.2. Comparison with Other Benchmark Dose Modeling Efforts 5-125
5.3.4.3. Kaplan-Meier Adjustment 5-128
5.3.4.4. EPA Time-to-Tumor Statistical Modeling 5-129
5.4. CONCLUSIONS FROM THE QUANTITATIVE ASSESSMENT OF
CANCER RISK FROM FORMALDEHYDE EXPOSURE BY INHALATION .. 5-133
5.4.1. Inhalation Unit Risk Estimates Based on Human Data 5-133
5.4.2. Inhalation Unit Risk Estimates Based on Rodent Data 5-133
5.4.3. Summary of Inhalation Unit Risk Estimates 5-135
5.4.4. Application of Age-Dependent Adjustment Factors (ADAFs) 5-136
5.4.5. Conclusions: Cancer Inhalation Unit Risk Estimates 5-137
6. MAJOR CONCLUSIONS IN THE CHARACTERIZATION OF HAZARD AND
DOSE-RESPONSE 6-1
6.1. SUMMARY OF HUMAN HAZARD POTENTIAL 6-1
6.1.1. Exposure 6-1
6.1.2. Absorption, Distribution, Metabolism, and Excretion 6-1
6.1.3. Noncancer Health Effects in Humans and Laboratory Animals 6-4
6.1.3.1. Sensory Irritation 6-4
6.1.3.2. Respiratory Tract Pathology 6-5
6.1.3.3. Effects on Pulmonary Function 6-8
6.1.3.4. Asthmatic Responses and Increased Atopic Symptoms 6-9
6.1.3.5. Effects on the Immune System 6-10
6.1.3.6. Neurological Effects 6-11
6.1.3.7. Reproductive and Developmental Effects 6-12
This document is a draft for review purposes only and does not constitute Agency policy.
III-viii DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
6.1.3.8. Effects on General Systemic Toxicity 6-13
6.1.3.9. Summary 6-14
6.1.4. Carcinogenicity in Human and Laboratory Animals 6-14
6.1.4.1. Carcinogenicity in Humans 6-14
6.1.4.2. Carcinogenicity in Laboratory Animals 6-20
6.1.4.3. Carcinogenic Mode(s) of Action 6-21
6.1.5. Cancer Hazard Characterization 6-24
6.2. DOSE-RESPONSE CHARACTERIZATION 6-25
6.2.1. Noncancer Toxicity: Reference Concentration (RfC) 6-25
6.2.1.1. Assessment Approach Employed 6-25
6.2.1.2. Derivation of Candidate Reference Concentrations 6-25
6.2.1.3. Adequacy of Overall Data Base for RfC Derivation 6-26
6.2.1.4. Uncertainties in the Reference Concentration (RfC) 6-29
6.2.1.5. Conclusions 6-32
6.2.2. Cancer Risk Estimates 6-32
6.2.2.1. Choice of Data 6-32
6.2.2.2. Analysis of Epidemiologic Data 6-33
6.2.2.3. Analysis of Laboratory Animal Data 6-36
6.2.2.4. Extrapolation Aporoaches 6-37
6.2.2.5. Inhalation Unit Risk Estimates for Cancer 6-41
6.2.2.6. Early-Life Susceptibility 6-41
6.2.2.7. Uncertainties in the Quantitative Risk Estimates 6-42
6.2.2.8. Conclusions 6-45
6.3. SUMMARY AND CONCLUSIONS 6-45
REFERENCES R-l
VOLUME IV
APPENDIX A: SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITIONS A-l
APPENDIX B: SIMULATIONS OF INTERINDIVIDUAL AND ADULT-TO-CHILD
VARIABILITY IN REACTIVE GAS UPTAKE IN A SMALL SAMPLE
OF PEOPLE (Garcia et aL 2009) B-l
APPENDIX C: LIFETABLE ANALYSIS C-l
APPENDIX D: MODEL STRUCTURE & CALIBRATION IN CONOLLY ET AL.
(2003, 2004) D-l
APPENDIX E: EVALUATION OF BBDR MODELING OF NASAL CANCER IN THE
F344 RAT: CONOLLY ET AL. (2003) AND ALTERNATIVE
IMPLEMENTATIONS E-1
APPENDIX F: SENSITIVITY ANALYSIS OF BBDR MODEL FOR FORMALDEHYDE
INDUCED RESPIRATORY CANCER IN HUMANS F-l
This document is a draft for review purposes only and does not constitute Agency policy.
III-ix DRAFT—DO NOT CITE OR QUOTE
-------
CONTENTS (continued)
APPENDIX G: EVALUATION OF THE CANCER DOSE-RESPONSE MODELING
OF GENOMIC DATA FOR FORMALDEHYDE RISK ASSESSMENT G-l
APPENDIX H: EXPERT PANEL CONSULTATION ON QUANTITATIVE
EVALUATION OF ANIMAL TOXICOLOGY DATA FOR
ANALYZING CANCER RISK DUE TO INHALED FORMALDEHYDE .. H-l
This document is a draft for review purposes only and does not constitute Agency policy.
III-x DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES
Table 2-1. Physicochemical properties of formaldehyde 2-2
Table 2-2. Ambient air levels by land use category 2-6
Table 2-3. Studies on residential indoor air levels of formaldehyde (non-occupational) 2-8
Table 3-1. Endogenous formaldehyde levels in animal and human tissues and body fluids 3-8
Table 3-2. Formaldehyde kinetics in human and rat tissue samples 3-21
Table 3-3. Allelic frequencies of ADH3 in human populations 3-25
Table 3-4. Percent distribution of airborne [14C]-formaldehyde in F344 rats 3-26
Table 3-5. Apparent formaldehyde levels in exhaled breath of individuals attending a
health fair, adjusted for methanol and ethanol levels which contribute to the
detection of the protonated species with a mass to charge ratio of 31 reported
as formaldehyde (m/z = 31) 3-29
Table 3-6. Measurements of exhaled formaldehyde concentrations in the mouth and nose,
and in the oral cavity after breath holding in three healthy male laboratory
workers 3-30
Table 3-7. Extrapolation of parameters for enzymatic metabolism to the human 3-53
Table 4-1. Cohort and case-control studies of formaldehyde cancer and NPC 4-59
Table 4-2. Case-control studies of formaldehyde and nasal and paranasal cancer 4-71
Table 4-3. Epidemiologic studies of formaldehyde and pharyngeal cancer (includes
nasopharyngeal cancer) 4-78
Table 4-4. Epidemiologic studies of formaldehyde and lymphohematopoietic cancers 4-98
Table 4-5. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of mice 4-112
Table 4-6. Respiratory effects of formaldehyde-induced reflex bradypnea in various
strains of rats 4-113
Table 4-7. Inhaled dose of formaldehyde to nasal mucosa of F344 rats and B6C3F1
mice exposed to 15 ppm 4-116
Table 4-8. Exposure regimen for cross-tolerance study 4-117
This document is a draft for review purposes only and does not constitute Agency policy.
IH-xi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-9. Summary of formaldehyde effects on mucociliary function in the upper
respiratory tract 4-127
Table 4-10. Concentration regimens for ultrastructural evaluation of male CDF rat
nasoturbinates 4-129
Table 4-11. Enzymatic activities in nasal respiratory epithelium of male Wistar rats
exposed to formaldehyde, ozone, or both 4-130
Table 4-12. Lipid analysis of lung tissue and lung gavage from male F344 rats exposed
to 0, 15, or 145.6 ppm formaldehyde for 6 hours 4-138
Table 4-13. Formaldehyde effects on biochemical parameters in nasal mucosa and lung
tissue homogenates from male F344 rats exposed to 0, 15, or 145.6 ppm
formaldehyde for 6 hours 4-139
Table 4-14. Mast cell degranulation and neutrophil infiltration in the lung of rats
exposed to formaldehyde via inhalation 4-140
Table 4-15. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde—short term studies 4-143
Table 4-16. Location and incidence of respiratory tract lesions in B6C3F1 mice
exposed to formaldehyde 4-146
Table 4-17. Formaldehyde effects (incidence and severity) on histopathologic changes in
the noses and larynxes of male and female albino SPF Wistar rats exposed to
formaldehyde 6 hours/day for 13 weeks 4-148
Table 4-18. Formaldehyde-induced nonneoplastic histopathologic changes in male
albino SPF Wistar rats exposed to 0, 10, or 20 ppm formaldehyde
and examined at the end of 130 weeks inclusive of exposure 4-149
Table 4-19. Formaldehyde-induced nasal tumors in male albino SPF Wistar rats
exposed to formaldehyde (6 hours/day, 5 days/week for 13 weeks) and
examined at the end of 130 weeks inclusive of exposure 4-150
Table 4-20. Formaldehyde effects on nasal epithelium for various concentration-by-
time products in male albino Wistar rats 4-153
Table 4-21. Rhinitis observed in formaldehyde-treated animals; data pooled for male
and female animals 4-154
This document is a draft for review purposes only and does not constitute Agency policy.
IH-xii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-22. Epithelial lesions found in the middle region of nasoturbinates of
formaldehyde-treated and control animals; data pooled for males and
females 4-155
Table 4-23. Cellular and molecular changes in nasal tissues of F344 rats exposed to
formaldehyde 4-156
Table 4-24. Percent body weight gain and concentrations of iron, zinc, and copper in
cerebral cortex of male Wistar rats exposed to formaldehyde via inhalation
for 4 and 13 weeks 4-158
Table 4-25. Zinc, copper, and iron content of lung tissue from formaldehyde-treated
male Wistar rats 4-158
Table 4-26. Total lung cytochrome P450 measurements of control and formaldehyde-
treated male Sprague-Dawley rats 4-159
Table 4-27. Cytochrome P450 levels in formaldehyde-treated rats 4-160
Table 4-28. Summary of respiratory tract pathology from inhalation exposures to
formaldehyde, subchronic studies 4-162
Table 4-29. Histopathologic findings and severity scores in the naso- and
maxilloturbinates of female Sprague-Dawley rats exposed to inhaled
formaldehyde and wood dust for 104 weeks 4-166
Table 4-30. Histopathologic changes (including tumors) in nasal cavities of male
Sprague-Dawley rats exposed to inhaled formaldehyde or HC1 alone and
in combination for a lifetime 4-170
Table 4-31. Summary of neoplastic lesions in the nasal cavity of f344 rats exposed to
inhaled formaldehyde for 2 years 4-173
Table 4-32. Apparent sites of origin for the SCCs in the nasal cavity of F344 rats
exposed to 14.3 ppm of formaldehyde gas in the Kerns et al. (1983)
bioassay 4-174
Table 4-33. Incidence and location of nasal squamous cell carcinoma in male F344
rats exposed to inhaled formaldehyde for 2 years 4-175
Table 4-34. Summary of respiratory tract pathology from chronic inhalation exposures
to formaldehyde 4-183
Table 4-35. Cell proliferation in nasal mucosa, trachea, and free lung cells isolated
from male Wistar rats after inhalation exposures to formaldehyde 4-194
This document is a draft for review purposes only and does not constitute Agency policy.
III-xiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-36. The effect of repeated formaldehyde inhalation exposures for 3 months on
cell count, basal membrane length, proliferation cells, and two measures of
cell proliferation, LI and ULLI, in male F344 rats 4-196
Table 4-37. Formaldehyde-induced changes in cell proliferation and (ULLI) in the nasal
passages of male F344 rats exposed 6 hours/day 4-198
Table 4-38. Cell population and surface area estimates in untreated male F344 rats and
regional site location of squamous cell carcinomas in formaldehyde
exposed rats for correlation to cell proliferation rates 4-199
Table 4-39. Summary of formaldehyde effects on cell proliferation in the upper
respiratory tract 4-202
Table 4-40. Summary of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 4 weeks 4-206
Table 4-41. Incidence of lesions observed in the gastrointestinal tracts of Wistar rats
after drinking-water exposure to formaldehyde for 2 years 4-209
Table 4-42. Effect of formaldehyde on gastroduodenal carcinogenesis initiated by
MNNG and NaCl in male Wistar rats exposed to formaldehyde (0.5%
formalin) in drinking water for 8 weeks 4-212
Table 4-43. Summary of benign and malignant gastrointestinal tract neoplasia
reported in male and female Sprague-Dawley rats exposed to
formaldehyde in drinking water at different ages 4-214
Table 4-44. Incidence of hemolymphoreticular neoplasia reported in male and
female Sprague-Dawley rats exposed to formaldehyde in drinking water
from 7 weeks old for up to 2 years (experiment BT 7001) 4-215
Table 4-45. Battery of immune parameters and functional tests assessed in female
B6C3F1 mice after a 3 week, 15-ppm formaldehyde exposure 4-218
Table 4-46. Summary of the effects of formaldehyde inhalation on the mononuclear
phagocyte system (MPS) in female B6C3F1 mice after a 3-week, 15 ppm
formaldehyde exposure ( 6 hours/day, 5 days/week) 4-219
Table 4-47. Formaldehyde exposure regimens for determining the effects of
formaldehyde exposure on pulmonary S. aureus infection 4-221
Table 4-48. Summary of immune function changes due to inhaled formaldehyde
exposure in experimental animals 4-226
This document is a draft for review purposes only and does not constitute Agency policy.
III-xiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-49. Study design for guinea pigs exposed to formaldehyde through different
routes of exposure: inhalation, dermal, and injection 4-232
Table 4-50. Sensitization response of guinea pigs exposed to formaldehyde through
inhalation, topical application, or footpad injection 4-233
Table 4-51. Cytokine and chemokine levels in lung tissue homogenate supernatants in
formaldehyde-exposed male ICR mice with and without Der f sensitization 4-240
Table 4-52. Correlation coefficients among ear swelling responses and skin mRNA
levels in contact hypersensitivity to formaldehyde in mice 4-249
Table 4-53. Summary of sensitization and atopy studies by inhalation or dermal
sensitization due to formaldehyde in experimental animals 4-251
Table 4-54. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 2 hours 4-259
Table 4-55. Fluctuation of behavioral responses when male AB mice inhaled
formaldehyde in a single 2-hour exposure: effects after 24 hours 4-259
Table 4-56. Effects of formaldehyde exposure on completion of the labyrinth test by
male and female LEW. IK rats 4-263
Table 4-57. Summary of neurological and neurobehavioral studies in inhaled
formaldehyde in experimental animals 4-279
Table 4-58. Effects of formaldehyde on body and organ weights in rat pups from
dams exposed via inhalation from mating through gestation 4-289
Table 4-59. Formaldehyde effects on Leydig cell quantity and nuclear damage in adult
male Wistar rats 4-298
Table 4-60. Formaldehyde effects on adult male albino Wistar rats 4-299
Table 4-61. Formaldehyde effects on testosterone levels and seminiferous tubule
diameters in Wistar rats following 91 days of exposure 4-300
Table 4-62. Effects of formaldehyde exposure on seminiferous tubule diameter and
epithelial height in Wistar rats following 18 weeks of exposure 4-302
Table 4-63. Incidence of sperm abnormalities and dominant lethal effects in
formaldehyde-treated mice 4-302
This document is a draft for review purposes only and does not constitute Agency policy.
III-XV DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-64. Body weights of pups born to beagles exposed to formaldehyde during
gestation 4-303
Table 4-65. Testicular weights, sperm head counts, and percentage incidence of
abnormal sperm after oral administration of formaldehyde to male
Wistar rats 4-305
Table 4-66. Effect of formaldehyde on spermatogenic parameters in male Wistar rats
exposed intraperitoneally 4-306
Table 4-67. Incidence of sperm head abnormalities in formaldehyde-treated rats 4-307
Table 4-68. Dominant lethal mutations after exposure of male rats to formaldehyde 4-308
Table 4-69. Summary of reported developmental effects in formaldehyde inhalation
exposure studies 4-311
Table 4-70. Summary of reported developmental effects in formaldehyde oral exposure
studies 4-317
Table 4-71. Summary of reported developmental effects in formaldehyde dermal
exposure studies 4-318
Table 4-72. Summary of reported reproductive effects in formaldehyde inhalation
studies 4-319
Table 4-73. Summary of reported reproductive effects in formaldehyde oral studies 4-322
Table 4-74. Summary of reported reproductive effects in formaldehyde intraperitoneal
studies 4-323
Table 4-75. Summary of chronic bioassays which address rodent leukemia and
lymphoma 4-329
Table 4-76a. Formaldehyde-DNA reactions (DPX formation) 4-340
Table 4-77. Formaldehyde-DNA reactions (DNA adduct formation) 4-343
Table 4-78. Formaldehyde-DNA interactions (single stranded breaks) 4-344
Table 4-79. Other genetic effects of formaldehyde in mammalian cells 4-346
Table 4-80. In vitro clastogenicity of formaldehyde 4-348
Table 4-81. Summary of mutagenicity of formaldehyde in bacterial systems 4-350
This document is a draft for review purposes only and does not constitute Agency policy.
III-xvi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 4-82. Mutagenicity in mammalian cell systems 4-355
Table 4-83. Genotoxicity in laboratory animals 4-361
Table 4-84. MN frequencies in buccal mucosa cells of volunteers exposed to
formaldehyde 4-364
Table 4-85. MN and SCE formation in mortuary science students exposed to
formaldehyde for 85 days 4-364
Table 4-86. Incidence of MN formation in mortuary students exposed to formaldehyde
for 90 days 4-365
Table 4-87. Multivariate regression models linking genomic instability/occupational
exposures to selected endpoints from the MN assay 4-369
Table 4-88. Genotoxicity measures in pathological anatomy laboratory workers and
controls 4-370
Table 4-89. Summary of human cytogenetic studies 4-372
Table 4-90. Summary of cohort and case-control studies which evaluated the
incidence of all LHP cancers in formaldehyde-exposed populations
(ICD-8 Codes: 200-209) and all leukemias (ICD-8 Codes: 204-207) 4-412
Table 4-91. Secondary analysis of published mortality statistics to explore alternative
disease groupings within the broad category of all lymphohematopoetic
malignancies 4-419
Table 4-92. Summary of studies which provide mortality statistics for myeloid
leukemia sub-types 4-420
Table 4-93. Summary of mortality statistics for Hodgkin's lymphoma, lymphoma and
multiple myeloma from cohort analyses of formaldehyde exposed workers 4-422
Table 4-94. Summary of chronic bioassays which address rodent leukemia and
lymphoma 4-424
Table 4-95. Incidence of lymphoblastic leukemia and lymphosarcoma orally dosed in
Sprague-Dawley rats 4-425
Table 4-96. Available evidence for susceptibility factors of concern for formaldehyde
exposure 4-469
This document is a draft for review purposes only and does not constitute Agency policy.
III-xvii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 5-1. Points of departure (POD) for nervous system toxicity in key human and
animal studies 5-19
Table 5-2. Effects of formaldehyde exposure on completion of the labyrinth test by
male and female LEW. IK rats 5-23
Table 5-3. Developmental and reproductive toxicity PODs including dose conversions
and duration adjustments - key human study 5-31
Table 5-4. Summary of candidate studies for formaldehyde RfC development by
health endpoint category 5-36
Table 5-5. Adjustment for non-occupational exposures to formaldehyde 5-64
Table 5-6. Summary of reference concentration (RfC) derivation from critical study and
supporting studies 5-68
Table 5-7. Relative risk estimates for mortality from nasopharyngeal malignancies
(ICD-8 code 147) by level of formaldehyde exposure for different
exposure metrics 5-78
Table 5-8. Regression coefficients from NCI log-linear trend test models for NPC
mortality from cumulative exposure to formaldehyde 5-79
Table 5-9. Extra risk estimates for NPC mortality from various levels of continuous
exposure to formaldehyde 5-80
Table 5-10. ECooos, LECooos, and inhalation unit risk estimates for NPC mortality from
formaldehyde exposure based on the Hauptmann et al. (2004) log-linear
trend analyses for cumulative exposure 5-81
Table 5-11. ECooos, LECooos, and inhalation unit risk estimates for NPC incidence from
formaldehyde exposure based on the Hauptmann et al. (2004) trend
analyses for cumulative exposure 5-82
Table 5-12. Relative risk estimates for mortality from Hodgkin lymphoma
(ICD-8 code 201) and leukemia (ICD-8 codes 204-207) by level of
formaldehyde exposure for different exposure metrics 5-90
Table 5-13. Regression coefficients for Hodgkin lymphoma and leukemia mortality
from NCI trend test models 5-90
Table 5-14. Extra risk estimates for Hodgkin lymphoma mortality from various levels
of continuous exposure to formaldehyde 5-91
This document is a draft for review purposes only and does not constitute Agency policy.
III-xviii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF TABLES (continued)
Table 5-15. Extra risk estimates for leukemia mortality from various levels of
continuous exposure to formaldehyde 5-91
Table 5-16. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin lymphoma
mortality from formaldehyde exposure based on Beane Freeman et al.
(2009) log-linear trend analyses for cumulative exposure 5-93
Table 5-17. ECoos, LECoos, and inhalation unit risk estimates for leukemia mortality
from formaldehyde exposure based on Beane Freeman et al. (2009)
log-linear trend analyses for cumulative exposure 5-93
Table 5-18. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin
lymphoma incidence from formaldehyde exposure, based on Beane
Freeman et al. (2009) log-linear trend analyses for cumulative exposure 5-94
Table 5-19. ECoos, LECoos, and inhalation unit risk estimates for leukemia incidence
from formaldehyde exposure based on Beane Freeman et al. (2009)
log-linear trend analyses for cumulative exposure 5-94
Table 5-20. Calculation of combined cancer mortality unit risk estimate at 0.1 ppm 5-100
Table 5-21. Calculation of combined cancer incidence unit risk estimate at 0.1 ppm 5-100
Table 5-22. Summary of tumor incidence in long-term bioassays on F344 rats 5-105
Table 5-23. BMD modeling of unit risk of SCC in the human respiratory tract 5-125
Table 5-24. Formaldehyde-induced rat tumor incidences 5-128
Table 5-25. Human benchmark extrapolations of nasal tumors in rats using
formaldehyde flux and DPX 5-134
Table 5-26. Summary of inhalation unit risk estimates 5-135
Table 5-27. Total cancer risk from exposure to a constant formaldehyde exposure
level of 1 |ig/m3 from ages 0-70 years 5-137
Table 6-1. Summary of candidate Reference Concentrations (RfC) for co-critical studies.... 6-27
Table 6-2. Effective concentrations (lifetime continuous exposure levels) predicted
for specified extra cancer risk levels for selected formaldehyde-related
cancers 6-36
Table 6-3. Inhalation unit risk estimates based on epidemiological and experimental
animal data 6-42
This document is a draft for review purposes only and does not constitute Agency policy.
III-xix DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES
Figure 2-1. Chemical structure of formaldehyde 2-1
Figure 2-2. Locations of hazardous air pollutant monitors 2-5
Figure 2-3. Modeled ambient air concentrations based on 1999 emissions 2-7
Figure 2-4. Range of formaldehyde air concentrations (ppb) in different environments 2-9
Figure 3-1. Formaldehyde-mediated protein modifications 3-2
Figure 3-2. 3H/14C ratios in macromolecular extracts from rat tissues following exposure
to 14C and 3H-labeled formaldehyde (0.3, 2, 6, 10, 15 ppm) 3-18
Figure 3-3. Formaldehyde clearance by ALDH2 (GSH-independent) and ADH3
(GSH-dependent) 3-20
Figure 3-4. Metabolism of formate 3-22
Figure 3-5. Scatter plot of formaldehyde concentrations measured in ppb in direct breath
exhalations (x axis) and exhaled breath condensate headspace (y axis) 3-31
Figure 3-6. Reconstructed nasal passages of F344 rat, rhesus monkey, and human 3-36
Figure 3-7. Illustration of interspecies differences in airflow and verification of CFD
simulations with water-dye studies 3-37
Figure 3-8. Lateral view of nasal wall mass flux of inhaled formaldehyde simulated in
the F344 rat, rhesus monkey, and human 3-38
Figure 3-9. CFD simulations of formaldehyde flux to human nasal lining at different
inspiratory flow rates 3-39
Figure 3-10. Single-path model simulations of surface flux per ppm of formaldehyde
exposure concentration in an adult male human 3-43
Figure 3-11. Pressure drop vs. volumetric airflow rate predicted by the CUT CFD
model compared with pressure drop measurements made in two hollow
molds (CI and C2) of the rat nasal passage (Cheng et al., 1990) or in rats
in vivo 3-45
Figure 3-12. Formaldehyde-DPX dosimetry in the F344 rat 3-47
Figure 4-1. Delayed asthmatic reaction following the inhalation of formaldehyde after
"painting" 100% formalin for 20 minutes 4-20
This document is a draft for review purposes only and does not constitute Agency policy.
III-xx DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-2. Formaldehyde effects on minute volume in naive and formaldehyde-
pretreated male B6C3F1 mice and F344 rats 4-115
Figure 4-3. Sagittal view of the rat nose (nares oriented to the left) 4-121
Figure 4-4. Main components of the nasal respiratory epithelium 4-122
Figure 4-5. Decreased mucus clearance and ciliary beat in isolated frog palates
exposed to formaldehyde after 3 days in culture 4-126
Figure 4-6. Diagram of nasal passages showing section levels chosen for morphometry
and autoradiography in male rhesus monkeys exposed to formaldehyde 4-135
Figure 4-7. Formaldehyde-induced cell proliferation in male rhesus monkeys exposed to
formaldehyde 4-136
Figure 4-8. Formaldehyde-induced lesions in male rhesus monkeys exposed to formaldehyde
4-137
Figure 4-9. Frequency and location by cross-section level of squamous metaplasia in
the nasal cavity of F344 rats exposed to formaldehyde via inhalation 4-172
Figure 4-10. Effect of formaldehyde exposure on cell proliferation of the respiratory
mucosa of rats and mice 4-190
Figure 4-11. Alveolar MP Fc-mediated phagocytosis from mice exposed to 5 ppm
formaldehyde, 10 mg/m3 carbon black, or both 4-223
Figure 4-12. Compressed air in milliliters as parameter for airway obstruction
following formaldehyde exposure in guinea pigs after OVA sensitization and
OVA challenge 4-235
Figure 4-13. OVA-specific IgGl (IB) in formaldehyde-treated sensitized guinea pigs
prior to OVA challenge 4-235
Figure 4-14. Anti-OVA titers in female Balb/C mice exposed to 6.63 ppm
formaldehyde for 10 consecutive days, or once a week for 7 weeks 4-236
Figure 4-15. Vascular permeability in the tracheae and bronchi of male Wistar rats
after 10 minutes of formaldehyde inhalation 4-238
Figure 4-16. Effect of select receptor antagonists on formaldehyde-induced vascular
permeability in the trachea and bronchi of male Wistar rats 4-239
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-17. The effects of formaldehyde inhalation exposures on eosinophil
infiltration (Panel A) and goblet cell proliferation (Panel B) after Der f
challenge in the nasal mucosa of male ICR mice after sensitization and
challenge 4-241
Figure 4-18. NGF in BAL fluid from formaldehyde-exposed female C3H/He mice
with and without OA sensitization 4-243
Figure 4-19. Plasma Substance P levels in formaldehyde-exposed female C3H/He
mice with and without OVA sensitization 4-244
Figure 4-20. Motor activity in male and female rats 2 hours after exposure to
formaldehyde expressed as mean number of crossed quadrants ± SEM 4-256
Figure 4-21. Habituation of motor activity was observed in control rats during the
second observation period (day 2, 24 hours after formaldehyde exposure) 4-257
Figure 4-22. Motor activity was reduced in male and female LEW. IK rats 2 hours
after termination of 10-minute formaldehyde exposure 4-258
Figure 4-23. The effects of the acute formaldehyde (FA) exposures on the
ambulatory and vertical components of SLMA 4-260
Figure 4-24. Effects of formaldehyde exposure on the error rate of female LEW. IK
rats performing the water labyrinth learning test 4-264
Figure 4-25. Basal and stress-induced trunk blood corticosterone levels in male
LEW. IK rats after formaldehyde inhalation exposures 4-269
Figure 4-26. NGF production in the brains of formaldehyde-exposed mice 4-274
Figure 4-27. Mortality corrected cumulative incidences of nasal carcinomas in the
indicated exposure groups 4-325
Figure 4-28. Leukemia incidence in Sprague-Dawley rats exposed to formaldehyde
in drinking water for 2 years 4-330
Figure 4-29. Unscheduled deaths in female F344 rats exposed to formaldehyde for
24 months 4-332
Figure 4-30. Cumulative leukemia incidence in female F344 rats exposed to
formaldehyde for 24 months 4-333
Figure 4-31. Cumulative incidence or tumor bearing animals for lymphoma in
female mice exposed to formaldehyde for 24 months 4-334
This document is a draft for review purposes only and does not constitute Agency policy.
Ill-xxii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 4-32. DNA-protein cross-links (DPX) and thymidine kinase (tk) mutants in
TK6 human lymphoblasts exposed to formaldehyde for 2 hours 4-357
Figure 4-33. Developmental origins for cancers of the lymphohematopoietic system 4-409
Figure 4-34A. Association between peak formaldehyde exposure and the risk of
lymphohematopoietic malignancy 4-415
Figure 4-34B. Association between average intensity of formaldehyde exposure and
the risk of lymphohematopoietic malignancy 4-416
Figure 4-35. Effect of various doses of formaldehyde on cell number in (A) HT-29
human colon carcinoma cells and in (B) human umbilical vein epithelial cells
(HUVEC) 4-433
Figure 4-36. Integrated MO A scheme for respiratory tract tumors 4-446
Figure 4-37. Location of intra-epithelial lymphocytes along side epithelial cells in
the human adenoid 4-450
Figure 5-1. Change in number of additions made in 10 minutes following formaldehyde
exposure at 32, 170, 390, or 890 ppb 5-21
Figure 5-2. Effects of formaldehyde exposure on the error rate of female LEW. IK rats
performing the water labyrinth learning test 5-24
Figure 5-3. Fecundity density ratio among women exposed to formaldehyde in the high
exposure index category with 8-hour time weighted average formaldehyde
exposure concentration of 219 ppb 5-27
Figure 5-4. Estimated reduction in peak expiratory flow rate (PEFR) in children in
relation to indoor residential formaldehyde concentrations 5-41
Figure 5.5. Odds ratios for physician-diagnosed asthma in children associated with in-
home formaldehyde levels in air 5-45
Figure 5-6. Prevalence of asthma and respiratory symptom scores in children associated
with in-home formaldehyde levels 5-48
Figure 5-7. Prevalence and severity of allergic sensitization in children associated with
in-home formaldehyde levels 5-49
Figure 5-8. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (eye irritation) 5-53
This document is a draft for review purposes only and does not constitute Agency policy.
Ill-xxiii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF FIGURES (continued)
Figure 5-9. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (burning eyes) 5-54
Figure 5-10. Age-specific mortality and incidence rates for myeloid, lymphoid, and
all leukemia 5-98
Figure 5-11. Schematic of integration of pharmacokinetic and pharmacodynamic
components in the CUT model 5-110
Figure 5-12. Fit to the rat tumor incidence data using the model and assumptions in
Conolly etal. (2003) 5-112
Figure 5-13. Spatial distribution of formaldehyde over the nasal lining, as
characterized by partitioning the nasal surface by formaldehyde flux to
the tissue per ppm of exposure concentration, resulting in 20 flux bins 5-120
Figure 5-14. Distribution of cells at risk across flux bins in the F344 rat nasal lining 5-120
Figure 5-15. MLE and upper bound (UB) added risk of SCC in the human nose for
two BBDR models 5-124
Figure 5-16. Replot of log-probit fit of the combined Kerns et al. (1983) and
Monticello et al. (1996) data on tumor incidence showing BMCio and
BMCLio 5-127
Figure 5-17. EPA multistate Weibull modeling: nasal tumor dose response 5-131
Figure 5-18. Multistage Weibull model fit 5-132
Figure 5-19. Multistage Weibull model fit of tumor incidence data compared with
KM estimates of spontaneous tumor incidence 5-132
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxiv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS
ACGIH
American Conference of Governmental Industrial Hygienists
ADAF
age-dependent adjustment factors
ADH
alcohol dehydrogenase
ADS
anterior dorsal septum
AIC
Akaike Information Criterion
AIE
average intensity of exposure
AIHA
American Industrial Hygiene Association
ALB
albumin
ALDH
aldehyde dehydrogenase
ALL
acute lymphocytic leukemia
ALM
anterior lateral meatus
ALP
alkaline phosphatase
ALS
amyotrophic lateral sclerosis
ALT
alanine aminotransferase
AML
acute myelogenous leukemia
AMM
anterior medial maxilloturbinate
AMPase
adenosine monophosphatase
AMS
anterior medial septum
ANAE
alpha-naphthylacetate esterase
ANOVA
analysis of variance
APA
American Psychiatric Association
ARB
Air Resources Board
AST
aspartate aminotransferase
ATCM
airborne toxic control measure
ATP
adenosine triphosphate
ATPase
adenosine triphosphatase
ATS
American Thoracic Society
AT SDR
Agency for Toxic Substances and Disease Registry
AUC
area under the curve
BAL
bronchoalveolar lavage
BALT
bronchus associated lymphoid tissue
BBDR
biologically based dose response
BC
bronchial construction
BCME
bis(chloromethyl)ether
BDNF
brain-derived neurotrophic factor
BEIR
biologic effects of ionizing radiation
B£R
German Federal Institute for Risk Assessment
BHR
bronchial hyperresponsiveness
BMC
benchmark concentration
BMCL
95% lower bound on the benchmark concentration
BMCR
binuclated micronucleated cell ratefluoresce
BMD
benchmark dose
BMDL
95% lower bound on the benchmark dose
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxv DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
BMR benchmark response
BN Brown-Norway
BrdU bromodeoxyuridine
BUN blood urea nitrogen
BW body weight
CA chromosomal aberrations
CalEPA California Environmental Protection Agency
CAP College of American Pathologists
CASRN Chemical Abstracts Service Registry Number
CAT catalase
CBMA cytokinesis-blocked micronucleus assay
CBMN cytokinesis-blocked micronucleus
CDC U.S. Centers for Disease Control and Prevention
CDHS California Department of Health Services
CFD computational fluid dynamics
CGM clonal growth model
CHO Chinese hamster ovary
CI confidence interval
CUT Chemical Industry Institute of Toxicology
CLL chronic lymphocytic leukemia
CML chronic myelogenous leukemia
CNS central nervous system
CO2 carbon dioxide
COEHHA California Office of Environmental Health Hazard Assessment
CREB cyclic AMP responsive element binding proteins
CS conditioned stimulus
C x t concentration times time
DA Daltons
DAF dosimetric adjustment factor
DDX DNA-DNA cross-links
DEI daily exposure index
DEN diethylnitrosamine
Der f common dust mite allergen
DMG dimethylglycine
DMGDH dimethylglycine dehydrogenase
DNA deoxyribonucleic acid
DOPAC 3,4-dihydroxyphenylacetic acid
DPC / DPX DNA-protein cross-links
EBV Epstein-Barr virus
EC effective concentration
ED effective dose
EHC Environmental Health Committee
ELISA enzyme-linked immunosorbent assay
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxvi DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
EPA
U.S. Environmental Protection Agency
ERPG
emergency response planning guideline
ET
ethmoid turbinates
FALDH
formaldehyde dehydrogenase
FDA
U.S. Food and Drug Administration
FDR
fecundability density ratio
FEF
forced expiratory flow
FEMA
Federal Emergency Management Agency
FEV1
forced expiratory volume in 1 second
FISH
fluorescent in situ hybridization
FSH
follicle-stimulating hormone
FVC
forced vital capacity
GALT
gut-associated lymphoid tissue
GC-MS
gas chromatography-mass spectrometry
GD
gestation day
GI
gastrointestinal
GO
gene ontology
G6PDH
glucose-6-phosphate dehydrogenase
GPX
glutathione peroxidase
GR
glutathione reductase
GM-CSF
granulocyte macrophage-colony-stimulating factor
GSH
reduced glutathione
GSNO
S-nitrosoglutathione
GST
glutathione S-transferase
HAP
hazardous air pollutant
Hb
hemoglobin
HC1
hydrochloric acid
HCT
hematocrit
HEC
human equivalent concentration
5-HI A A
5-hydroxyindoleacetic acid
hm
hydroxymethyl
HMGSH
S-hydroxymethylglutathione
HPA
hypothalamic-pituitary adrenal
HPG
hypothalamo-pituitary-gonadal
HPLC
high-performance liquid chromatography
HPRT
hypoxanthine-guanine phosphoribosyl transferase
HR
high responders
HSA
human serum albumin
HSDB
Hazardous Substances Data Bank
Hsp
heat shock protein
HWE
healthy worker effect
I cell
initiated cell
IARC
International Agency for Research on Cancer
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxvii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
ICD
International Classification of Diseases
IF
interfacial
IFN
interferon
Ig
immunoglobulin
IL
interleukin
LP.
intraperitoneal
IPCS
International Programme on Chemical Safety
IRIS
Integrated Risk Information System
Km
Michaels-Menton constant
KM
Kaplan-Meier
LD50
median lethal dose
LDH
lactate dehydrogenase
LEC
95% lower bound on the effective concentration
LED
95% lower bound on the effective dose
LHP
ly mphohematopoi eti c
LI
labeling index
LM
Listeria monocytogenes
LMS
linearized multistage
LLNA
local lymph node assay
LOAEL
lowest-observed-adverse-effect level
LPS
lipopolysaccharide
LR
low responders
LRT
lower respiratory tract
MA
methylamine
MALT
mucus-associated lymph tissues
MCH
mean corpuscular hemoglobin
MCHC
mean corpuscular hemoglobin concentration
MCS
multiple chemical sensitivity
MCV
mean corpuscular volume
MDA
malondialdehyde
MEF
maximal expiratory flow
ML
myeloid leukemia
MLE
maximum likelihood estimate
MMS
methyl methane sulfonate
MMT
medial maxilloturbinate
MN
micronucleus, micronuclei
MNNG
N-methyl -N-nitro-N-nitrosoguani dine
MOA
mode of action
MoDC
monocyte-derived dendritic cell
MP
macrophage
MPD
multistage polynomial degree
MPS
mononuclear phagocyte system
MRL
minimum risk level
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxviii DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
mRNA messenger ribonucleic acid
MVE-2 Murray Valley encephalitis virus
MVK Moolgavkar, Venzon, and Knudson
N cell normal cell
NaCl sodium choride
NAD+ nicotinamide adenine dinucleotide
NADH reduced nicotinamide adenine dinucleotide
NALT nasally associated lymphoid tissue
NATA National-Scale Air Toxics Assessment
NCEA National Center for Environmental Assessment
NCHS National Center for Health Statistics
NCI National Cancer Institute
NEG Nordic Expert Group
NER nucleotide excision repair
NGF nerve growth factor
NHL non-Hodgkin's lymphoma
NHMRC/ARMCANZ National Health and Medical Research Council/Agriculture and Resource
Management Council of Australia and New Zealand
NNK nitrosamine nitrosamine 4-(methylnitrosamino)-l-(3-pyridyl)-butanone
N6-hmdA N6-hydroxymethyldeoxyadenosine
N4-hmdC N4-hydroxymethylcytidine
N2-hmdG N2-hydroxymethyldeoxyguanosine
NICNAS National Industrial Chemicals Notification and Assessment Scheme
NIOSH National Institute for Occupational Safety and Health
NLM National Library of Medicine
NMDA N-methyl-D-aspartate
NO nitric oxide
NOAEL no-ob served-adverse-effect level
NPC nasopharyngeal cancer
NRBA neutrophil respiratory burst activity
NRC National Research Council
NTP National Toxicology Program
OR odds ratio
OSHA Occupational Safety and Health Administration
OTS Office of Toxic Substances
OVA ovalbumin
PBPK physiologically based pharmacokinetic
PC Philadelphia chromosome
PCA passive cutaneous anaphylaxis
PCMR proportionate cancer mortality ratio
PCNA proliferating cell nuclear antigen
PCR polymerase chain reaction
PCV packed cell volume
This document is a draft for review purposes only and does not constitute Agency policy.
Ill-xxix DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
PEC AM
platelet endothelial cell adhesion molecule
PEF
peak expiratory flow
PEFR
peak expiratory flow rates
PEL
permissible exposure limit
PFC
plaque-forming cell
PG
peri glomerular
PHA
phytohemagglutinin
PLA2
phospholipase A2
PI
phagocytic index
PLM
posterior lateral meatus
PMA
phorbol 12-myristate 13-acetate
PMR
proportionate mortality ratio
PMS
posterior medial septum
PND
postnatal day
POD
point of departure
POE
portal of entry
PTZ
pentilenetetrazole
PUFA
polyunsaturated fatty acids
PWULLI
population weighted unit length labeling index
RA
reflex apnea
RANTES
regulated upon activation, normal T-cell expressed and secreted
RB
reflex bradypnea
RBC
red blood cells
RD50
exposure concentration that results in a 50% reduction in respiratory rate
REL
recommended exposure limit
RfC
reference concentration
RfD
reference dose
RGD
regional gas dose
RGDR
regional gas dose ratio
RR
relative risk
RT
reverse transcriptase
SAB
Science Advisory Board
see
squamous cell carcinoma
SCE
sister chromatid exchange
SCG
sodium cromoglycate
SD
standard deviation
SDH
succinate dehydrogenase; sarcosine dehydrogenase
SEER
Surveillance, Epidemiology, and End Results
SEM
standard error of the mean
SEN
sensitizer
SH
sulfhydryl
SHE
Syrian hamster embryo
SLMA
spontaneous locomotor activity
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxx DRAFT—DO NOT CITE OR QUOTE
-------
LIST OF ABBREVIATIONS AND ACRONYMS (continued)
SMR
standardized mortality ratio
SNP
single nucleotide polymorphism
SOD
superoxide dismutase
SOMedA
N6-sulfomethyldeoxy adenosine
Spl
specificity protein
SPIR
standardized proportionate incidence ratio
SSAO
semicarbozole-sensitive amine oxidase
SSB
single strand breaks
STEL
short-term exposure limit
TBA
tumor bearing animal
TH
T-lymphocyte helper
THF
tetrahydrofolate
TK
toxicokinetics
TL
tail length
TLV
threshold limit value
TNF
tumor necrosis factor
TP
total protein
TRI
Toxic Release Inventory
TRPV
transient receptor potential vanilloid
TWA
time-weighted average
TZCA
thiazolidine-4-carboxylate
UCL
upper confidence limit
UDS
unscheduled DNA synthesis
UF
uncertainty factor
UFFI
urea formaldehyde foam insulation
ULLI
unit length labeling index
URT
upper respiratory tract
USD A
U.S. Department of Agriculture
VC
vital capacity
VOC
volatile organic compound
WBC
white blood cell
WDS
wet dog shake
WHO
World Health Organization
WHOROE
World Health Organization Regional Office for Europe
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxxi DRAFT—DO NOT CITE OR QUOTE
-------
This page intentionally left blank.
This document is a draft for review purposes only and does not constitute Agency policy.
III-xxxii DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5. QUANTITATIVE ASSESSMENT: INHALATION EXPOSURE
This chapter presents the quantitative assessments conducted by EPA for both cancer and
noncancer health effects associated with formaldehyde exposure. The quantitative assessment is
focused on the inhalation route of exposure. The current IRIS reference dose (RfD) is not
reevaluated in this assessment. Although there is some evidence of formaldehyde
carcinogenicity via the oral route of exposure, these data are not evaluated herein nor is an oral
slope factor considered at this time. Therefore, the following sections address derivation of a
reference concentration (RfC) and cancer unit risk estimate for inhalation exposures.
For noncancer effects, the RfC is an estimate (with uncertainty spanning perhaps an order
of magnitude) of a continuous inhalation exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.
It can be derived from a NOAEL, LOAEL, or benchmark concentration, with uncertainty factors
generally applied to reflect limitations of the data used. Data from the previous chapters are
evaluated to determine the health effects associated with formaldehyde exposure and which
studies may best inform the exposure response relationship for RfC derivation. Section 5.1
summarizes the observed noncancer health effects, selecting key studies and critical effects for
consideration. Candidate RfCs are derived for each identified key study. Several alternatives are
considered for uncertainty factors addressing human variability for key studies and alternatives
presented (Section 5.1.2.3). Options for addressing the overall database uncertainty factor are
provided which may modify the final RfC (Section 5.1.3).
The derivation of the cancer inhalation unit risk estimate considered data regarding both
respiratory tract cancers and lymphohematopoietic malignancies. Exposure-response modeling
from epidemiologic studies was used to derive a combined unit risk estimate for nasopharyngeal
cancer and lymphohematopoietic cancers (Section 5.2). This unit risk estimate is supported by
an analysis of exposure-response modeling of respiratory tract cancer risk using data from
experimental animal studies (Section 5.3). Analysis of the animal bioassays includes an
evaluation of a published biologically based dose-response model as well as an appraisal of
published dose-response modeling of genomics data and a presentation of benchmark dose
modeling approaches. Finally, Section 5.4 provides a summary and conclusions from the cancer
exposure-response modeling, presenting the final unit risk estimate based on the combined risk
of nasopharyngeal cancer and lymphohematopoietic cancers observed in the human studies.
This document is a draft for review purposes only and does not constitute Agency policy.
5_1 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5.1. INHALATION REFERENCE CONCENTRATION (RfC)
Prior to the current assessment, the EPA IRIS file for formaldehyde did not provide an
inhalation RfC. As presented in the hazard identification in Chapter 4, a number of noncancer
health effects are associated with formaldehyde exposure. Section 5.1.1 describes each of the
health effect categories considered for RfC derivation and the specific endpoints considered for
each category. The identified effect categories are: sensory irritation (eye, nose, and throat);
upper respiratory tract (URT) pathology; pulmonary function; increased asthma and atopic
sensitization; altered immune function; neurotoxicity and reproductive and developmental
toxicity. For each health effect category,, studies that may adequately inform the exposure-
response relationship for specific critical effects are identified for consideration in RfC
derivation.
EPA employed a screening process across the different health effect categories to select
key studies that would best support the derivation of an inhalation RfC (as described in
Section 5.1.2.1). The following factors were considered in this evaluation: characteristics of the
study population, exposure regimen, quality of exposure assessment, quality of exposure-
response assessment, exposure levels at which effects were seen and statistical power of the
study. Based on this analysis, seven studies were considered for RfC derivation. Candidate RfC
derivation from a key study includes the following steps: 1) define the critical effect(s); 2)
determine appropriate point(s) of departure (PODs) on the basis of inhaled concentration; 3)
adjust each POD by endpoint/study-specific uncertainty factors (UFs), to account for
uncertainties in the extrapolation of study results to conditions of human environmental
exposure. All of the identified key studies were human studies and several studies included
potentially susceptible individuals (e.g. children, asthmatics). The uncertainty factor for human
variability has sometimes been reduced for studies of susceptible populations or lifestages.
However, for five of the seven key studies it was unclear if an uncertainty factor of 3 or 1 for
human variability was most appropriate. Therefore, alternatives are presented for consideration.
Candidate RfCs (cRfCs) are derived for sensory irritation, decreased pulmonary function in
children, increased asthma incidence in children, increased allergic sensitization to common
allergens in children, and decreased fecundability density ratio (FDR) in women (increased time
to pregnancy) (Table 5-7). All of these cRfCs are derived from endpoints identified in
residential studies, with the exception of decreased FDR (observed in an occupational study of
women in the woodworking industry).
The overall literature database of both human and laboratory animal studies examining
the health effects from formaldehyde exposure is large; however, the available studies for some
types of effects are limited. Limitations in the existing database are discussed in Section 5.1.3,
This document is a draft for review purposes only and does not constitute Agency policy.
5_2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
specifically regarding understanding the reproductive and developmental effects and the
exposure-response relationship for the observed neurological and behavioral effects from
formaldehyde exposure. EPA considers 3 options for addressing these database uncertainties in
the final RfC: (1) providing an RfC derived from studies of respiratory and allergenic responses
and protective of sensory irritation effects, without further adjustment for uncertainties in the
database (noting the need for further research to elucidate reproductive, developmental and
neurotoxic effects); (2) providing an RfC with a database uncertainty factor incorporated to
reflect the potential that reproductive, developmental, or neurotoxic effects might occur at lower
doses; or (3) provide a range for the RfC which encompasses the above two options for the
database uncertainty factor.
5.1.1. Candidate Critical Effects by Health Effect Category
The following subsections describe the best available studies and endpoints for
quantitative RfC derivation within each health effect category. These studies are considered
representative of the health effects attributed to formaldehyde exposure. For more details on
specific studies discussed here, see Sections 4.1.1 and 4.2.1. The identified health effect
categories are: sensory irritation (eye, nose, and throat); upper respiratory tract (URT) pathology;
pulmonary function; increased asthma and allergic sensitization; altered immune function;
neurotoxicity and reproductive and developmental toxicity. Discussions in each subsection
below describe the various health effects observed in human and animal studies for each
category.
For each health effect category, specific studies that may adequately inform the exposure-
response relationship for critical effects are identified for consideration in RfC derivation. In
general, studies are included where study quality and ability to define exposures are considered
adequate for RfC derivation. Whenever possible, greater consideration is typically given to
human data for derivation of an RfC. When laboratory studies conducted in rodents are
considered for RfC derivation, the potential confounding effects of formaldehyde-induced reflex
bradypnea (RB) are evaluated (Section 4.1.1 for a discussion of RB). If the exposure levels are
expected to cause RB, results are evaluated to ensure the effects are not in part attributable to
primary or secondary effects of RB in the rodents.
5.1.1.1. Sensory Irritation of the Eyes, Nose, and Throat
Eye, nose, and throat irritation are common effects of chemically induced sensory
irritation; specific effects include lacrimation, burning of the eyes and nose, rhinitis, burning of
the throat, and cough (Feron et al., 2001). Chemical irritants such as formaldehyde bind to
This document is a draft for review purposes only and does not constitute Agency policy.
5.3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
protein receptors of the trigeminal nerve, triggering a burning and painful sensation. This
process is distinct from taste and smell (Cometto-Muniz and Cain, 1992; Nielsen, 1991). The
trigeminal nerve has three branches (ophthalmic, maxillary, and mandibular) and not only acts as
an afferent nerve relaying these sensations to the central nervous system but has efferent nerve
activity as well (Meggs, 1993). Stimulation of the trigeminal nerve may result in reflex
responses, including lacrimation, coughing, and sneezing. In this assessment, both the reflex
responses and the sensations (such as burning, pain, and itching of the eyes, nose, and throat) are
considered adverse effects (see Section 4.1 for a full discussion of available human data).
There are studies noting irritant effects in rodents (Sarsilmaz et al., 1999; Holmstrom et
al., 1989; Dubreuil et al., 1976) and monkeys (Monticello et al., 1989; Rusch et al., 1983).
These animal studies are supportive of the health effects reported in humans. However, given
the uncertainties in extrapolation from responses in laboratory animals to expected responses in
humans, the available human studies are preferred.
In human studies, the endpoints for assessing irritation include subjective self reporting
of symptoms (e.g., pain, burning, itching, increased cough) via questionnaires or objective
measures of irritation that can be assessed during controlled acute exposures (e.g., eye-blink
counts, lacrimation). Several acute chamber studies support development of a concentration-
response relationship for sensory irritation, identifying an effect level for various exposure
durations (Kulle, 1993; Andersen and Mjzflhave, 1983; Bender et al., 1983; Weber-Tschopp et al.,
1977). Arts et al. (2006b) reviewed several studies and performed BMD analyses, reporting
10% extra risk BMCL values for reported eye discomfort of 560 and 240 ppb for 3 and 5 hour
exposures, respectively. LOAELs of 1,000 ppb and 1,700 ppb were reported for 1-2 minute
exposures (Bender et al., 1983; Weber-Tschopp et al., 1977). These acute studies support a role
for both concentration and duration in the effect level for eye irritation. Although exposure
concentrations are well-defined in these chamber studies, the chamber studies are not appropriate
for RfC derivation because they are of acute duration and the exposure levels used are much
higher than those reported for chronic exposure scenarios, both occupational and residential.
A study of industrial workers assessed sensory irritation and provided an average
exposure derived from in-plant exposure measurements and the work history of each study
participant (Holmstrom and Wilhelmsson, 1988). Although average daily exposures were
estimated for each employee, these data were not used to explore an exposure-response
relationship within the worker cohort. The symptom prevalence for sensory irritation (e.g., nasal
discomfort, eye discomfort, and airway discomfort) relative to the referent group was reported
for the cohort as a whole, where worker exposure ranged from 0.05 to 0.5 mg/m3 formaldehyde
8-hr time-weighted average (TWA), with a mean of 0.26 mg/m3 (210 ppb). The daily TWA does
This document is a draft for review purposes only and does not constitute Agency policy.
5.4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
not reflect the peak exposures experienced during specific work tasks. Although this study
demonstrated marked increases in symptoms of sensory irritation in the workplace due to
formaldehyde exposure, it provided little data to inform the exposure-response relationship,
especially in the range of environmental exposures.
There are three studies that report sensory irritation in humans from chronic exposures in
a residential environment and provide sufficient exposure data to support quantitative assessment
(Liu et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). Each study reports site-
specific exposure measurements and presents some metric of individual exposure. These
residential studies employ in-home measurements for each study participant, either as average
exposure level (Ritchie and Lehnen, 1987; Hanrahan et al., 1984) or as calculated cumulative
exposure based on the time in the home (Liu et al., 1991). Eye irritation is reported at similar
levels of residential formaldehyde exposure in the three studies (Figures 5-8 and 5-9). Each
study provides an exposure-response relationship for prevalence of sensory irritation in relation
to in-home formaldehyde exposure based on individual level data. The detailed exposure
information and chronic nature of the exposures support the selection of these studies as
potential priniciple studies for RfC derivation. Each of these studies is further evaluated and a
cRfC developed for consideration (Section 5.1.2).
5.1.1.2. Upper Respiratory Tract Pathology
Formaldehyde-induced respiratory tract pathology includes inflammation, rhinitis, goblet
cell hyperplasia, metaplastic changes, squamous cell hyperplasia, and impaired mucociliary
transport. A series of laboratory animal studies assessing formaldehyde-induced changes in the
nasal mucosa suggests that these changes may be a protective or adaptive response and that
increased mucus flow and metaplastic changes will progress in relation to the concentration and
duration of exposure to protect the underlying tissue (Swenberg et al., 1983). The degree of
inflammation, hyperplasia, and metaplastic change that is due to sensory irritation-induced
inflammatory responses versus inflammation and tissue remodeling from formaldehyde-induced
direct cell damage cannot be distinguished. These changes have been noted as sensitive
indicators of formaldehyde-induced effects, occurring before gross cellular damage and focal
lesions (Monticello et al., 1989). These responses are considered for RfC derivation, especially
for exposure concentrations where gross damage of the underlying tissue is not expected.
Although well-documented studies demonstrating formaldehyde-induced upper respiratory tract
(URT) pathology have been performed in laboratory animals, including the rat (Zwart et al.,
1988; Woutersen et al., 1987; Morgan et al., 1986a, b, 1983; Swenberg et al., 1986, 1983) and
This document is a draft for review purposes only and does not constitute Agency policy.
5.5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
monkey (Rusch et al., 1983), robust human data are available, and these human data are
preferred for RfC derivation.
Six human studies examined the effects of formaldehyde exposure on URT pathology
(Pazdrak et al., 1993; Boysen et al., 1990; Holmstrom et al., 1989; Edling et al., 1988;
Holmstrom and Wilhelmsson, 1988; Andersen and Mjzflhave, 1983). Of these studies,
Holmstrom and Wilhelmsson (1988) and Holmstrom et al. (1989) were identified as the most
robust and sensitive and are included as candidate studies for RfC derivation (see Table 5-1).
Both studies address the same cohort and, thus, were considered together. The Holmstrom and
Wilhelmsson (1988) study is discussed above under sensory irritation effects. In this study of 70
factory workers exposed to a TWA formaldehyde concentration of 210 ppb, impaired
mucociliary clearance was reported in 20% of the exposed workers and 3% of the 36 nonexposed
workers. Using rhinomanometry, Holmstrom and Wilhelmsson (1988) also found an increase in
nasal resistance due to mucosal swelling, though this increase was not statistically significant. In
Holmstrom et al. (1989), nasal biopsy samples were collected from 62 of the 70 formaldehyde-
exposed factory workers (these 62 had been exposed to a TWA formaldehyde concentration of
240 ppb) and also from 32 of the nonexposed workers. A pathologist scored each sample by
using a scale of 0 (normal respiratory epithelium) to 8 (carcinoma). Biopsy scores for both the
exposed and control groups ranged from 0 (normal respiratory epithelium) to 4 (stratified
squamous epithelium with marked horny layer). The mean scores for the two groups—2.16 for
the formaldehyde-exposed workers and 1.56 for the unexposed workers—however, the
difference was statistically significant and the authors reported that the loss of cilia, goblet cell
hyperplasia, and the incidence of cuboidal and squamous cell metaplasia replacing the columnar
epithelium were more frequent in the group exposed to formaldehyde. There was no correlation
between the duration of exposure and histologic changes or between smoking habits and biopsy
scores. The URT effects, taken together (decreased mucous flow, increased inflammation,
decreased nasal flow, and degradation of the respiratory epithelium), demonstrate a range of
formaldehyde-induced URT pathology consistent with effects observed in controlled animal
studies.
5.1.1.3. Pulmonary Function Effects
The potential effects of formaldehyde exposure on pulmonary function in humans can be
examined on several time-scales of interest. There are reports examining effects from acute
exposures among naively exposed anatomy graduate students (Kriebel et al., 1993; 2001),
anatomy graduate students with several weeks of episodic exposure (Kriebel et al., 1993), as
well as post-shift versus pre-shift differences in pulmonary function in workers with regular
This document is a draft for review purposes only and does not constitute Agency policy.
5_6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
occupational exposure (Malaka and Kodama, 1990; Herbert et al., 1994; Alexandersson et al.,
1982; Alexandersson and Hedenstierna, 1989). Depending on whether the exposures are naive,
the epidemiologic studies that assessed the pulmonary effects of acute exposures to
formaldehyde may be assessing different biological responses, namely, the acute effect alone or
the acute effect(s) in people who may have already been sensitized to formaldehyde effects.
Pulmonary effects of acute formaldehyde exposure have been studied in both healthy
volunteers and sensitive populations under controlled conditions (e.g. acute chamber studies).
Although acute chamber studies have the advantage of measured controlled exposures, other
factors can limit the usefulness of the studies for RfC derivation including: acute duration, small
study populations and lack of statistical power to assess the measured parameters. The acute
chamber studies are more fully evaluated in Section 4.1.1 and will not be further considered here
for RfC derivation.
The observed effects in the previously unexposed anatomy students provide additional
information on acute exposures in two naive populations (Kriebel et al., 1993; 2001), as well as
insight into the intermediate stages of possible sensitization (Kriebel et al., 1993). Kriebel and
colleagues (1993) examined the pre-laboratory and post-laboratory peak expiratory flow (PEF)
in students attending anatomy classes once per week. They found the strongest pulmonary
response when examining the average cross-laboratory decrement in peak expiratory flow in the
first 2 weeks of the study when formaldehyde concentrations collected in the breathing zones
had a geometric average concentration of 0.73 ppm. Overall, the students exhibited a 2%
decrement in PEF, while the students with any history of asthma showed a 7.3% decrement in
PEF. These findings of acute decreases in PEF following students' initial anatomy sessions were
corroborated by the Kriebel et al. (2001) study, which used a similar study design applied to
another class of anatomy students.
The first Kriebel et al. (1993) study also shows how the acute effects of formaldehyde
exposure were altered following several weeks of weekly episodic exposure. By the 5th week of
class, the pre- and post-laboratory measurements of PEF were no longer reflecting a clearly
demonstrated acute effect but following the 7th week of episodic exposure, both pre-and post-
laboratory PEF continued to drop steadily until the class adjourned after 10 weeks time. While
the acute effects of formaldehyde exposure appeared to diminish after several weeks of
exposure, the intermediate effect across 9 weeks was a 24 liter/minute drop in PEF that was
statistically significant (P<0.01) after statistical control for random person effects, asthma, an
interaction between time and asthma and eye and nose symptoms of irritation. The Kriebel et al.
(1993) study is considered of sufficient quality to support an acute RfC but the quantitative
details on the initial acute effects among the naively exposed students are not adequately
This document is a draft for review purposes only and does not constitute Agency policy.
5.7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
provided. The findings of the Kriebel et al. (2001) study were confounded by decreased class
attendance, which dropped from 37 in the first week to 20 in week 6 and to just 10 students by
week 10. While the Kriebel et al. (2001) study could be useful as a supportive study for naively
exposed students, the longitudinal component is not strong enough to support RfC development.
Several studies of workers assess both cross-shift and chrinic effects of formaldehyde
exposure (Malaka and Kodama, 1990; Herbert et al., 1994; Alexandersson et al., 1982;
Alexandersson and Hedenstierna, 1989). Since formaldehyde exposure may have cumulative
effects over chronic exposures, occupational studies generally showed clinically small but
statistically significant decrements in pulmonary function across shifts. In general these studies
did not identify, have information on or have appropriate statistical control of, potential
confounding co-exposures. While these occupational studies provide evidence that is clearly and
consistently supportive of an acute effect on pulmonary function, they do not directly support
RfC development of an acute effect divorced of the concomitant chronic effects.
Several studies allowed for the examination of potential chronic effects of formaldehyde
exposure. These included an occupational study (Malaka and Kodama, 1990) that reported pre-
shift pulmonary function as a percentage of expected among the formaldehyde exposed
compared to comparable people not exposed to formaldehyde. Studies that did not report pre-
shift pulmonary function as a percentage of expected function (Herbert et al., 1994;
Alexandersson et al., 1982; Alexandersson and Hedenstierna, 1989) contribute less to an
assessment of potential chronic effects because, post-hoc, it is difficult to calibrate for cross-
study comparison the multiple pulmonary function data without knowledge of the age, gender,
smoking status, height, year of birth, etc. that are important determinants of the pulmonary
function metrics of concern. The single study (Malaka and Kodama, 1990) that did report
functional measures in relation to expected value, found that an average 8-hour time weighted
average formaldehyde exposure of 1.13 ppm from area samples was associated with statistically
significant decrements in FEVi, FEVi/FVC and FEF25-75 compared to a referent population. The
strongest response was for FEF25.75. which showed a 12% drop in observed function compared to
expected function in the unexposed, but it is unclear how to interpret the potential chronic health
effect(s) with just the magnitude of the decrement and the length of the average occupational
tenure at this plywood facility (6.5 years), which was not reported by exposure status.
One study reported on the longitudinal follow-up of workers exposed to formaldehyde
(Alexandersson and Hedenstierna, 1989). This investigation not only examined the acute effects
of exposure across shift, but was able to do so among some of the same workers that had been
studied five years earlier (Alexandersson et al., 1982). Statistically significant decreases in
FEVi/FVC and FEF25.75 were noted over the intervening five years in non-smokers after
This document is a draft for review purposes only and does not constitute Agency policy.
5.8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
correction for normal aging. The decrease in FEF25-75 was 0.212 liters/s (SD=0.066 liters/s) for
each year of exposure and was highly significant (p<0.01). For comparison with the 12% drop in
the same pulmonary metric reported by Malaka and Kodama (1990) over an estimated 6.5 years,
EPA computed the extrapolated percentage decrease in FEF25-75 for the Alexandersson and
Hedenstierna (1989) using the reported yearly decrement applied to the pre-shift values at the
time of the initial study period. EPA calculated that from the predicted value of 4.57 liters/s, a
decrease of 0.168 liters/s could be estimated for each year of exposure regardless of smoking
status. For 6.5 years of exposure, this would result in a 24% drop in FEF25-75- Formaldehyde
concentrations were estimated at 0.42 ppm in the first Alexandersson et al. (1982) study and at
0.50 ppm in the second study, but without better exposure measures, the results of the
longitudinal follow-up cannot support quantitative RfC development.
Information is lacking in these studies such as length or tenure of employment associated
with the pre-shift pulmonary function or how long the residents had lived in their homes.
Likewise, knowledge of how occupational or residential exposure may have changed over time
would have allowed for an examination of the progression of any decrement in function
associated with long-term episodic exposure. Among these studies, the best designed and
executed of the cross-sectional studies was that of Kryzanowski and colleagues (1990).
Municipal employees and their children (613 adults and 298 children) were randomly sampled
and were considered to be representative of a diverse local population. Residential exposures to
formaldehyde were based on repeated samples from each individual's kitchen, living area and
bedroom. The average formaldehyde concentration was 26 ppb, with a maximum sample value
of 140 ppb. The majority of subjects (83%) lived in homes with 2-week average concentrations
below 40 ppb. Subjects' peak expiratory flow rates (PEFR) were determined 4 times daily in the
morning, at noon, in the early evening and before bed for 2 weeks. A statistically significant
linear relationship between increased formaldehyde exposure and decreased peak expiratory
flow rate was reported in children but not adults. All statistical models controlled for
socioeconomic status, tobacco smoking (current active or environmental tobacco smoking) and
nitrogen dioxide concentrations. In children, formaldehyde concentrations of 60-140 ppb
increased the prevalence of physician-diagnosed asthma and bronchitis. Among adults, there was
a statistically significant non-linear relationship with decreased morning PEFR for formaldehyde
concentration <40 ppb. Nonetheless, this strong study had only minor weaknesses such as
measurement error and the fact that it was a cross-sectional study. However, random
measurement error tends to attenuate any true effect and is unlikely to have produced a spurious
effect. It is unlikely that these findings were the product of unmeasured or residual confounding
as the analyses controlled for smoking as well as nitrogen dioxide levels and there is no evidence
This document is a draft for review purposes only and does not constitute Agency policy.
5.9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
of alternative factors that were correlated with formaldehyde concentrations and more strongly
associated with pulmonary function. This study of a large and representative sample from a
diverse study population with a well-quantified concentration-response function and is further
considered for RfC derivation.
5.1.1.4. Asthma and Allergic Sensitization (Atopy)
Sensitization to inhalational chemical exposure may manifest as an allergic or asthmatic
response that is characterized by bronchial constriction (BC) or bronchial hyperresponsiveness
(BHR). This sensitization may be a result of immune involvement, as in the case of
hypersensitivity, or a neurogenic sensitization, where a chemical may directly stimulate
inflammation. Asthma is a specific manifestation of IgE-mediated hypersensitivity,
characterized by BHR and airway inflammation, resulting in lower airway obstruction (Fireman,
2003; Kuby, 1991).
A variety of hypersensitivity reactions have been reported following exposure to
formaldehyde. Rashes and skin reactions have been reported in some individuals after dermal
exposures to formaldehyde. Increased expression of Th-2 cytokines in the lymph nodes of mice
given dermal applications of formaldehyde indicates the involvement of an immune component
to the observed sensitization (Dearman et al., 2005; Hilton et al., 1998; Arts et al., 1997).
However, the response does not appear to be IgE mediated (Arts et al., 1997; Lee et al., 1984).
Gorski et al. (1992) observed an increase in formaldehyde-mediated neutrophil burst in
dermatitis patients exposed in a controlled chamber study and suggests a putative role of
oxidative stress and reactive oxygen species (ROS).
Epidemiologic Studies:
Inhalation exposure has been associated with increased asthmatic responses in asthmatics
in occupational settings. While few available case reports of bronchial asthma suggest direct
respiratory tract sensitization to formaldehyde gas (Lemiere et al., 1995; Burge et al., 1985;
Hendrick et al., 1982; Hendrick and Lane, 1977, 1975), a greater body of human data provides
evidence of an association between formaldehyde exposure and exacerbation of asthmatic
responses in compromised individuals (Kriebel et al., 1993) and particularly in children
(Rumchev et al., 2002; Garrett et al., 1999; Krzyzanowski et al., 1990). Increased asthma
incidence reported after inhalation exposure to formaldehyde led to a NOAEL of 30 ppb
(Rumchev et al., 2002). An increased frequency of respiratory symptoms associated with
asthmatic responses and formaldehyde exposure led to a LOAEL of 30 ppb (Garrett et al., 1999).
This document is a draft for review purposes only and does not constitute Agency policy.
5-10 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
The association between formaldehyde and asthma has been studied by examining
occupational exposures (Fransman et al., 2003; Malaka and Kodama, 1990), school-related
exposures (Zhao et al., 2008; Smedje and Norback, 2001; Norback et al., 2000) and residential
exposures (Matsunaga et al., 2008; Tavernier et al., 2006; Gee et al., 2005; Delfino et al., 2003;
Rumchev et al., 2002; Garrett et al., 1999; Palczynski et al., 1999; Norback et al., 1995;
Krzyzanowski et al., 1990). The two occupational studies examined the respiratory health of
plywood workers (Fransman et al., 2003; Malaka and Kodama, 1990). The most recent of these
was conducted in New Zealand by Fransman et al. (2003). Personal samples of formaldehyde
exposure were taken. The mean level of exposure was 0.08 mg/m3 (65 ppb) and the majority of
samples were below the limit of detection which was reported to be 0.03 mg/m3 (24 ppb).
Compared with those with low levels of formaldehyde exposure, workers with high levels of
exposure were more likely to report having asthma (OR=4.3 [95% CI]: 0.7-27.7]). The
association was not seen when examining formaldehyde exposure and use of asthma medication.
The second study of plywood workers was completed in Indonesia. Background levels of
formaldehyde ranged from 0.003 to 0.07 ppm. The highest concentration of formaldehyde
detected in an air sample was in the particleboard unit (range 1.16 to 3.48 ppm). The occurrence
of asthma was found to be positively associated with formaldehyde exposure, where asthma was
defined as, "Have you ever had an attack of wheezing that made you feel short of breath?",
(Malaka and Kodama, 1990).
Studies of exposure to formaldehyde at schools have been performed in China (Zhao et
al., 2008) and in Sweden (Smedje and Norback, 2001). In the study from China (Zhao et al.,
2008), mean levels of formaldehyde were reported to be 2.3 |ig/m3 (range 1.0-5.0 |ig/m3)
indoors and 5.8 |ig/m3 (range 5.0-7.0 |ig/m3) outdoors. Cumulative asthma (i.e., physician-
diagnosed asthma since birth) and daytime attacks of breathlessness were found to be associated
with outdoor formaldehyde levels. Neither of these outcomes was associated with indoor
concentrations of formaldehyde; however, indoor levels were found to be associated with
nocturnal attacks of breathlessness. In Sweden (Smedje and Norback, 2001), the levels of
formaldehyde measured indoors were higher (mean 4, range <5.0-72 |ig/m3). One difference
between the Swedish study and the study conducted in China is that the Swedish study examined
the incidence of asthma over a 4-year period and did not report an association between
formaldehyde exposure and the incidence of asthma (OR 1.2 [95% CI: 0.8-1.7]) among the
whole study population. However, when the investigators stratified based on history of atopy,
they reported that among children without a history of atopy, a new diagnosis of asthma was
significantly more likely at higher concentrations of formaldehyde (OR 1.7 per 10 |ig /m3 [95%
CI: 1.1-2.6]) and at higher total concentrations of mold (OR=4.7 per 10-fold increased in total
This document is a draft for review purposes only and does not constitute Agency policy.
5_11 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
molds [95% CI: 1.2-18.4] in the classroom air. The finding in increase health effects due to
formaldehyde and mold exposures did not appear to control for the other exposure and no
information on the potential correlation between the two exposures was provided. In order to
evaluate the potential for confounding of the reported formaldehyde association by the reported
mold association, the magnitude of effects must be compared on a appropriate scale since the
magnitude of an odds ratio depends on the magnitude of the change in exposure level that is
expected to produce increased risk. After standardizing the units to the reported geometric mean
standard deviation, the results for formaldehyde (GSM=2.3 |ig /m3) is ORl=1.13 per GSD and
the results for mold is OR2=1.02 for a comparison of risks at the GSM to 10*GSM and
OR3=1.06 for a comparison of risks at the minimum value of total molds (5*103/m3) to
10*minimum. As it appears that the magnitude of the formaldehyde effect is stronger than that
of the mold effect (following standardization of exposure increment), it can be concluded that the
reported formaldehyde effect could not have been due to uncontrolled confounding by mold.
The results of studies measuring residential exposure to formaldehyde and asthma are
varied, with some demonstrating an association and others finding no relationship. A recent
study (Matsunaga et al., 2008) found no association between 24-hour formaldehyde and
prevalence of asthma when pregnant women with an exposure to >47 ppb were compared to
those with exposure to <18 ppb. However, they reported an increased risk of atopic eczema.
This study did not assess the risk of incident asthma. A study utilizing self-reported asthma
prevalence as an outcome also found no association with levels of formaldehyde (mean
25.9 |ig/m3, range 2.0-66.8 |ig/m3) (Palczynski et al., 1999), although they noted the incidence
of allergic diseases was greatest in the highest formaldehyde exposure group but that the groups
were too small for statistical evaluation.
A study performed by Tuthill (1984) measured formaldehyde exposure for children
grades K through 6 by using a combination of proxy variables. Overall, there was no
association, but some individual variables showed an increased risk. For example, the reported
risk ratio for having new construction or remodeling performed in the house in the past 4 months
was 2.5 (95% CI: 1.7-3.9). The risk ratio for having new or upholstered furniture in the house
(within the past 4 months) was 2.2 (95% CI: 1.2-3.9).
The study by Delfino et al. (2003) assessed whether the ambient formaldehyde
concentration measured at a central monitoring site was associated with asthma symptoms. The
study examined 22 10-15 year olds with at least 1 year of physician-diagnosed asthma and living
1 OR per GSD=exp[ln(OR per |ig /m3)/10 |ig /m3 * 2.3 |ig /m3]=exp[ln(1.7)/10*2.3]=1.13
2 OR per GSD=exp[ln(OR per 10-fold increase)/ (9*GSM)*1.6 |ig /m3]=exp[ln(4.7)/162*1.6]=1.02
3 OR per GSD=exp[ln(OR per 10-fold increase)/ (9*Minimum)*1.6 |ig /m3]=exp[ln(4.7)/45*1.6]=1.06
This document is a draft for review purposes only and does not constitute Agency policy.
5-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
in a nonsmoking household. The mean levels of formaldehyde were measured to be 7.21 ppb
(range 4.27-14.02 ppb). There was a positive association between asthma symptom scores
(comparing children who report symptoms interfering with their daily activities versus those
with no symptoms or symptoms not great enough to affect their daily activities) and high current
levels of formaldehyde (OR 1.90 [95% CI: 1.13-3.19]).
Three studies (Tavernier et al., 2006; Gee et al., 2005; Garrett et al., 1999) were
performed by matching children with and without asthma and comparing the levels of
formaldehyde in their homes. Gee et al. (2005) reported median formaldehyde levels of 0.03
ppm in living rooms and 0.04 ppm in bedrooms. Analyses were limited to univariate
comparisons of formaldehyde levels for cases of existing asthma and controls without asthma.
The concentrations did not differ in a statistically significant manner. The study by Gee et al
(2005) was followed up with a more sophisticated analysis of the same children in the same
home. Tavernier et al. (2006) reiterated the earlier finding by Gee et al (2005) that formaldehyde
was not found to be associated with existing asthma. Tavernier et al. (2006) did not report the
measured levels of formaldehyde but gave the OR for the highest tertile of exposure compared
with the lowest tertile of exposure as 0.99 (95% CI: 0.39-2.50). The width of this confidence
interval suggests that, while no effect was observed, these findings would still be consistent with
two-fold increase in risk.
Garrett et al. (1999) reported on the risk of allergy and asthma-like respiratory symptoms
due to formaldehyde exposure in a cross-sectional survey of households with children with (n =
53) or without (n = 88) doctor-diagnosed asthma. Formaldehyde exposure was characterized by
4 seasonal in-home sampling events across the year for bedrooms and 4-day passive samples
collected in living rooms, kitchens and outdoors. Statistically significant linear trends for
increased risk of having asthma were seen with increasing formaldehyde levels (p < 0.02);
however, the ORs for the association did not remain statistically significant after controlling for
parental allergy and asthma (exact ORs and 95% CIs not given). Garrett et al (1999) also
evaluated the prevalence and severity of allergic sensitization to 12 common allergens and
reported increased prevalence with increasing formaldehyde concentration in the home. The
respiratory symptom score was also increased and demonstrated a significant effect for
formaldehyde in a multiple regression after adjusting for multiple risk factors and interactions.
For the atopy and respiratory symptom endpoints, severity/incidence was increased in the
medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the low (<20 |ig/m3)
exposure group, based on the highest of four seasonal 4-day formaldehyde measurements in the
home. The associations between formaldehyde concentrations and severity of allergic
sensitization are clearly shown and further substantiated with multivariate regression controlling
This document is a draft for review purposes only and does not constitute Agency policy.
5-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
for potential confounders. In logistic regressions, both the prevalence and severity of allergic
sensitization to 12 common allergens increased with increasing formaldehyde concentration in
the home. The crude association for atopy with an increase in formaldehyde concentration per
10 |ig/m3 was OR=1.34 which increased when adjusted for parental asthma and gender to and
odds ratio of 1.42 per 10 |ig/m3 (95% CI: 0.99-2.04). Passive smoking, the presence of pets,
indoor nitrogen dioxide concentrations, airborne fungal spores and house-dust-mite allergens did
not influence the effect estimates and were unlikely to be confounders. Additionally, a
calculated respiratory symptom score was increased and demonstrated a significant relationship
to increased formaldehyde concentration in a multiple linear regression after adjusting for
multiple risk factors and interactions. For each of these endpoints, severity/incidence was
increased in the medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the
low (<20 |ig/m3) exposure group, based on the highest of four seasonal 4-day formaldehyde
measurements in the home.
Residential formaldehyde exposure was associated with an increased risk of asthma in a
population-based case-control study of 192 children aged 6 months to 3 years (Rumchev et al.,
2002). The study, which comprises 88 cases of children discharged from the emergency
department of a children's hospital in Perth, Australia, with a primary diagnosis of asthma and
104 controls, provides a positive exposure-response relationship. Seasonal in-home
formaldehyde measurements taken in the living room and subject's bedroom were used to assess
exposure (8-hour passive sampler). The odds ratios (ORs) for risk of asthma by formaldehyde
exposure level category were adjusted for numerous risk factors both familial and environmental
including, familial history of asthma, age, sex, smoking, presence of pets, and attributes of the
home. Of these, age, allergic sensitization to common allergens, and family history of allergy
were independent risk factors for asthma (ORs of 1.09, 2.57, and 2.66, respectively). Categorical
analysis of the data indicates the ORs for asthma were increased in the two highest formaldehyde
exposure groups, reaching statistical significance for household exposures > 60 |ig/m3 (48 ppb)
(OR of 1.39). Analysis of the data with formaldehyde as a continuous variable indicated there
was a statistically significant increase in the risk of asthma (3 % increase in risk per every
10 ug/m3 increase in formaldehyde level. All analyses controlled for other indoor air pollutants,
allergen levels, relative humidity, and indoor temperature as well as other risk factors.
A study of 202 households (mean formaldehyde level of 26 ppb) found that among children aged
6-15 years old and exposed to environmental tobacco smoke, the prevalence of asthma was
45.5% for those with measured levels of formaldehyde in the kitchen >60 ppb. The prevalence
of asthma dropped to 15.1% for levels <40 ppb and 0% for 41-60 ppb. No trend in asthma
This document is a draft for review purposes only and does not constitute Agency policy.
5-14 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
prevalence was seen for children who were not exposed to environmental tobacco smoke
(Krzyzanowski et al., 1990).
Finally, a study by Norback et al. (1995) reported mean levels of formaldehyde were
29 |ig/m3 (range <5-110 |ig/m3) in the bedrooms of individuals experiencing nocturnal
breathlessness compared with formaldehyde levels of 17 |ig/m3 (<5-60 |ig/m3) among those
without nocturnal breathlessness. The OR for this association was 12.5 (95% CI: 2.0-77.9) and
the effect was substantially stronger in magnitude than the associations observed for toluene,
terpenes and volatile organic compounds which makes confounding by those co-exposures
unlikely.
Supporting animal studies:
Several animal studies report increased airway resistance and BC due to inhalation
exposures to formaldehyde (Nielsen et al., 1999; Swiecichowski et al., 1993; Biagini et al., 1989;
Amdur, 1960). Changes in pulmonary resistance were observed as early as 10 minutes after
exposure (Biagini et al., 1989), and reported effect levels ranged from 0.3-13 ppm. Other
pulmonary effects were reported in conjunction with BHR, such as increased tracheal reactivity
and decreased pulmonary elasticity (Swiecichowski et al., 1993; Amdur, 1960). Although BHR
is a common result of Type I hypersensitivity reaction to an allergen, the observation of BHR
alone is not sufficient to demonstrate that an agent induces Type 1 hypersensitivity.
BHR may be directly induced both pharmacologically and neurogenically (Joos, 2003;
Cain, 2001; Meggs, 1995). There is little evidence that formaldehyde itself is an allergen
recognized by the immune system, especially via inhalation (Lee et al., 1984). Although
formaldehyde exposure has been reported to alter cytokine levels and immunoglobulins in some
experimental systems, these immunomodulatory effects do not support a type 1 hypersensitivity.
IgE was unchanged (Fujimaki et al., 2004a; Lee et al., 1984), and cytokine profiles were not
consistent with the Th-2 cytokines expected in IgE mediated hypersensitivity (Fujimaki et al.,
2004a; Ohtsuka et al., 2003).
Formaldehyde-induced dermal sensitization show parallel results. The physical signs of
irritation and sensitization are consistently shown (e.g., rashes, edema). Some involvement of
the immune response has been demonstrated with positive LLNA assays, indicating proliferation
of lymphocytes in lymph nodes draining the affected area (Hilton et al., 1998; Arts et al., 1997).
Increased expression of Th-2 cytokines in the lymph nodes of mice given dermal applications of
formaldehyde does indicate an immune component to the observed sensitization. However, the
response does not seem to be mediated by IgE (Arts et al., 1997; Lee et al., 1984).
This document is a draft for review purposes only and does not constitute Agency policy.
5-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Ito et al. (1996) reported that a tachykinin NKi receptor, but not the histamine Hi or
bradykinin B2 receptors, is involved in formaldehyde-induced vascular permeability.
Neuropeptides NGF and substance P were affected in BAL and stimulated splenocytes from
formaldehyde-exposed mice, with greater effects seen in OVA-immunized mice. Tachykinins
(e.g., substance P and neurokinin A) are produced by nerve cells and can directly stimulate
bronchoconstriction (Van Schoor et al., 2000). Substance P is also a mediator of neurogenic
inflammation. Therefore, although formaldehyde may induce some of the symptoms of type 1
hypersensitivity, these symptoms are more likely neurogenic than immunogenic in origin.
In contrast, formaldehyde enhances immunogenic hypersensitivity of known allergens
(Sadakane et al., 2002; Riedel et al., 1996; Tarkowski and Gorski, 1995). This potentiation
varied based on sensitization protocols (respiratory tract versus systemic, frequency and timing
of immunization, allergen, etc.) and formaldehyde exposure regimens (concentration, continuous
versus intermittent exposures). Taken as a whole, the results support the finding that
formaldehyde exposure can aggravate a type 1 hypersensitivity response (Table 4-53).
The mechanism underlying this response has not been elucidated. Formaldehyde-
induced IgE production has been reported in some studies (Vandenplas et al., 2004; Wantke et
al., 1996a). Other studies suggest that this effect does not appear to be immunogenic in nature
(Fujimaki et al., 2004; Lee et al., 1984). Although formaldehyde exposure has been reported to
alter cytokine levels and immunoglobulins in some experimental systems (Fujimaki et al., 2004a;
Ohtsuka et al., 2003), these immunomodulatory effects do not support immunogenically
mediated type 1 hypersensitivity.
These decrements may be mediated via neurogenic potentiation (Sadakane et al., 2002;
Riedel et al., 1996; Tarkowski and Gorski, 1995). Tarkowski and Gorski (1995) suggest that
formaldehyde may increase permeability of respiratory epithelium and destruction of
immunologic barriers. Tachykinin NKI receptor and various neuropeptides (NGF and substance
P) have been implicated in formaldehyde-induced sensitization and lend weight of evidence to a
neurogenic MOA (Van Schoor et al., 2000; Ito et al. 1996).
5.1.1.5 Immune Function
Although there are some indications of formaldehyde-induced immunomodulation in
laboratory animal studies (Jakab, 1992; Morgan et al., 1986a, b, c; Leach et al., 1983) and
reports of increased upper respiratory tract infections in formaldehyde-exposed workers
(Lyapina et al., 2004; Krzyzanowski et al., 1990; Holness and Nethercott, 1989), the overall
database for toxic effects on immune function and competence is very limited. A study of
workers using carbamide-formaldehyde glue indicates decreased neutrophil respiratory burst
This document is a draft for review purposes only and does not constitute Agency policy.
5-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
activity (NRBA) (Lyapina et al., 2004). NRBA was reduced in workers with URT inflammation
and long-lasting respiratory tract infections, compared with healthy controls, and in
formaldehyde-exposed workers with slight or no respiratory infections. The authors
hypothesized that the decreased NRBA in symptomatic workers may be an indication of
formaldehyde effects in a susceptible population. Since the workers have increased respiratory
tract infections as compared with controls, a formaldehyde-specific effect cannot be excluded.
These indications of a functional deficit of the immune system are considered adverse and
appropriate for consideration as a critical effect. Although this was a small study (n = 29), the
exposed workers had increased chronic URT infections and decreased resistance to infections
compared with a control population. Additionally, duration of employment was negatively
correlated with both erythrocyte count and hematocrit. Measured formaldehyde concentrations
for a work shift were 870 ± 390 |ig/m3 (722 ± 324 ppb). This average work-shift concentration
is considered to be the LOAEL for increased respiratory tract inflammation and decreased
resistance to infections in a worker population.
5.1.1.6. Neurological and Behavioral Toxicity
Studies evaluating the effects of formaldehyde on nervous system structure or function
are described in detail in Section 4.2.1.4 and summarized in Table 4-60. Taken together the
animal and human data support the conclusion that formaldehyde exposure results in
neurological and behavioral toxicity. Observed health effects include impaired memory and
learning, developmental effects seen as both structural changes in the brain and behavioral
changes, and a potential for increased mortality from amyotrophic lateral sclerosis (ALS).
Although studies appropriate for RfC derivation do not exist for each potential neurological and
behavioral health effect, several studies are available that may inform the formaldehyde RfC.
Seven of the available neurotoxicity studies were considered as candidates for RfC
development (listed in Table 5-1). All seven studies provided reliable documentation of
exposure, study design, and evaluation procedures, and all demonstrated robust findings of
changes in nervous system structure or function following formaldehyde exposure. All but one
of the candidate studies present information at multiple exposure levels to provide an
understanding of the exposure response relationship. One selected study (Senichenkova, 1991)
provided less robust information, with evaluation at only a single exposure level, but was
considered useful as supporting the findings of two other studies (Sarsilmaz et al., 2007; Asian et
al., 2006) regarding neurological sequelae of developmental exposure. All of the selected
studies using experimental animals were conducted in rats, although several studies in mice
demonstrated dose-related neurotoxic effects following formaldehyde exposure. These studies in
This document is a draft for review purposes only and does not constitute Agency policy.
5-17 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
mice were not considered for RfC development because of the possibility that results might be
confounded by reflex bradypnea at the doses tested in each study (see Section 4.2.1.4 for
details).
In order to improve transparency and facilitate comparison of health effect levels across
study types and health effects, Table 5-1 summarizes the PODs and exposure scenarios for each
selected study and describes the effects on which the selected POD is based. Dose conversions
used to adjust from actual experimental exposure concentrations to continuous exposure
concentrations are detailed. It should be noted that available studies providing dose-response
information regarding the effects of formaldehyde exposure on the nervous system were all of
short duration, and thus information regarding the relationship between formaldehyde toxicity
and exposure duration (i.e., whether toxicity increases with longer exposures at a given exposure
level, or is more related to the maximum exposure concentration) is limited. However, the
rodent study by Pitten et al. (2000) and the epidemiology study by Weisskopf et al. (2009)
provide strong support for an association between increasing neurotoxicity and increasing
duration of exposure.
Although chronic human studies are preferred for RfC derivation, no adequate human
study of chronic duration is available. The available human studies were sufficiently strong to
raise concern regarding formaldehyde effects on the nervous system; however, most did not
provide sufficient exposure information to permit derivation of a POD for use in quantitative
dose-response assessment. Available epidemiologic studies (most notably Weisskopf et al.
[2009] and Kilburn et al. [1987, 1985]) provided limited exposure information. Weisskopf et al.
(2009) demonstrated increased mortality from ALS associated with increased duration of
formaldehyde exposure among 987,229 people followed by an American Cancer Society study,
but no information regarding exposure concentrations was available. Interpretation of the
findings of Kilburn et al. (1987, 1985) is complicated by concomitant exposure of many subjects
to other solvents. Although the chamber study by Lang et al. (2008) included a concentration-
response assessment of changes in reaction time, as previously discussed, the effects detected
were difficult to interpret and the study was not considered useful for RfC derivation
This document is a draft for review purposes only and does not constitute Agency policy.
5-18 DRAFT—DO NOT CITE OR QUOTE
-------
5 ^
0
rs
a ^3
Grj St
!"§
Crq o
^ a
a -
1 §.-§,
m
>!
5^
W
§•
rs
sj
3
5
>!
6
Table 5-1. Points of departure (POD) for nervous system toxicity in key human and animal studies.
Reference
PODa
Exposure scenario
POD duration adjustments'"
Effect
Species
Type
ppb
Hours/day
Days/week
Duration
POD x Hours/day x Days/week = ppb
Ratio0
Developmental neuropathology effects
Sarsilmaz
et al.
(2007)
Rat
LOAEL
6,000
6
5
30 days
6,000 x 6/24 x 5/7 = 1,070
5.6
Volume and cell number
change in brain regions
following neonatal
exposure
Asian et al.
(2006)
Rat
LOAEL
6,000
6
5
30 days
6,000 x 6/24 x 5/7 = 1,070
5.6
Volume and cell number
change in brain regions
following neonatal
exposure
Human neurobehavioral outcomes
Bach et al.
(1990)d
Human
NOAEL
170
5.5
1
1 day
170d x x =170
1
Changes in short-term
memory and ability to
concentrate. Single
5.5-hour exposure
Psychomotor effects
Senichenko
va (1991)
Rat
LOAEL
400
4
7
GD 1-19
400 x 4/24 x 7/7 = 67
6
Changes in open field
motor activity (exploratory
activity and habituation in
offspring following in
utero exposure
Malek et
al. (2003c)
Rat
LOAEL
130e
2
1
1 day
130 x 2/4e x =65
2
Concentration-dependent
decreases in activity by a
variety of measures
following a single
exposure
Cognitive effects
Pitten et al.
(2000)f
Rat
LOAEL
2,600
0.17
7
90 days
2,600f
-
Impaired memory in a
spatial maze. Magnitude
of effect increased with
continued exposure
through 12 weeks
Y*
a
&
a
o
-------
a ^3
Grj St
li
>!
5^
W
§•
rs
sj
3
5
>!
ca
a
Malek et
al. (2003a)
Rat
LOAEL
100e
10 days
100
2/4e
7/7
= 50
Impaired learning in a
water maze. Short-term
(10 day) exposure with
testing conducted 2 hours
following daily exposure.
to
^ <*>
Crq O
a
a
>!
0
1
al mg/m3 = 0.813 ppm.
bBoth actual levels of experimental exposures, and duration adjusted PODs are shown.
°POD unadjusted dose / duration-adjusted dose.
dTesting was conducted during or following exposure, duration was not adjusted.
"Testing was conducted 2 hours postexposure; duration was adjusted to 4 hours to include the entire period between start of exposure and testing.
fDue to the uncertainty in continuous exposure adjustments and the unusually short (10 minutes) exposure in this study, no adjustment to continuous exposure is
presented.
to
o
O
H
b
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
17
18
19
20
21
22
23
One acute human study, Bach et al. (1990), which evaluated changes in cognitive
function following a single formaldehyde exposure, was considered for evaluation of a cRfC as
the chamber exposures were well defined and effects at multiple levels of exposure were
reported. In that study, concentration-related changes in short-term memory and ability to
concentrate were seen during a single 5.5-hour exposure at a range of levels (32, 170, 390, and
890 ppb). The study was designed as a comparison of effects of short-term formaldehyde
exposure in previously occupationally exposed individuals with effects in controls without
previous occupational exposure. Because occupational exposure levels were not assessed,
exposure measurements from the previously exposed workers are not appropriate for use in RfC
derivation. The authors reported a significant exposure-response relationship for three related
cognitive measures (number of additions completed, number of errors, and reaction time) in the
'addition test' assessment indicating a deficit in performance. Complete data were not
presented, but graphical presentations in the article indicated that the effect was seen at all doses
tested, with an apparent NOAEL of 170 ppb (see Figure 5-1).
N!0
o
"O
0.5 1.0
Formaldehyde concentration
0.1
Figure 5-1. Change in number of additions made in 10 minutes following
formaldehyde exposure at 32,170, 390, or 890 ppb.
Note: Vertical bars are the standard errors of the means, dashed line shows the
95% CI.
Source: Bach et al. (1990).
This document is a draft for review purposes only and does not constitute Agency policy.
5-21 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
No BMD modeling could be performed on these data because the graphical
representation could not be accurately digitized. The statistical analysis indicated no interaction
between formaldehyde effect and previous occupational exposure (i.e., the magnitude and
direction of the effect were similar in previously exposed and previously unexposed subjects)
and separate data were not presented for the two groups; thus, the LOAEL represents effects in
the combined study groups. Overall, the published paper lacks detail and it is difficult to
evaluate some aspects of the reported findings, in particular where magnitude and direction of
effect are not provided. Finally, the authors noted that controls and the high-exposure group
were not well matched on two key parameters (age and education level), adding uncertainty to
the reported exposure-response relationship (at the high dose). Although this study was
considered valuable in documenting neurological effects in humans following exposure to
relatively low concentrations of formaldehyde, the above concerns limit its utility for
quantitative human health risk assessment. Therefore, this study is not considered of sufficient
quality for RfC derivation.
In the absence of adequate human data, controlled studies in laboratory animals are
considered. There are no chronic studies and only one subchronic animal study evaluating
neurological and behavioral effects of formaldehyde exposure. Pitten et al. (2000) demonstrated
impaired retention of a previously learned task in rats exposed at concentrations of 2,600 or
4,600 ppb, 10 minutes per day, 7 days/week, for 90 days. In this study, the magnitude of the
impairment increased over time, even though testing was performed 22 hours after exposure,
indicating that repeated formaldehyde exposure led to a worsening of effect. The study design,
test methods, and reporting of the results are all of adequate quality for both hazard assessment
and quantitative risk assessment. However, the short duration (10 minutes) of the repeated daily
exposures is a severe limitation to establishing a chronic RfC based on this study, due to
uncertainties in extrapolating from 10 minutes to a 24-hour exposure (see Table 5-1). Because
this study as designed indicates an accumulation of effect with repeated exposure, it is useful in
documenting the existence of a duration component to the exposure-response relationship. It
follows that concentration alone, without an adjustment for duration of exposure, would be
inadequate as an exposure metric; however inadequate information is available to inform the
appropriate magnitude of the duration effect. Therefore, although Pitten et al. (2000) is a well-
conducted study, the data are of limited utility for RfC derivation.
Finally, there are several well-documented acute and subacute animal studies that provide
exposure-response information for neurological and behavioral endpoints relevant for RfC
derivation. Several laboratory animal studies that evaluate neurological effects following in
utero or neonatal exposure address potentially susceptible life stages. Sarsilmaz et al. (2007) and
This document is a draft for review purposes only and does not constitute Agency policy.
5_22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Asian et al. (2006) observed changes in brain structure (cell number and/or volume changes in
specific brain regions) following 30 days of exposure to neonatal rats. A related finding by
Senichenkova (1991) demonstrated changes in behavior (open field motor activity, including
habituation) in young rats following in utero exposure. Effects of concern were seen at all doses
in these studies, resulting in PODs of 67 ppb following in utero exposure and 1,070 ppb
following early postnatal exposure, based on LOAEL values adjusted for continuous exposure
(see Table 5-1). These studies support the possibility of neurodevelopmental effects attributable
to in utero or early postnatal formaldehyde exposure, at levels similar to or below those causing
other types of effects.
The other three studies in Table 5-1 evaluate behavioral changes in rats following
exposure to formaldehyde. Malek et al. (2003c) found concentration-related changes in motor
activity following a single 2-hour exposure at concentrations from 130-5,180 ppb (with testing
2 hours following cessation of exposure). In a second study, Malek et al. (2003a) found
concentrated-related changes in performance on a learning task at similar exposure levels (100-
5,400 ppb) when 2-hour exposures were repeated for 10 consecutive days; performance was
evaluated 2 hours after cessation of exposure, and concentration-related learning deficits were
seen at all exposure levels (see Table 5-2 and Figure 5-2).
Table 5-2. Effects of formaldehyde exposure on completion of the labyrinth
test by male and female LEW.1K rats
Male rats
Swimming time (sec)
Error rate (mean)
Day 1
Day 6
Day 10
Day 1
Day 6
Day 10
Control
105
12.2
6.33
7.4
0.5
0.0
0.1 ppma
100
12.9
6.07
7.7
5.0C
3.2C
0.5 ppm
97
16.7C
7.60b
7.6
4.4C
1.8°
5.4 ppm
105
25.7C
10.9C
7.7
5.0C
2.8C
Female
rats
Swimming time (sec)
Error rate (mean)
Day 1
Day 6
Day 10
Day 1
Day 6
Day 10
Control
103
12.5
6.47
7.9
0
0.0
0.1 ppm
96
12.3
7.53
7.1
5.2C
3.0C
0.5 ppm
97
14.6C
7.60b
8.0
4.6C
2.2C
5.4 ppm
98
23.5C
9.73c
7.9
5.2C
2.6C
"Rats were exposed to formaldehyde for 2 hours/day, for 10 consecutive days.
bDifferent from control, p < 0.05.
Different from control, p < 0.005.
Source: Malek et al. (2003a).
This document is a draft for review purposes only and does not constitute Agency policy.
5_23 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
1 23456789 10
Day of labyrinth test
Figure 5-2. Effects of formaldehyde exposure on the error rate of female
LEW.1K rats performing the water labyrinth learning test
Source: Drawn from data reported in Malek et al. (2003a).
Although other studies evaluating neurobehavioral effects were available in the
formaldehyde database (see Chapter 4), these studies by Malek et al. (2003a, c) were considered
to be the most robust, documenting effects at relatively low exposure levels. Both studies also
included evaluation at multiple concentrations and showed concentration-related increases in
effect. In the Malek et al. (2003a) study with repeated exposures, it is unclear whether or not the
measured effect primarily reflects the most recent exposure or cumulative exposure; therefore,
the adjustment for continuous exposure was made over the final exposure period and the 2 hours
following exposure (4 hours total), as was done for the single-exposure study (Malek et al.,
2003c). After appropriate duration adjustments, PODs for these studies range from 50 to 67 ppb
(based on LOAELs), and the types of effects seen provide support for the Bach et al. (1990)
study that detected cognitive impairments in humans following a single exposure (with a
NOAELof 170 ppb).
This document is a draft for review purposes only and does not constitute Agency policy.
5_24 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Summary of neurological and behavioral effects:
In summary, the available studies for formaldehyde and nervous system outcomes have
demonstrated that the nervous system is a sensitive target following inhalation of formaldehyde.
In experimental animals, changes in nervous system function were seen following acute and
subchronic exposures; studies evaluating neurological changes following chronic exposure were
unavailable. Available human studies that evaluated nervous system effects following inhalation
exposure were found to have many study-specific uncertainties and, thus, were not suitable to
serve as the primary basis for a chronic RfC. The Weisskopf et al. (2009) study of ALS, in
particular, suggests that humans may be at risk for severe neurological effects from
formaldehyde exposure; however, this study lacked the exposure concentration information
necessary to derive an RfC. Neurological findings from the rodent inhalation (acute and
subchronic) studies that were judged to be adequate for dose-response assessment identified
unadjusted LOAELs ranging from 100 to 6000 ppb, with LOAELs adjusted for continuous
exposure in the range of 50 to 1070 ppb. Use of these PODs in risk assessment would require
addressing uncertainties regarding animal-to-human extrapolation, short study durations, and
extrapolation from LOAELs.
Among the adequate studies, EPA considered Malek et al. (2003a) to be the most
appropriate for calculation of a cRfC for neurological and behavioral toxicity, based on the
exposure level at which effects were seen (100 ppb), the type of effect (impaired learning),
which is relevant to humans, and the use of a repeated-exposure paradigm (2 hours/day over a
period of 10 days), which addresses different exposure durations. This choice is supported by
similar effects seen in other studies (Lu et al., 2008; Pitten et al., 2000; Bach et al., 1990) and by
other neurologic effects seen at similar exposure levels (Malek et al., 2003c; Senichenkova,
1991; Sheveleva, 1971).
5.1.1.7. Developmental and Reproductive Toxicity
As described in Sections 4.1 and 4.2 and summarized in Tables 4-68 and 4-71, both
human epidemiologic data and experimental animal studies demonstrate an association between
formaldehyde inhalation exposure and adverse developmental and reproductive effects, where
adversity is characterized as per EPA risk assessment guidelines (U.S. EPA, 1991; U.S. EPA,
2006). Adverse outcomes were observed across the various manifestations of developmental
toxicity, including fetal death, structural alterations (including congenital malformations),
growth retardation, and functional development. Additionally, in spite of the lack of a
comprehensive database of studies for the evaluation of the overall effects of formaldehyde on
the reproductive system and its function, the available evidence demonstrates toxicity to the male
This document is a draft for review purposes only and does not constitute Agency policy.
5_25 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
reproductive system in multiple animal studies, as well as effects on the female reproductive
system in both rodents and epidemiologic studies, where an association with impaired fertility
and increased spontaneous abortions were noted.
Potential principal studies for specific adverse outcomes are presented and evaluated
below including reproductive effects (male and female), fetal death, growth retardation, and
structural alterations. The only available evidence for functional alterations is based on
developmental neurotoxicity studies which are presented and evaluated in Section 5.1.1.6.
Table 5-3 summarizes animal studies deemed suitable for deriving quantitative dose-response
information for reproductive and developmental outcomes and their corresponding PODs,
adjusted for continuous exposure. Calculations that were used in dose conversions and exposure
duration adjustments for the POD values are included. In general, repeated daily exposures of
laboratory animals are adjusted from a partial day to a 24-hour exposure and then weighted for
the number of days per week the exposures occurred. No chronic animal studies evaluating
these endpoints were available, so only subchronic and acute studies are considered. Exposure
duration adjustments to the only suitable human study (Taskinen et al., 1999) are more complex
due to uncertainties in the exposure data and the potential for non-occupational exposures. For
this discussion the reported 8-hr TWA exposures will be used for the Taskinen et al. (1999)
study. Further duration adjustments to this study are discussed in Section 5.1.2.2.5 for cRfC
derivation.
Spontaneous abortion andfetal death.
Increased risk of spontaneous abortion following maternal occupational formaldehyde
exposure was reported in a number of epidemiologic studies (Taskinen et al., 1999, 1994; John et
al., 1994; Seitz and Baron, 1990; Axelsson et al., 1984). The studies did not appear to be overtly
influenced by common principle biases found in epidemiologic studies. Considered together, the
studies are consistent with an adverse effect of formaldehyde exposure on pregnancy loss, where
adversity is characterized as per EPA risk assessment guidelines (U.S. EPA, 1991; U.S. EPA,
2006). . Of these studies, Taskinen et al. (1999) had the superior quantitative data reporting
reduced fecundity and spontaneous abortion in the exposed workers. Taskinen et al. (1999) is an
occupational study with a well-considered study design, including measurements of exposure
and outcomes, and relatively high study power. The study population consisted of 602 female
workers in Finland who had at least one successful childbirth and first employment in the wood-
working industry beginning at least 6 months prior to the studied pregnancy. Mean daily
formaldehyde inhalation exposures during the time-to-pregnancy period were estimated for each
worker, based on task-level exposure measurements and work history.
This document is a draft for review purposes only and does not constitute Agency policy.
5_26 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Exposure was reported as a daily exposure index representing the average daily exposure
for the time-to-pregnancy period, and three exposure classes were defined as low, medium and
high with equivalent mean work-shift TWA exposure of 18, 76 and 219 ppb, respectively.
Fecundity density ratio (FDR) was significantly reduced in the high exposure group compared to
the referent group (P=0.02), and the risk of spontaneous abortions was increased with reported
ORs of 3.2 (1.2-8.3), 1.8 (0.8-4) and 2.4 (1.2-4.8) for the high, medium and low exposure
groups, respectively. The effect on FDR remained in workers both with and without the use of
protective gloves (n=39) but lost statistical significance in the exposed group of workers who
wore gloves (n=22). Figure 5-3 shows the study results stratified by glove use in women in the
high-exposure group. Although this suggests that a component of dermal exposure might
contribute to the effect, it is unclear what, if any, dermal exposure is expected based on the
nature of the work. Regardless, there remains uncertainty as to whether effects are solely due to
inhalation exposure.
1.4
1.2
IO
o>
0.8
a:
Q
0.6
0.4
0.2
All Women (N=39)
No Gloves (ISM 7)
Gloves (N=22)
Figure 5-3. Fecundity density ratio among women exposed to formaldehyde
in the high exposure index category with 8-hour time-weighted average
formaldehyde exposure concentration of 219 ppb (Taskinen et al., 1999)
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
5_27 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
In some available rodent studies (Kitaev et al., 1984; Sheveleva, 1971), evidence of
increased embryo degeneration in early gestation or of preimplantation loss (findings that are
generally comparable to spontaneous abortion in humans) was observed. In the Kitaev et al.
(1984) study, early implantation losses resulted following treatment of dams prior to mating.
This may support a possible contribution of pre-pregnancy exposures to the spontaneous
abortions observed in Taskinen et al. (1999). Quantification of the findings by Kitaev et al.
(1984) and Sheveleva (1971) resulted in adjusted PODs of 50 and 70 ppb, respectively, based
upon study LOAELs (Table 5-3).
Structural alterations.
Studies of occupational exposures to formaldehyde examined the incidence of congenital
malformations, but exposure and outcome data were not fully characterized and therefore could
not be carried forward to RfC development. Animal studies (Senichenkova and Chetobar, 1996;
Senichenkova, 1991) reported increases in internal organ anomalies; the most frequently
observed structural anomaly was a delay in fetal testis descent (at times characterized as
cryptorchidism in the study reports). For both studies, which exposed rats to formaldehyde for 4
hours/day during gestation, adjusted PODs based upon LOAELs were 70 ppb (Table 5-3). These
studies included only one treatment level, precluding the ability to establish a dose-response
relationship, and the observed outcomes were not noted in other developmental toxicity studies
with similar exposure scenarios, thus limiting the strength of the studies for use in RfC
derivation.
Growth retardation.
Decreased fetal weight was observed in a number of animal studies that exposed pregnant
rats to formaldehyde during gestation. Of these, based on adequacy of dose-response
information, Saillenfait et al. (1989) was considered appropriate for consideration for RfC
development. In this study, rats were administered formaldehyde 6 hours/day on gestational
days (GDs) 6-20. Decreased male fetal body weight (BW) was modeled with a BMR of 5%
mean change, a BMCL was established, and, as shown in Table 5-3, the resulting duration-
adjusted POD of 325 ppb was derived. The relevance of this finding to human exposures was
qualitatively supported by a population-based study by Grazuleviciene et al. (1998) that reported
an association between atmospheric formaldehyde exposure and low birth weight; although a
dose-response relationship could not be adequately quantified from the information provided.
This document is a draft for review purposes only and does not constitute Agency policy.
5_28 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Male reproductive toxicity.
Evidence of adverse effects on male reproductive system endpoints following inhalation
exposure to formaldehyde was observed in a number of animal studies, where adversity is
characterized as per EPA risk assessment guidelines (U.S. EPA, 1991; U.S. EPA, 2006). The
effects include decreased testes weight, changes in Leydig cell quantity and quality, degeneration
of seminiferous tubules, decreased testosterone levels, alterations in biomarkers of toxicity in the
testes, and alterations in sperm count, morphology, and/or motility (Golalipour et al., 2007; Xing
et al., 2007; Zhou et al., 2006; Ozen et al., 2005, 2002; Sarsilmaz et al., 1999; Guseva, 1972).
Several of these studies included inhalation exposure of rats to formaldehyde 8 hours/day,
5 days/week for 4 and/or 13 weeks (Ozen et al., 2005, 2002; Sarsilmaz et al., 1999) and included
exposure-response information that was considered adequate for RfC derivation. In a study by
Ozen et al. (2002), increased severity of statistically significant testes weight decreases was
related to both dose and duration of treatment. Similarly, in the study by Golalipour et al.
(2007), seminiferous tubular diameter and epithelial height were reduced in rats following 18
weeks of formaldehyde inhalation exposure, with the severity of outcome positively correlated to
the number of hours/week that the animals were exposed. Sarsilmaz et al. (1999) noted dose
dependent decreases in Leydig cell quantity after 4 weeks of treatment, while decreased testis
weight and atrophy of seminiferous tubules were observed by Zhou et al. (2006) after only 2
weeks of treatment. The reported outcomes in these independent studies illustrate a biologically
consistent toxicological profile of treatment-related male reproductive toxicity. PODs, adjusted
for continuous exposure, ranged from 1,190 to 4,025 ppb, where the lowest POD was associated
with the longest exposure period and vice verse (Table 5-3).
Female reproductive toxicity.
Evidence of decreased fecundability was observed in the study by Taskinen et al. (1999),
which was described above for spontaneous abortions. Delays in the time to conception that
characterized this outcome, as well as increases in the incidence of endometriosis, were
statistically significantly associated with occupational exposures to formaldehyde. As these
effects were observed in the high exposure group, the unadjusted NOAEL for each of these
effects is 76 ppb (8 hr-TWA) based on the next lowest exposure group. Uncertainties included
lack of information human variability, as well as on the extrapolation of data from studies of
short duration to risk estimates for chronic exposures. As discussed above for spontaneous
abortions, the use of these data for cRfC derivation could result in values that would likely be an
underestimation of risk because they assume that all the risk was from inhalation exposure and
ignore the apparent contribution of dermal exposure (i.e., the dermal-exposure-adjusted
This document is a draft for review purposes only and does not constitute Agency policy.
5_29 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
candidate inhalation RfCs might be higher). For decreased fecundability, a POD can also be
identified based on the data from only the women who wore gloves. The fecundability density
ratio (FDR) for the women in the highest exposure group who wore gloves was 0.79 (95%
confidence interval [CI] 0.47-1.23). Although this FDR is not statistically significant, it can
reasonably be assumed to be part of a trend of decreased FDR with increasing inhalation
exposure, based on the overall data for the association of decreased FDR with formaldehyde
exposure. Whether the highest exposure level is considered to be a NOAEL or a LOAEL for
decreased fecundability in women who wore gloves, the unadjusted POD is 219 ppb (8 hr-
TWA). Evidence of spontaneous abortions in the same study, as described above, may also be
indicative of female reproductive toxicity.
In animal studies, assessment of the female reproductive system was quite limited. An
increase in the mean follicle-stimulating hormone (FSH) levels in rats, observed at the highest
exposure level tested in Kitaev et al. (1984) was found to be sufficient to derive a duration-
adjusted POD of 50 ppb (Table 5-3).
Summary of developmental and reproductive toxicity studies suitable for RfC development.
A review of the developmental and reproductive toxicity studies in humans and animals
that would be suitable for cRfC development demonstrated that the developing organism and the
reproductive system are targets for toxicity following formaldehyde exposure by inhalation. In
the animal studies, effects during early development were observed following maternal
premating or gestational exposures at duration-adjusted PODs ranging from 50-325 ppb. The
minimal data available on female reproductive toxicity demonstrated an adjusted POD of 50 ppb
with subchronic (4-month) premating exposure, while more extensive evaluation of male
reproductive outcomes identified adjusted PODs of 1,190-4,025 for testicular and sperm
abnormalities after exposures of from 2 weeks to 3 months in duration. The animal studies
demonstrate the broad range of adverse outcomes to the reproductive system and the developing
organism following inhalation exposure to formaldehyde and highlight concerns regarding the
inadequacy of the database for the assessment of these outcomes (as described in Chapter 4).
These data also support the human relevance of female reproductive and/or embryonic and fetal
developmental effects, since some outcomes were similarly observed in both human and animal
studies.
This document is a draft for review purposes only and does not constitute Agency policy.
5_30 DRAFT—DO NOT CITE OR QUOTE
-------
a 2
o S.
a ^3
Grj St
li
>!
5^
W
§•
rs
sj
3
5
>!
6
Si
Table 5-3. Developmental and reproductive toxicity POPs including duration adjustments - key animal studies
POD
Exposure Scenario
POD Duration Adjustments
Hours/
Days/
POD
Hours/
Days/
Adjusted
POD
Reference
Species
Type
ppba
day
week
Duration
(ppb)
day
week
(ppb)
Ratio b
Effect; comments
Spontaneous abortion and
fetal death
Kitaev et al.
Rat
LOAEL
400
4
5
6 months
400
X
4/24
X
5/7
=
50
8
Increased (>threefold) embryo
(1984)
premating
degeneration on gestational
days 2-3 after 4 months
maternal premating treatment
Sheveleva (1971)
Rat
LOAEL
400
4
7
GDs 1-19
400
X
4/24
X
7/7
=
70
5.7
Increased (50%)
preimplantation loss8
Structural alterationsc
Senichenkova
Rat
LOAEL
400
4
7
GDs 1-19
400
X
4/24
X
7/7
=
70
5.7
Increased (13%) litter
(1991)
incidence of internal organ
anomalies, including 20%
increase in undescended
testes; 9% decreased fetal
incidence of hyoid
ossification8
Senichenkova
Rat
LOAEL
400
4
7
GDs 1-19
400
X
4/24
X
7/7
=
70
5.7
Increased (21%) fetal and
and Chetobar
litter incidences of
(1996)
cryptorchidism and increased
(6%) fetal incidences of total
anomalies8
Growth retardation
Saillenfait et al.
Rat
BMCL
1,300
6
7
GDs 6-20
1,300
X
6/24
X
5/7
=
325
4
Decreased male fetal body
(1989)
weights8 (BMR = 5%)
Functional developmentd
Male reproductive toxicity
Ozen et al.
Rat
LOAEL
10,000
8
5
4 or 13
10,000
X
8/24
X
5/7
=
2,380
4.2
Decreased testis weight at
(2002)
weeks
4 weeks (2%) and 13 weeks
(8%)
Ozen et al.
Rat
LOAEL
5,000
8
5
91 days
5,000
X
8/24
X
5/7
=
1,190
4.2
Decreased (40%) serum
(2005)
testosterone levels at 91 days
Sarsilmaz et al.
Rat
LOAEL
10,000
8
7
4 weeks
10,000
X
8/24
X
7/7
=
2,380
4.2
Decreased (5%) Leydig cell
(1999)
numbers at 4 weeks
Crq o
§ 4
O
>!
Lh
U>
o
H
o
o
-------
a 2
o S.
a ^3
Grj St
li
>!
5^
W
§•
rs
sj
3
5
>!
6
c>
"^S
Table 5-3. Developmental and reproductive toxicity POPs including duration adjustments - key animal studies
Zhou et al.
(2006)
Rat
LOAEL
8,050
12
7
2 weeks
8,050
X
12/24
X
7/7
4,025
2
Decreased (-25%) testis
weight; alteration of
epididymal sperm [decreased
(38%) count, decreased (19%)
motility, and increased (>3-
fold) abnormal morphology]
at 2 weeks
Female reproductive toxicity
Kitaev et al.
(1984)
Rat
NOAEL
400
4
5
4 months
premating
400
X
4/24
X
5/7
50
8
Increased (-66%) follicle-
stimulating hormone at
4 months
Crq o
.rs
a
^ &
a
§•
>!
0
1
GDs = Gestation days
a 1 mg/m3 = 0.813 ppm.
b POD unadjusted dose / duration-adjusted dose
0 Neuropathological alterations following exposures during postnatal development (from the studies by Asian et al. [2006] and Sarsilmaz et al. [2007]) are
addressed in the neurobehavioral toxicity Section 4.2.1.6 and Table 5-2.
d Functional developmental endpoints (from the study by Senichenkova [1991]) are addressed in the neurobehavioral toxicity Section 4.2.1.6 and Table 5-2.
LtJ
to
o
H
b
O
-------
1
2
3
4
5
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
The animal study data were not selected for RfC derivation, since a high-quality human
study (Taskinen et al., 1999) was available for the purpose of deriving a chronic RfC. This
study, a well-designed population-based case-control study of women who were occupationally
exposed to formaldehyde, included a well-defined study population which was adequately
selected to allow for meaningful comparisons of health effects among individuals with different
levels of exposure to formaldehyde. Potential confounding factors such a selection bias and
inaccurate self-reporting were not considered to have had a significant influence on the study
findings. The increased risk of spontaneous abortion observed in Taskinen et al. (1999), and
perhaps the observed decrease in fecundity, is internally consistent and coherent with other
reports of increased risk of pregnancy loss associated with exposure to formaldehyde (John et al.,
1994; Taskinen et al., 1994; Seitz and Baron, 1990; Axelsson et al., 1984). It is also supported
by similar adverse outcomes observed in the animal data (Kitaev et al., 1984; Sheveleva, 1971).
5.1.2. Summary of Critical Effects and Candidate RfCs
5.1.2.1. Selection of Studies for Candidate RfC Derivation
The above reviews of data from both human and animal studies identified health effects
associated with formaldehyde exposure. Detailed information on these findings is given in
Chapter 4 (sections 4.1 and 4.2), and a qualitative summary of the noncancer hazard
identification is provided in Section 4.4 for each of the identified health effect categories:
sensory irritation, upper respiratory tract pathology, respiratory effects, increased atopic
response, immune function, reproductive and developmental toxicity, and neurobehavioral
toxicity. In this chapter, results for each health effect category are reviewed and studies are
identified which are adequate to inform the exposure-response relationship for health effects
from inhalation exposure (Section 5.1.1). Although the database of published studies that are
currently available does not provide adequate quantitative data to derive cRfCs for all
qualitatively identified endpoints, at least one adequate study was identified for each of the
health effect categories discussed above. For all but one of the categories, at least one study was
available that provided epidemiologic (human) data, based on occupational or residential
exposures, which was judged adequate to provide a quantitative basis for a cRfC.
In order to select the principal study or studies most appropriate for use as the basis of the
RfC for formaldehyde, the relative merits of these studies were evaluated with respect to study
quality, characteristics of the study population, the quality and frequency of exposure
measurements, and the exposure levels at which effects are observed. The ideal RfC would be
derived from a reported exposure level without an appreciable risk of deleterious effects in
humans, including sensitive populations, with little uncertainty. Additionally, where possible,
This document is a draft for review purposes only and does not constitute Agency policy.
5-33 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
the RfC should be derived with consideration of all of the identified health effects. The several
factors that were collectively taken into consideration for these studies (in no particular order)
included the following:
• Were studies of laboratory animals or humans?
o Human studies were generally preferred over laboratory animal studies for similar
health effects, when both were of good quality, given the uncertainties in
interspecies extrapolation.
• What was the study size?
o Larger studies were generally preferred over smaller studies because they can
give more precise estimates of response levels associated with specific exposure
levels.
• Among the epidemiologic (human) studies, were exposures from an occupational setting or
from a residential setting?
o Studies of health effects from residential exposures were generally preferred over
studies of health effects from occupational exposures because residential
exposures tend to have a smaller range of variability and are less prone to large
intermittent exposure peaks.
o Residential exposures are more representative of the exposures of the general
population.
• Among the epidemiologic (human) studies, were children among the study population in
which health effects were observed?
o Studies of health effects that assessed the effect of formaldehyde on children's
health, representing a potentially more susceptible life-stage for some effects,
were given some preference because they provide formaldehyde-specific data
relevant to the components of the RfC derivation that address potentially sensitive
life-stages and populations.
• Relative to the other studies under consideration for RfC development, how accurately were
formaldehyde concentrations measured?
o Studies based on relatively more accurately measured formaldehyde
concentrations were generally preferred over studies that estimated exposures.
• Studies that reported effects at relatively lower formaldehyde concentrations, potentially
indicative of more sensitive endpoints, were generally preferred.
This document is a draft for review purposes only and does not constitute Agency policy.
5.34 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Taking all the factors into consideration collectively, the individual studies are presented
in Table 5-4.
For sensory irritation, four studies are identified with adequate exposure information for
RfC derivation, and all are human studies (Table 5-4). Of these, 3 studies were conducted in
residential populations, including children and the elderly (Liu et al., 1991; Ritchie and Lehnen,
1987; Hanrahan et al., 1984). Each of these studies includes in-home formaldehyde
measurements for each participant. Liu et al. (1991) provide the best exposure measurements,
with 7-day in-home passive air samples collected in two seasons. The occupational study by
Holmstrom and Wilhelmsson (1988) provides evidence of sensory irritation in workers;
however, only the mean and range of exposures for all workers is given. Furthermore,
occupational exposures can include high peak exposures. The residential studies are preferred
for development of candidate RfC. Although there are differences in study size and the quality
of exposure measurements between the three residential studies, their results are mutually
supportive, defining similar effect levels in similar populations, and the use of the three
residential studies was considered to provide adequate consideration of the sensory irritation
endpoint. Therefore, all 3 studies are selected (Liu et al., 1991; Ritchie and Lehnen, 1987;
Hanrahan et al., 1984) and will be evaluated together in the following section.
Histological changes in the upper respiratory tract are well documented in animal studies
and have been observed in several worker studies (Section 4.4). Although the study of resin
production workers (Holmstrom and Wilhelmsson, 1988; Holmstrom et al., 1989) provides the
best documentation of effect level for this health category in humans, it is not carried through for
development of a candidate RfC. As with the sensory irritation endpoint reported in these
studies, exposure is described for the worker cohort by a simple mean, with a range of exposures
given for all workers. Therefore, these data do not provide an exposure-response relationship
and the POD would be the mean exposure level of all workers, regardless of effect. This is less
exact than other available studies which provide exposure-response relationships. Additionally,
animal studies provide a broad database which supports sensory irritation as a more sensitive
endpoint than histological changes in the nasal mucosa.
This document is a draft for review purposes only and does not constitute Agency policy.
5.35 DRAFT—DO NOT CITE OR QUOTE
-------
3 5
o 5.
^ sZ
Co SS
II
>}
§•
Cl
s
s
>}
a
Table 5-4. Summary of candidate studies for formaldehyde RfC development by health endpoint category
Health
Endpoint
Category
Study
Species
Setting
Children
Study
size
Formaldehyde
measurements
Specific
Endpoints
Observed
effects"
(ppb)
POD (ppb)
Sensory
Irritation
Liuetal. (1991)
Human
Residential
Yes
1,394
Two locations at one
time period (winter or
summer); 7-day
passive monitors
Eye irritation
95
LOAEL=95
Ritchie and
Lehnen (1987)
Human
Residential
Yes
2,007
Two locations at one
time period; 30-
minute sample
Eye, nose, and
throat sensory
irritation
200
NOAEL=50
Hanrahan et al.
(1984)
Human
Residential
Yes
(teenagers)
61
Two locations at one
time period; 60-
minute sample
10% increased
prevalence of
burning eyes
130
BMCL10=70
Holmstrom and
Wilhelmsson
(1988)
Human
Occupational
No
106
Several measurements
at factory
workstations taken
over 7 years
Eye irritation
210
NOAEL=70
Upper
Respiratory
Tract Pathology
Holmstrom and
Wilhelmsson
(1988);
Holmstrom et al.
(1989)
Human
Occupational
No
132
68 with
pathology
Several measurements
at factory
workstations taken
over 7 years
Loss of ciliated
epithelium;
goblet cell
hyperplasia;
squamous cell
metaplasia
240
LOAEL=240
Sensitization:
Asthma and
atopy
Garrett et al.
(1999)
Human
Residential
Yes
148
Four locations over up
to four time periods;
4-day passive
monitors
Increased
allergy;
increased
asthma-like
symptoms
28
LOAEL=28
Rumchev et al.
(2002)
Human
Residential
Yes
192
Two locations at two
time periods (Winter
& Summer); 8-hour
passive monitors
Initial diagnosis
of asthma
45
NOAEL=33
ck) o
^ a
,rj ^
^ S
o
§•
•3
Q\
a
&
H
o
o
-------
^ ^2
Co SS
II
ck} o
^5
0
1
s
I
§•
^n
3?
W
o-
C>
s
s
>!
%•
e5
Health
Endpoint
Category
Study
Species
Setting
Children
Study
size
Formaldehyde
measurements
Specific
Endpoints
Observed
effects"
(ppb)
POD (ppb)
Pulmonary
Function
Krzyzanowski et
al. (1990)
Human
Residential
Yes
208
Four locations over
two time periods
(opposite seasons);
7-day passive
monitors
10% Reduction
in PEFR
27
BMCL10=17
Neurological
Malek et al.
(2003a)
Rat
Laboratory
-
120
Intentional exposures
at specific levels
Impaired
learning
100
LOAEL=100
Reproductive
and
Developmental
effects
Taskinen et al.
(1999) (FDR)
Human
Occupational
No
602
Actual and surrogate
measurements
estimated by
occupational hygienist
Decreased
fecundity
density ratio
(FDR)
226b
NOAF.L=86
Taskinen et al.
(1999) (SAB)
Human
Occupational
No
602
Actual and surrogate
measurements
estimated by
occupational hygienist
Increased risk of
spontaneous
abortion (SAB)
26b
LOAEL=26
Immune
Function
Lyapina et al.
(2004)
Human
Occupational
No
29
Average shift
concentrations based
on measures 8-hour
exposures
Increased
respiratory tract
infections,
decreased
neutrophil
respiratory burst
activity
722
LOAEL=722
ij a This is the lowest level of exposure at which adverse effects were observed, the LOAEL, in effect, or the cut-off point for adversity for BMCLs.
^ b See Section 5.1.2.6.2 for methods to adjust exposure levels from Taskinen et al. (1999).
O
H
b
O
-------
1
2
3
4
5
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
Reduced pulmonary function is associated with formaldehyde exposure in several human
studies (students and workers). The best single study demonstrating decreased pulmonary
function is the moderate residential study by Krzyzanowski et al. (1990). The study was
specifically designed to include homes with children between the ages of 5-15. Results
presented for children (n = 208) provide an exposure-response relationship for reduced PEFR.
Data quality is considered high for this study, both in terms of the in-home exposure
measurements (7-day passive monitors, two time periods) and the contemporaneous in-home
measurement of pulmonary function. Sources of potential confounding or bias were considered
by the study authors and adequately taken into account in the study. Therefore, this study is
retained for derivation of a candidate RfC.
Several studies report increased asthma and/or allergic sensitization in children
associated with increased formaldehyde exposure in school or homes (Section 5.1.4). Of these,
two studies are further evaluated here (Garrett et al., 1999; Rumchev et al., 2002). The study by
Rumchev et al. (2002) is a case-control study of asthma incidence in children, and the study by
Garrett et al. (1999) is designed to study several related health effects (asthma, sensitization and
respiratory symptoms) in asthmatic and non-asthmatic children. Both studies measure in-home
formaldehyde levels with multi-day passive samples. Survey data and health outcome data are
considered of high quality in each study. Additionally, sources of potential confounding or bias
were considered by the study authors and adequately taken into account in the study. Therefore,
both studies are retained for derivation of a candidate RfCs. Although several studies of school
children support these findings, the residential studies were considered more appropriate for RfC
derivation because individual in-home formaldehyde levels were associated with the health
outcome data.
Multiple lines of evidence support the occurrence of neurotoxicity following exposure to
formaldehyde, however, none of the available human studies were considered to be of adequate
quality for derivation of a point of departure for use in quantitative assessment. Of the available
neurotoxicity studies, Malek et al. (2003a), in which impaired learning was seen in rats
following exposure at 100 ppb, was selected as a potential candidate for RfC development (see
Section 5.1.6). A NOAEL was not identified for this effect. In view of the other studies
available in the formaldehyde database (including multiple human studies of potentially sensitive
populations), and considering the uncertainty in extrapolating from the exposure conditions in
the Malek et al. (2003a) study (two hour exposures, repeated on ten consecutive days) to a
chronic exposure scenario, this study was not carried forward for derivation of a candidate RfC.
It is important to note that the resulting RfC may therefore not fully consider the documented
neurotoxic effects of formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
5-38 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Of the various reproductive and developmental effects associated with formaldehyde
exposure, reduced fecundity and increased risk of spontaneous abortions are primarily studied in
humans (Section 5.1.7). Of the available human studies, only one study provides individual
exposure estimates of adequate quality to support RfC development (Taskinen et al., 1999).
Exposure-response relationships for decreased fecundability density ratio and increased risk of
spontaneous abortions are seen with increased categories of worker exposures. Several potential
confounding exposures are evaluated in the study, and the association of decreased fecundability
density ratio observed in the study is most convincingly associated with increased formaldehyde
exposure (Taskinen et al., 1999). Potential sources of bias were also adequately addressed in the
study. This is considered a high quality study and is retained for cRfC derivation.
Although Lyapina et al. (2004) have documented decreased neutrophil respiratory burst
activity in exposed workers, the overall weight of evidence for deficit in immune function due to
formaldehyde exposure is weak. There is a trend for increased respiratory tract infections in
formaldehyde-exposed individuals, but it is a direct result of impaired immune function or,
perhaps, increased infection due to direct effects on the protective barriers of the nasal mucosa.
Animal studies do not support a finding of a deficit in immune function with formaldehyde
exposure. The study by Lyapina et al. (2004) is a small study, and the findings of decreased
neutrophil respiratory burst activity were in those individuals with more upper respiratory tract
infections, so there is some question of causality. The data evaluation does not provide an
exposure-response relationship, but, rather, exposure for the cohort is expressed as a mean
exposure of 722 ppb. Although the potential for impairment of immune function is an important
health effect, the overall evidence for this effect and this specific study are relatively weak
compared to other data available to support RfC derivation for formaldehyde. Therefore, this
study is not carried further in the quantitative analysis.
In summary, the best studies evaluated herein for the derivation of an RfC for
formaldehyde exposure and the related health effects are: 1) Sensory irritation (Liu et al.,1991;
Ritchie and Lehnen, 1987; Hanrahan et al., 1984); 2) reduced pulmonary function
(Krzyzanowski et al., 1990); 3) sensitization (atopy and asthma) (Garrett et al., 1999 and
Rumchev et al., 2002); and 4) reduced fecundity and increased spontaneous abortion (Taskinen
et al., 1999). It is recognized that not all identified health effects are represented in these
studies.
This document is a draft for review purposes only and does not constitute Agency policy.
5.39 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5.1.2.2. Derivation of Candidate RfCs from Key Studies
Candidate RfC derivation for Krzyzanowski et al. (1990) (Pulmonary function)
The study by Krzyzanowski et al. (1990) is a high quality epidemiology (human) study of
health effects in a random sample of residents and their families. The study was specifically
designed to include only households that had children 5-15 years of age, a sensitive life-stage for
respiratory effects. The study was of moderate size, when the effects in children were analyzed
separately from adults, with the final analysis based on 208 children—a cohort large enough to
show statistically significant results. The formaldehyde monitors were prepared by the
Lawrence Berkeley Laboratories and were considered to be precise and highly reliable. The 7-
day passive formaldehyde monitors generally provide the lowest limit of formaldehyde
detection. The investigators specifically tested an a priori hypothesis and conclusively
demonstrated to a high level of statistical significance that increased residential formaldehyde
exposures were associated with decreased pulmonary function as measured by peak expiratory
flow rate (PEFR) in children. This effect was clearly shown at relatively low concentrations of
formaldehyde as the mean concentration in the homes was 26 ppb with more than 83% of homes
having measured concentration less than 40 ppb. This study also reported specific regression
modeling results that allowed EPA to calculate the point of departure for RfC development using
a BMCL as the point of departure.
The effects of formaldehyde exposure on pulmonary function represent a sensitive
endpoint with a reported 10% reduction in PEFR at 27 ppb. Among children with physician-
diagnosed asthma, the observed effects of increased formaldehyde exposure on decreased PEFR
were more pronounced - a clear indication of variability in response. The American Thoracic
Society (ATS, 2000) considers decreased pulmonary function an adverse health effect, even
when it is transient and subclinical. "Assuming that the relationship between the risk factor and
the disease is causal, the committee considered that such a shift in the risk factor distribution,
and hence the risk profile of the exposed population, should be considered adverse, even in the
absence of the immediate occurrence of frank illness" (ATS, 2000). The ATS (2000) stated that
individuals in an exposed population experiencing a shift in the distribution of pulmonary
function were at potential risk from another agent due to the reduction in their reserve capacity to
address additional insults. In the study by Krzyzanowski et al. (1990), the investigators
demonstrated statistically significant interaction between formaldehyde exposures, smoking, and
chronic cough. That is, a formaldehyde concentration that caused decreased pulmonary function
at residential levels also caused chronic cough in the presence of environmental tobacco
exposures. Higher prevalence rates of physician-diagnosed asthma and chronic bronchitis were
This document is a draft for review purposes only and does not constitute Agency policy.
5-40 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
also shown at higher concentrations of formaldehyde (60-140 ppb), an effect that was
exacerbated by environmental tobacco exposures.
Figure 5-4 illustrates the reductions in peak expiratory flow rate (PEFR) in children (<15
years of age) in relation to indoor residential formaldehyde concentrations estimated by a
random effects model based on 3,021 observations in 208 subjects. Formaldehyde levels in the
home were significantly related to reductions in PEFR in children both at bedtime and in the
morning (p < 0.05). PEFR measurements in the morning versus at bedtime were significantly
different (p < 0.05). Formaldehyde-related reductions in PEFR were greater in the morning in
asthmatic children than in non-asthmatic children (p < 0.05).
— Bedtime
-- Morning, Non-Asthmatics
— Morning, Asthmatics
360
340
320
£ 300
E
^ 280
oc
260
Ll_
LU
Q_
240
220
200
0
20
40
60
80
100
HCHO(ppb)
Figure 5-4. Estimated reduction in peak expiratory flow rate (PEFR) in
children in relation to indoor residential formaldehyde concentrations.
Source: Krzyzanowski et al. (1990).
Candidate RfC derivation based on Krzyzanowski et al. (1990):
Critical effect: Based on this study, which specifically included a susceptible
population, the critical effect is reduction in PEFR in children. PEFR was the most
sensitive measure of disease or impaired lung function reported in this population, with
decreases in lung function reported in children who lived in homes with average
measured formaldehyde concentrations as low as 30 ppb (Krzyzanowski et al. (1990).
Children were more sensitive to formaldehyde-associated decreases in PEFR than adults,
so the cRfC derived focused on the results in the 208 children.
This document is a draft for review purposes only and does not constitute Agency policy.
5-41 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Point of departure: A BMR of 10% reduction in PEFR was selected as a cut-off point
for adversity, based on rationales articulated by the ATS (2000)4. Using this BMR and
the model coefficient in Table 5 of Krzyzanowski et al. (1990), a BMCLio of 17 ppb
(BMCio = 27 ppb) was derived for all children.5 Although the authors noted that
asthmatic children were more sensitive, the necessary data were not provided in the
report to calculate a BMCL for asthmatic children alone. Thus, 17 ppb, the BMCL based
on all children in the study, was used as the POD.
Application of study-specific Uncertainty Factors (UFs):
Interspecies UF = 1: No interspecies adjustment is needed, as this is a human study.
LOAEL-to-NOAEL UF = 1: Because a BMCL was used for the POD and the BMR of
10% reduction in PEFR was considered to be a cut point for adversity, no
LOAEL-to-NOAEL UF was needed (UFL = 1).
Subchronic-to-chronic UF = 1: The study addresses ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. It was judged that a population-
based study of residential exposures is sufficient to derive a chronic RfC without
adjusting for a subchronic observation period — at least for adults and older
children, and the children in this study were mostly older children (e.g., older than
7 years).
4
The ATS (2000) recommended that "a small, transient loss of lung function, by itself, should not automatically be
designated as adverse" and cited EPA's 1989 review of ozone, which offered a graded classification of lung function
changes in persons with asthma as "mild," "moderate," or "severe" for reductions of less than 10, 10-20, and more
than 20%, respectively (U.S. EPA, 1989). ATS (2000) concluded that, in evaluating the adverse health effects of air
pollution at the level of population health (compared to individual risk), "[a]ssuming that the relationship between
the risk factor and the disease is causal, the committee considered that such a shift in the risk factor distribution, and
hence the risk profile of the exposed population, should be considered adverse." This was specifically considered by
ATS (2000) even when "[e]xposure to air pollution could shift the distribution towards lower levels without bringing
any individual child to a level that is associated with clinically relevant consequences." A moderate adverse effect at
functional decrements of 10-20% was considered the best indicator of adverse effects in the study population. This
criterion had been similarly applied in EPA's Air Quality Criteria for Ozone and Related Photochemical Oxidants
(U.S. EPA, 2006d) for pulmonary function.
5 According to the regression model in Table 5 in Krzyzanowski et al. (1990), the coefficient ± standard error for
formaldehyde (in ppb) is -1.28 ± 0.46 and the background PEFR is 349.6 L/minute. Thus, a 10% reduction in PEFR
is -35 L/minute and the 95% (one-sided) upper bound on the slope for PEFR as a function of formaldehyde exposure
is -1.28 - (1.645 x 0.46), or -2.04 L/minute-ppb. Dividing 35 L/minute by 2.04 L/minute-ppb yields 17 ppb as the
BMCL.
This document is a draft for review purposes only and does not constitute Agency policy.
5-42 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Human variability UF = 3: The study was designed to include homes with children,
and a POD can be established based on reduced PEFR in children, who were
more sensitive to the health effects than the adults in the study. Therefore, the
POD represents data for a sensitive life stage, an aspect of human
(intraindividual) variability. With respect to the human (interindividual)
variability UF, although environmental tobacco smoke and socioeconomic status
did not affect the formaldehyde results in children, asthmatic children were more
sensitive to the effects of formaldehyde exposure on PEFR; thus, asthmatic
children represent a population with increased susceptibility for this effect. The
prevalence rate for physician-diagnosed asthma in the children was 15.8% in this
study, which is higher than the national prevalence of about 5.9% for ages 5 to
17 years.6 Thus the BMCL based on all children may be influenced by a higher
prevalence of susceptible children for the critical effect. The authors do report
that the PEFR was reduced to a greater degree in asthmatic children (as shown in
Figure 5-4), and a lower BMC of 17 ppb can be calculated in this subgroup versus
a BMC of 27 ppb for all children. However, the published regression statistics do
not provide sufficient detail to calculate a BMCL specific for asthmatic children.
In addition, other potentially sensitive populations (for example, elderly
individuals or individuals with respiratory diseases) may not be adequately
represented in the study. Therefore, an UF for human variability of 3 is applied to
address the observed increased sensitivity of asthmatic children in lieu of a
calculated BMCL specific to asthmatic children and to ensure adequate protection
for other potentially sensitive populations.
Rfr_ BMCLl0 _ 11 ppb
pFAxUFLxUFsxUFH) (lxlxlx3)
UFa = 1 (interspecies UF)
UFl= 1 (LOAEL-to-NOAEL UF)
UFS= 1 (subchronic-to-chronic UF)
UFh = 3 (human variability UF)
6 The national prevalence rate of asthma in children ages 5-17 is according to the Centers for Disease Control and
Prevention (CDC) (MMWR 49(40):908-911, 2000). Although the Krzyzanowski et al. (1990) study was conducted
in the late 1980s, prevalence data from the National Health Interview Survey for 1997 were used for comparison
because that is the earliest year for which data are available after a 1997 redesign of the survey. Previously, the
survey asthma question was not specific for physician-diagnosed asthma, so the redesigned results were considered
to be more comparable to the physician-diagnosed asthma definition in the Krzyzanowski et al. (1990) study.
This document is a draft for review purposes only and does not constitute Agency policy.
5.43 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5.1.2.2.1. Candidate RfC derivation for Rumchev et al. (2002) (Asthma)
Residential formaldehyde exposure was associated with an increased risk of asthma in a
population-based case-control study of 192 children aged 6 months to 3 years (Rumchev et al.,
2002). While it is acknowledged that accurately diagnosing asthma in young children is
difficult, as the diagnosing physician was unaware of the formaldehyde level in the children's
home, any diagnostic error would be unrelated to formaldehyde concentrations and would not
induce a spurious association. It is noted that the endpoint is physician-diagnosed asthma. The
study, which comprises 88 cases of children discharged from the emergency department of a
children's hospital in Perth, Australia, with a primary diagnosis of asthma and 104 controls,
provides a positive exposure-response relationship adequate for RfC derivation. Seasonal in-
home formaldehyde measurements taken in the living room and subject's bedroom were used to
assess exposure (8-hour passive sampler). The ORs for risk of asthma by formaldehyde
exposure level category were adjusted for numerous risk factors, both familial and
environmental, including familial history of asthma, age, sex, socioeconomic status, smoking,
presence of pets, air conditioning, humidifier, and gas appliances. Of these, age, allergic
sensitization to common allergens, and family history of allergy were independent risk factors
for asthma (OR = 1.09, 2.57, and 2.66, respectively). Odds ratios were further adjusted for the
effects of the measured indoor air pollutants (see Rumchev et al., 2004), indoor allergen levels of
dust mites, relative humidity, and indoor temperature. Categorical analysis of the data indicates
that the ORs for asthma were increased in the two highest formaldehyde exposure groups,
reaching statistical significance for household exposures > 60 |ig/m3 (48 ppb) (OR= 1.39)
(Figure 5-5). Analysis of the data with formaldehyde as a continuous variable provides a
statistically significant increase in the risk of asthma (3% increase in risk per every 10 |ig/m3
increase in formaldehyde level.)
5.1.2.2.2. Candidate RfC derivation based on Rumchev et al. (2002):
Critical effect: Diagnosis of childhood asthma (case-control study).
Point of departure: ANOAEL of 33 ppb (40 |ig/m3; midpoint of the 30-49 |ig/m3
category) was selected because the OR for asthma in the next highest exposure category
was considered to be part of an exposure-related trend of increasing asthma risk and,
therefore, biologically significant.
This document is a draft for review purposes only and does not constitute Agency policy.
5.44 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
,o
13
-o
"O
O
1.8-
1.6-
1 4-
1 2-
10-
08-
06-
10-29 30-49 50-59
Formaldehyde figm-3
60+
Figure 5-5. Odds ratios for physician-diagnosed asthma in children
associated with in-home formaldehyde levels in air.
Source: Rumchev et al. (2002).
Application of Study-Specific Uncertainty Factors (UFs):
Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
LOAEL-to-NOAEL UF = 1: No LOAEL-to-NOAEL UF was needed because the POD
was a NOAEL (UFL = 1).
Subchronic to chronic UF = 3: The study addresses ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. Study participants were 3 years or
younger, therefore the duration of exposure could not meet the expected
definition for a chronic study of one-tenth the lifespan. However, asthma often
develops during childhood, indicating a less-than chronic duration of exposure.
Since asthma may develop throughout childhood it is unclear whether a study of
children under 3 years of age would be of adequate duration for this
developmental window. Therefore, an uncertainty factor of 3 was applied as a
subchronic to chronic adjustment.
Human variability UF = 1 or 3: As a case-control study, all new cases of childhood
asthma which met the study criteria were eligible for inclusion and the cases
likely included children predisposed to asthma. Individuals with a family history
of asthma and/or genetic markers for genes believed to predispose individuals to
asthma would represent a susceptible population. Therefore, the cases in this
This document is a draft for review purposes only and does not constitute Agency policy.
5.45 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
study address children as a susceptible population for first diagnosis of asthma.
Additionally, there was an association of a familial history of asthma with the
diagnosis of children's asthma in this cohort (OR = 2.66). Not all sources of
human variability which may contribute to a diagnosis of asthma are known, and
there are likely additional sources of inter-individual variability among children
and among individuals with a family history of asthma, thus it is unlikely that all
sources of human variability were adequately represented in the study population
The two alternatives are described below and cRfCs are derivedfor each alternative.
Alternative A: Rumchev et al. (2002)
Human variability UF = 3:
To account for potentially susceptible individuals beyond those represented in the study
population, an uncertainty factor of 3 for human variability is applied.
NOAEL 33 ppb
(UFaxUFlxUFsxUFh) (1x1x3x3) 'PP
UFa = 1 (interspecies UF)
UFl = 1 (LOAEL-to-NOAEL UF)
UFS = 3 (subchronic-to-chronic UF)
UFh = 3 (human variability UF)
This document is a draft for review purposes only and does not constitute Agency policy.
5-46 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
Alternative B: Rumchev et al. (2002)
Human variability UF = 1:
EPA's Technical Report of the RfD and RfC Processes Technical Report (US EPA, 2002a)
indicates that UFH of 1 has been applied in cases where there are data "very specific
about the particular vulnerability of infants and children within certain age ranges to an
agent." Asthma and allergic sensitization to common allergens develop during childhood
and young adulthood defining a developmental window during which individuals are
most susceptible to the development of asthma. Since this study includes only children
up to 3 years of age, the UF for subchronic exposure is applied above acknowledging that
this study does not cover the susceptible developmental window. No additional
adjustment is applied for inter-individual variability among children. It is acknowledged
that additional sources of human variability are possible - but it is believed that
childhood is a key developmental window for initial diagnosis of asthma. The technical
report acknowledges that applying a UFH of 1 may be appropriate where "even within
these populations it is possible that some variability still exists.
NOAEL 33 ppb
{UFaxUFlxUFsxUFh) (lxlx3xl) PP
UFa = 1 (interspecies UF)
UFl = 1 (LOAEL-to-NOAEL UF)
UFS = 3 (subchronic-to-chronic UF)
UFh = 1 (human variability UF)
5.1.2.2.3. Candidate RfC derivation for Garrett et al. (1999) (Asthma, respiratory
symptoms, atopy and severity of allergic sensitization)
Garrett et al. (1999) reported on the risk of allergy and asthma-like respiratory symptoms
due to formaldehyde exposure in a cross-sectional survey of households with children 7-14
years old with (n = 53) or without (n = 95) doctor-diagnosed asthma. Formaldehyde exposure
was characterized by four seasonal in-home sampling events using 4-day passive samples
collected in bedrooms, living rooms, kitchens, and outdoors. In logistic regressions, both the
This document is a draft for review purposes only and does not constitute Agency policy.
5.47 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
prevalence and severity of allergic sensitization to 12 common allergens increased with
increasing formaldehyde concentration in the home. Additionally, a calculated respiratory
symptom score was increased and demonstrated a significant relationship with increased
formaldehyde concentration in a multiple linear regression after adjusting for multiple risk
factors and interactions. For each of these endpoints, severity/incidence was increased in the
medium (20-50 |ig/m3) and high (>50 |ig/m3) exposure groups relative to the low (<20 |ig/m3)
exposure group, based on the highest of four seasonal 4-day formaldehyde measurements in the
home (Figures 5-6 and 5-7).
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
44%
39%
16%
<20
20-50
>50
Exposure group, formaldehyde level
(ug/m3)
< 20 (ig/m3 20-50 ng/m3 > 50 jig/m3
Highest recorded formaldehyde level
Figure 5-6. Prevalence of asthma and respiratory symptom scores in children associated
with in-home formaldehyde levels. Trend analysis indicates statistical significance in these
increases {percent asthmatic children, unadjusted (p=0.03) and respiratory symptom score
(p=0.03)}.
Source: Garrett et al. (1999).
This document is a draft for review purposes only and does not constitute Agency policy.
5-48 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
80%
_ 70%
— 60%
C
£! 50%
2
¦£ 40%
O
o 30%
o" 20%
< 10%
0%
75%
33%
<20
20-50
>50
Exposure group, formaldehyde level
(ug/m3)
2 SO
E
| 40.
& 3 0'
JU
"5
60
2 0
1 0'
00L
Positive skin prick lestsj
Allergen wheal ratio
J
I
< 20 ng/ni' 20-50 Jig/m' > 50 |ig/m3
Highest recorded formaldehyde level
Figure 5-7. Prevalence and severity of allergic sensitization in children
associated with in-home formaldehyde levels. Trend analysis indicates
statistical significance in these increases {percent atopic children (p=0.002),
positive skin prick tests (p=0.001) and severity as allergen wheal ratio
(p=0.004)}.
Note: Skin prick tests included 12 environmental allergens (cat, dog, grass [two
types], house dust, dust mite [two strains] and fungi [five strains]).
Source: Garrett et al. (1999).
The findings of Garrett et al. (1999) are supported by the observation of an increased
bronchial responsiveness to mite allergen in a chamber study of 19 sensitized adult asthmatics
exposed to formaldehyde at a concentration of 100 |ig/m3 for 30 minutes (Casset et al., 2006).
Additionally, inhalation exposures to formaldehyde have been shown to increase an animal's
response to other common allergens via inhalation (Fujimaki et al., 2004; Sadakane et al., 2002;
Riedel et al., 1996; Tarkowski and Gorski, 1995).
Candidate RfC derivation for increased allergic sensitization from Garrett et al. (1999):
Critical effects: Allergic sensitization - Increase in allergic sensitization (proportion of
atopic children). Severity of allergic sensitization measured both as number of positive
skin tests to common allergens and the recorded allergen wheal ratio for those tests.
Asthma - increase in proportion of asthmatic children. Respiratory symptoms -
Increased respiratory symptom score.
Point of departure: For all critical effects, categorical analyses are presented that show
an increase in the mid-exposure group (16-40 ppb) and high exposure group (>40 ppb)
This document is a draft for review purposes only and does not constitute Agency policy.
5.49 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
relative to the low-exposure group (<16 ppb) (Figures 5-6 and 5-7). However, it is
unknown if the findings in the low-exposure group are comparable to the responses that
would be observed in an unexposed population. Therefore, the low-exposure group
cannot be considered a NOAEL but rather serves as a referent group for the two other
exposure groups. Thus, the LOAEL is based on health effects observed in the mid-
exposure group (16-40 ppb) for all three critical effects. As neither the mean or median
exposure levels are provided for the exposure categories used to analyze the health
effects data, the mid-point of the exposure category is selected for the LOAEL: 28 ppb.
Application of study-specific Uncertainty Factors (UFs):
Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
LOAEL-to-NOAEL UF = 3: As discussed, the mid-exposure group is selected as the
LOAEL since the low-exposure group is the referent group; there is no true
unexposed control. It is unclear whether or not a full LOAEL to NOAEL
uncertainty factor is warranted for these data. The authors did provide evidence
for increased atopy for every increase of 16 ppb of exposure with borderline
statistical significance when adjusted for several potential confounders (OR = 1.4;
95% CI: 0.98-2.00). An UF of 3 adjusts the LOAEL to a similar range and is
consistent with this alternative presentation of the data.
Subchronic to chronic UF = 1: The study addresses ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. It is judged that a population-based
study of residential exposures is sufficient for derivation of a chronic RfC without
adjusting for a subchronic observation period.
Human variability UF = 1 or 3: This study was designed to assess allergic
sensitization, asthma prevalence and respiratory symptoms in children with
relation to in-home formaldehyde levels. The recruitment of participants was
designed to include households (50%) with asthmatic children, resulting in
43 households with at least one asthmatic child and 37 without asthmatic children
for a total of 148 children (35% asthmatic). Parental allergy and asthma were
also assessed and included as adjustment variables in the data evaluation.
Therefore the study population includes individuals reflecting several key aspects
of human variability for asthma and allergic sensitization (age, familial history of
disease), and addresses the links between allergic sensitization and asthma. Both
This document is a draft for review purposes only and does not constitute Agency policy.
5-50 DRAFT—DO NOT CITE OR QUOTE
-------
asthma and allergic sensitization are risk factors for increased respiratory
symptoms.
The two alternatives are described below and cRfCs derived for each alternative
Alternative A: Garrett et al. (1999)
Human variability UF = 3: It is unclear whether the effect levels in the study truly
reflect the effect levels in sensitive populations, since study findings controlled for
both asthma and family history. Therefore, a value of 3 was used for the human
variability UF.
RfC
LOAEL
28 ppb
"> 8 ppb
UFa = 1 (interspecies UF)
UFl= 3 (LOAEL-to-NOAEL UF)
UFS= 1 (subchronic-to-chronic UF)
UFh= 3 (human variability UF)
This document is a draft for review purposes only and does not constitute Agency policy.
5.51 DRAFT—DO NOT CITE OR QUOTE
-------
Alternative B: Garrett et al. (1999)
Human variability UF = 1:
Individuals with a family history of asthma and/or genetic markers for genes are
believed to be predisposed to asthma and this would define a susceptible population within
children. In this study parental disease status is a marker for potential genetic susceptibility.
Although exposure-response relationships are not provided for individuals with a familial
history of disease, analyses provided suggest the results reflect responses from these
individuals. Among children with parental allergy, allergic children were exposed to higher
formaldehyde levels than non-allergic children (p = 0.02), relating higher formaldehyde
exposure to sensitization even among those with a likely genetic susceptibility. As shown in
Figure 5-8, formaldehyde levels are related to increased asthma incidence with a significant
linear trend (p = 0.02), yet this relationship loses significance when controlling for parental
allergy and asthma, suggesting the measured response on which the POD is based is driven
by children with a potential for genetic susceptibility.
An EPA Technical Report of the RfD and RfC Processes (US EPA, 2002a) indicates
that a UFh of 1 can be applied in cases where data are "very specific about the particular
vulnerability of infants and children within certain age ranges to an agent." Asthma and
allergic sensitization to common allergens develop during childhood and young adulthood.
Therefore no additional adjustment is applied for human variability. The technical report
acknowledges that "even within these populations it is possible that some variability still
exists", but that a UFH of 1 is still applied.
RfC
LOAEL
28 ppb
9 3 ppb
UFa = 1 (interspecies UF)
UFl= 3 (LOAEL-to-NOAEL UF)
UFS = 1 (subchronic-to-chronic UF)
UFh = 1 (human variability UF)
1
This document is a draft for review purposes only and does not constitute Agency policy.
5-52 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
5.1.2.2.4. Candidate RfC derivation for Ritchie and Lehnen, 1987; Hanrahan et al., 1984
and Liu et al., 1991 (Sensory irritation).
There are three studies that report sensory irritation in humans from chronic exposures in
a residential environment and provide sufficient exposure data to support quantitative assessment
(Liu et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). Each study reports site-
specific exposure measurements and presents some metric of individual exposure. These
residential studies employ in-home measurements for each study participant, either as average
exposure level (Ritchie and Lehnen, 1987; Hanrahan et al., 1984) or as calculated cumulative
exposure based on the time in the home (Liu et al., 1991). Eye irritation is reported at similar
levels of residential formaldehyde exposure in the three studies (Figures 5-8 and 5-9). Each
study provides an exposure-response relationship for prevalence of sensory irritation in relation
to in-home formaldehyde exposure based on individual level data.
FORMHLOmDE CONCENTRRT I ON IN FPU
Panel A: Regression of prevalence of eye irritation
versus indoor formaldehyde concentration (ppm) in
mobile homes (30-60 minute air sample in each
home).
Note: Dashed lines show upper and lower 95th percentile
confidence intervals on model results. Model based on
reported eye irritation from individuals in 42 mobile homes.
ro 3Q%
Qi 20%
= 100 100-300 >300
Formaldehyde Concentration (ppb)
Panel B: Prevalence of eye irritation in groups
defined by in-home formaldehyde exposure (30-
60 minute air sample in each mobile home).
Note: Eye irritation rate is given by smoking status:
active smokers (n = 143), passive exposure to smoke
(n = 133) and nonsmokers (n = 180).
Data source: Ritchie and Lehnen (1987).
Figure 5-8. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (eye irritation).
This document is a draft for review purposes only and does not constitute Agency policy.
5.53 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25% i
~ Summer
¦ Winter
<7 7-12 >12
Formaldehyde Exposure (ppm-hr per week)
Figure 5-9. Positive exposure-response relationships reported for in-home
formaldehyde exposures and sensory irritation (burning eyes).
Note: Cumulative formaldehyde exposure was estimated for each participant
from measured in-home formaldehyde levels (7-day passive air sample) and
reported hours spent in the home. Prevalence rates are given for both summer
(n = 1,388) and winter (n = 1,093) survey periods.
Data source: Liu et al. (1991).
Ritchie and Lehnen (1987) examined formaldehyde-associated effects on eye, nose, and
throat irritation in a large residential study with 2,007 participants from 841 homes. Based on
in-home measurements of formaldehyde concentration, participants were categorized into three
exposure groups: low (<100 ppb), mid (100-300 ppb) and high (>300 ppb) (average of two 30-
60 minute air samples per home). Ritchie and Lehnen (1987) observed clear exposure-response
relationships in the percentage of residential occupants reporting eye, nose, and throat irritation.
For example, in nonsmoking mobile home residents, incidence scores for eye irritation were 1%,
18% and 86%, and for nose/throat irritation were 5%, 17% and 78%, respectively, for the three
exposure groups. The exposure-response relationships were similar regardless of type of home,
mobile (n = 851) or conventional (n = 1,156). Although smoking status was also a predictor of
irritation, in-home formaldehyde concentrations were a stronger predictor of health effects. The
study included children and the elderly and results were consistent across age groups. Children
This document is a draft for review purposes only and does not constitute Agency policy.
5.54 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
<7 years of age were only included in the eye irritation analyses because of concerns about the
quality of parental reporting for nose and throat effects in young children. The selection criteria
for participants indicate that more sensitive individuals may have been over-represented in the
study population.7 All study participants were self-selected, with a physician's approval,
perhaps resulting in a higher proportion of individuals experiencing various irritant and upper
respiratory tract symptoms, which may represent a sensitive population for eye, nose, or throat
irritation.
Hanrahan et al. (1984) reported an exposure-response relationship for burning eyes and
eye irritation in a study of 61 teenage and adult residents of mobile homes. As in the Ritchie and
Lehnen (1987) study, in-home formaldehyde measurements were obtained for all participants
and measured formaldehyde levels were used to characterize average in-home exposures (30-
60 minute air sample). Eye irritation was associated with in-home formaldehyde exposures
(p < 0.05) (both as "burning eyes" and "eye irritation"), and the authors provided a graphical
representation of the best-fitting regression model for exposures between 100 and 800 ppb.
From inspection of this graph, the prevalence of eye irritation predicted at 100 ppb is
approximately 4% with an upper bound of 18% (95th percentile CI) (Figure 5-8, Panel A).
Because the limit of detection for formaldehyde in indoor air was 100 ppb, data or model results
are not provided below 100 ppb.
The third residential study is a random-sample study of over 1,000 mobile home residents
(1,394 in the summer; 1,096 in the winter) that included both young children and the elderly (Liu
et al., 1991). Cumulative weekly exposures were based on in-home formaldehyde sampling and
a participant survey of time spent at home. Air sampling was conducted for a 7-day period using
a passive sampler in each home (summer and winter). The resulting estimates of cumulative
exposure assumed no formaldehyde exposure outside of the home. Cumulative formaldehyde
exposure was a significant predictor of numerous irritant symptoms in a multivariate linear
logistic regression, including "burning eyes" (p < 0.05). The prevalence of eye irritation
increased with increasing cumulative exposure in a categorical analysis of participants 20-64
years old for both summer and winter exposure estimates (Figure 5-9). Eye irritation was above
10% in the lowest exposure group (0-7.0 ppm-hours/week) and increased to 17.1% and 21.4 %
in the mid- and high-exposure group, respectively, for the summer survey time; winter rates were
slightly lower but showed a similar increase with increasing cumulative exposure.
7
Participants in this study were self-selected residents who were concerned about possible formaldehyde exposure
and had obtained a written request from a physician to have the Minnesota Department of Health test their homes as
part of a free program; thus, people with symptoms may be overrepresented in this study compared with the general
population. This potential overrepresentation does not necessarily imply a selection bias because it is unlikely that it
was associated with the measured formaldehyde exposure levels in participants' homes.
This document is a draft for review purposes only and does not constitute Agency policy.
5-55 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Taken together, these three studies report increased eye irritation from residential
exposures that are below the BMCLs calculated from acute exposures in the laboratory. Each
study has the strength of having individual in-home exposure measurements and demonstrates a
positive exposure-response relationship for sensory irritation within a range of residential
formaldehyde exposures (both conventional and mobile homes). Potentially confounding factors
(such as allergens and some other in-home exposures) have been taken into account and
statistical analyses of the data include relevant covariates (e.g., age, sex, smoking status). As
such, these studies provide a basis for development of a cRfC for sensory irritation.
Additionally, the study populations have been drawn from the general population, including
children and the elderly, and have not been limited to those healthy enough for full-time
employment (as is often the case in occupational cohorts).
All three studies support a finding of increased eye irritation for exposures above 100 ppb
(Figures 5-8 and 5-9). However, the shape of the exposure-response curve below 100 ppb, or an
indication of a no-effect level, is less clear. Two of the studies indicate 1-4% eye irritation in
residents where formaldehyde exposures were measured at 100 ppb or less (Ritchie and Lehnen,
1987; Hanrahan et al., 1984). Thus, there is uncertainty in considering 100 ppb as a no-effect
level for increased eye irritation for these studies. When modeled, the 95% CIs around the point
estimate of 4% eye irritation were 1—18% eye irritation, illustrating the range of response rates at
100 ppb that are consistent with the observed data (Hanrahan et al., 1984). Additionally, the
presentation of results by exposure category in Ritchie and Lehnen (1987) is inexact and has
individuals with exposures at the low end of the categorical range being grouped with those at
higher exposures in the range, obscuring any exposure-response relationship within the
categorical range. For these reasons, a POD for RfC derivation from either of these studies
should reflect these uncertainties. Therefore, for the NOAEL representing the category of
individuals with <100 ppb, in which 1-2 % eye irritation was observed, the upper end of this
exposure category is not used, but rather the midpoint, 50 ppb (Ritchie and Lehnen, 1987).
Although Hanrahan et al. (1984) provided no model results below 100 ppb, an extrapolation of
the graphical results (Figure 5-8, Panel A) provides an estimated BMCLio of 70 ppb8. No
additional duration adjustments were made from the in-home exposure measurements to
g
Figure 1 of Hanrahan et al. (1984) shows predicted values and 95% confidence intervals (CIs) for the percent
prevalence of a burning-eyes response for formaldehyde concentrations >100 ppb (See Panel A in Figure 5-9 above).
A short extension of the upper 95% CI to the concentration associated with 13% prevalence (i.e., a 10% increased
prevalence above an assumed background response rate of 3%; this assumed background rate was chosen to be
conservatively high to err on the side of not underestimating the actual value, given that the value was approximated
from a visual extension of the upper 95% CI curve) suggests a BMCL of approximately 70 ppb for 10% increased
prevalence. The actual value is unknown but is clearly below 100 ppb, which is the minimum exposure
concentration depicted in the figure.
This document is a draft for review purposes only and does not constitute Agency policy.
5-56 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
continuous exposure because neither time away from the home, nor potential exposures outside
of the home, were characterized in either study.
Of the three studies, only Liu et al. (1991) provides exposure measurements below
100 ppb, with a reported detection limit of 10 ppb formaldehyde for the in-home air monitoring.
Additionally, air samples were collected using a 7-day passive sampler which is more
representative of average residential exposures than a one-time, 30-60 minute, air sample.
Therefore, the data collected by Liu et al. (1991) are more suited to understanding the exposure-
response relationship for eye irritation of exposures below 100 ppb. In addition to controlling
for age, gender, and smoking status, Liu et al. (1991) controlled for the presence of chronic
respiratory disease when assessing the effects of formaldehyde on symptoms of sensory
irritation. Finally, this study provides results for both summer and winter survey periods,
addressing seasonal variation in both formaldehyde levels and sensory irritation. The use of the
cumulative exposure metric considers not only the concentration of formaldehyde but also the
number of hours during the week each participant spent in their residence. Linear logistic
regression indicates that cumulative formaldehyde exposure was a statistically significant
predictor of burning eyes for both winter and summer survey periods. However, no BMCL can
be calculated because no regression coefficients were provided in the report. Data were
provided for the categorical analysis illustrating a positive exposure-response relationship
(redrawn in Figure 5-9). Based on the categorical results, the mid-exposure group (7-12 ppm-
hours/week) demonstrated an increased response compared with the low-exposed group. Since
the prevalence rate in the low-exposed group was above 10% for burning eyes, this exposure
group does not represent a NOAEL, but rather serves as a referent for the mid-exposure group.
Therefore, the POD is derived from the midpoint of 7-12 ppm-hours/week, 9.5 ppm-hours/week.
Using a conversion factor applied by the authors, the cumulative exposure of this mid-exposure
group corresponds to a continuous home exposure of 70-120 ppb for an individual who spends
60% of the week in the home, with a mid-point of 95 ppb.
Candidate RfC derivation for sensory irritation:
Critical effect: Prevalence of sensory irritation (eye irritation, burning eyes).
Point of departure: Each of the studies discussed above has different strengths and
weaknesses for the determination of a POD for sensory irritation. Nevertheless, the
effect levels and PODs derived from each study are in relatively close agreement with
less than a twofold span from lowest to highest. Therefore each POD is carried through
to calculate a cRfC:
This document is a draft for review purposes only and does not constitute Agency policy.
5.57 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
NOAEL
BMCLio
LOAEL
= 50 ppb
i = 70 ppb
= 95 ppb
(Ritchie and Lehnen, 1987)
(Hanrahan et al., 1984)
(Liu et al., 1991)
Application of Uncertainty Factors (UFs)
Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
LOAEL-to-NOAEL UF: An uncertainty factor of 1 is applied to the NOAEL and
BMCL io established as PODs from Ritchie and Lehnen, (1987) Hanrahan et al.
(1984) studies. An uncertainty factor of 3 is applied to the LOAEL of 95 ppb
based on the Liu et al. (1991) study, as the prevalence rates for this exposure level
are below 20% for an effect that is of relatively low severity. In addition, the
LOAEL is not significantly above the NOAEL and BMCLio from the other
studies that evaluated the same endpoint.
Subchronic to chronic UF = 1: These studies address ongoing residential exposure to
formaldehyde. Although information on the duration of exposure for each
participant is not provided, the residential nature of the study suggests a longer
term exposure than the duration of the study. It is judged that a population-based
study of residential exposures is sufficient for derivation of a chronic RfC without
adjusting for a subchronic observation period.
Human variability UF = 1 or 3: All three studies were population-based and included
children, the elderly and both sexes. Sample sizes for two of the studies were
very large (1,394 for Liu et al. [1991]; 2,007 for Ritchie and Lehnen [1987]),
increasing the likelihood that sensitive populations were included. Analysis of
the data controlled for sex, smoking status, and age group.
This document is a draft for review purposes only and does not constitute Agency policy.
5-58 DRAFT—DO NOT CITE OR QUOTE
-------
1 The two alternatives are described below and cRfCs derived for each alternative
2
Alternative A
Sensory irritation studies:
Human variability UF = 3: For all studies, the analysis was based on prevalence rates,
decreasing the likelihood that effects on sensitive individuals would be lost due to
response averaging. For Ritchie and Lehnen (1987), the prevalence rate in the
<100 ppb exposure group (represented by a NOAEL of 50 ppb, the midpoint) was 1-
4%. For Hanrahan et al. (1984), the POD is a
BMCL corresponding to a 10%
response rate. Given these prevalence rates and the fact that the sensory irritation
effects assessed are considered minimally adverse, a human variability UF of 3 was
considered adequate for this endpoint.
Ritchie and Lehnen (1987):
NOAEL
~ x x x ) ~
50 ppb
"h i f ^=17 ppb
(1x1x1x3)
Hanrahan et al. (1984):
Rrc BMCL„
(UFaxUFlxUFsxUFh)
70 ppb
i r i\ = 23PPb
(1x1x1x3)
Liu et al. (1991):
s LOAEL
~(E/F(xE/FIxE/FsxE/Fj~
95 ppb
h 2 1 ,\ = 9-5PPb
(1x3x1x3)
4
5
6
This document is a draft for review purposes only and does not constitute Agency policy.
5.59 DRAFT—DO NOT CITE OR QUOTE
-------
1
Alternative B
Sensory irritation studies
Human variability UF = 1: Two studies included a broad age range allowing some
assessment of human variability due to life stage. Ritchie and Lehnen (1987)
evaluated the influence of age on sensory irritation in the following age groups
65 years.
An age effect for eye irritation was not evident in these data and pooled data are
presented for this endpoint. Liu et al. (1991) report that greater eye irritation was
reported in participants of 20-64 years than in those younger than 20 or older than
65 years. The elderly population (>65 years) was well-represented in this study (39%
of participants in the summer and 34% in the winter). The modeled results on which
the BMCLio is based for Hanrahan et al. (1984) are normalized to 48 years of age (the
mean age of respondents), which is consistent with the age group considered the most
responsive in the Liu et al. (1999) study. Therefore the PODs derived from these
studies do account somewhat for human variability across the life stage.
The critical effects of sensory irritation (eye, nose, and throat irritation) are considered
minimally adverse health effects. The nominal response rates for eye irritation of
1-4% for in-home exposures below 100 ppb from which the PODs were derived
suggest that the PODs are below significant response levels. Additionally, as the data
are reported as prevalence rates, there is no masking of effect from sensitive
individuals (as may occur when benchmark responses are average values of biometric
parameters).
Finally, sensory irritation is a POE effect. Therefore, sources of human variability such as
absorption, distribution, and metabolism of a compound are unlikely to influence
incidence rates for this endpoint. There may be human variability in the sensitivity of
the trigeminal nerve to formaldehyde binding and stimulation.
Taken together, these studies address many potential sources of human variability.
Therefore, it is judged that further adjustment to address human variability is not
warranted for the minimally adverse health effect of sensory irritation. Thus a UFH of
1 is applied to all three studies. It is acknowledged that there is the potential for
sources of variability not captured in these studies.
(Continued on next page.)
This document is a draft for review purposes only and does not constitute Agency policy.
5-60 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Continued from previous page:
Ritchie and Lehnen (1987):
NOAEL _ 50 ppb
{UFAxUFLxUFsxUFH) (lxlxlxl) PP
Hanrahan et al. (1984):
IQppb
(UFa X UFl X UFS X UFh ) (lxlxlxl)
Liu et al (1991):
_ LOAEL _ 95 ppb
(lJFa xUFl xUFs xUFh ) ~ (l x 3 x 1 x l) ~ PP
5.1.2.2.5. Candidate RfC derivation for Taskinen et al. (1999) (Fecundity ratio)
On review of the candidate developmental and reproductive toxicity studies in humans
and animals (presented in Section 5.1.3.2.7), the Taskinen et al. (1999) human study was
considered to be the strongest for the purpose of deriving a chronic RfC. This study was a well-
designed population-based case-control study of women who were occupationally exposed to
formaldehyde. The study population was well defined and adequately selected to allow for
meaningful comparisons of health effects among individuals with different levels of exposure to
formaldehyde. Potential selection bias and the self-reporting of spontaneous abortion are not
considered to have had a significant influence on the study findings. Additionally, the decreased
FDR and increased risk of spontaneous abortion observed in Taskinen et al. (1999) are internally
consistent and coherent with other reports of increased risk of pregnancy loss associated with
exposure to formaldehyde (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990;
Axelsson et al., 1984) and is supported by animal data (Kitaev et al., 1984; Sheveleva, 1971).
The Taskinen et al. (1999) study allows the consideration of three potential critical
effects: endometriosis, increased spontaneous abortion, and decreased FDR. However, there is
little independent support for the finding of increased endometriosis and the ORs for organic
solvent exposure within this study (OR = 14.7; 95% CI: 3.1-70) were much greater than for
formaldehyde (OR = 4.5, 95% CI: 1.0-20), indicating a potential for confounding. Both
increased spontaneous abortions and decreased FDR are supported by independent findings in
This document is a draft for review purposes only and does not constitute Agency policy.
5-61 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
other formaldehyde-exposed cohorts (John et al., 1994; Taskinen et al., 1994; Seitz and Baron,
1990; Axelsson et al., 1984). As this study was designed to examine the effect of workplace
formaldehyde exposures on FDR, the study design and data collection best support this finding.
The exposure measurements were conducted to represent what the researchers considered the
relevant time-to-pregnancy exposures. Although data on miscarriages were collected to control
the time-to-pregnancy findings for confounding from formaldehyde-related spontaneous
abortions, it is less certain that the exposure measurements coincide with the defined
spontaneous abortion cases. Spontaneous abortions were only included in calculations of
exposure-specific ORs if a participant indicated that she was employed at the same location
when she had the spontaneous abortion and when the time-to-pregnancy exposure assessment
was done. The analysis showed that there were statistically significantly increased risks of
spontaneous abortion in the lowest exposure group. While this finding was consistent with other
studies showing adverse reproductive effects of formaldehyde and appears to be causal, the
Taskinen et al. (1999) spontaneous abortion results did not clearly control for all the potential
confounders that were controlled for in the FDR analyses (i.e. organic solvents and phenols).
While the other coexposures were not associated with FDR and therefore not confounders,
endometriosis was strongly associated with organic solvents. Therefore, for these endpoints, the
study design and strength of results best support the use of decreased FDR in formaldehyde-
exposed women as the critical effect for this study.
It is preferable that the critical effect be the most sensitive of the effects which is well
supported by the given study. As spontaneous abortions are significantly increased in the low-
exposure group and the response in the mid-exposure group is considered a no-effect level for
decreased FDR, there is uncertainty that an RfC based on the FDR NOAEL would be protective
for the more sensitive effect. Although qualitatively the finding of increased spontaneous
abortion is convincing, there is more uncertainty in the applicability of the exposure assessment
for quantitative risk assessment. Additionally, there is greater uncertainty in the use of the
exposure adjustments for the low-exposure group on which the LOAEL is based because they
account for more of the work time in the low-exposure group than the medium and high
exposure groups (Table 5-5).
There are several sources of uncertainty in the exposure estimates for use in RfC
derivation. As discussed above, the average exposure estimate for the low exposure group
includes a greater proportion of non-assessed background exposures. This is evidenced in part
by the reported average exposure being below background levels for these workers, even with
exposure measurements as high as 300 ppb. The unaccounted for non-task exposures may
represent time during the day spent in the work facility, or time in a different job or work
This document is a draft for review purposes only and does not constitute Agency policy.
5-62 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
environment. Additionally, task-level exposure measurements were available for only 27% of
women in the low exposure group, versus 38% and 69% of women in the medium and high
exposure groups, indicating less certainty in exposure classification for the low exposure group.
Duration adjustment for candidate study points of departure:
Normally, exposures from occupational studies are adjusted to account for the daily
breathing volume appropriate to an environmental (versus occupational) setting and for exposure
every day of the year (EPA, 1993). However, with formaldehyde, there is potential for exposure
outside of work from in-home and environmental sources of formaldehyde (Chapter 2). A
contemporaneous study of formaldehyde exposures in Finland reports average exposure of 21.4
ppb (measured over 48 hours with a personal monitor) (Jurvelin et al., 2001). Furthermore, both
the mean exposure (18 ppb 8hr TWA) and lowest reported exposure (10 ppb 8hr TWA) of the
'low exposed' category are below the reported average ambient exposures for Finland (21.4
ppb). Thus, it is likely that exposure estimates for study participants include time during the
workday when women reported no formaldehyde exposure and a zero exposure was assessed for
a non-formaldehyde related task. Additionally, participants may have qualified for the study
based on employment date but may not have been working with formaldehyde during the entire
time-to-pregnancy period. In both cases, the investigators in Taskinen et al. (1999) appear to
have assumed that, while the women were away from their "exposed" workplace, their exposure
to formaldehyde was zero, not accounting for background occupational exposures and ambient
levels of formaldehyde. This explains why both the mean exposure as well as lower end of
workshift exposures for women in the low exposure group were reported at and below expected
ambient levels. The women in the low exposure category had task-level workplace exposures of
up to 300 ppb in addition to experiencing some work time at background exposure levels.
Compared to women who only experienced background exposure levels, those in the low
exposure category were at significantly higher risk of spontaneous abortion.
The reported data do not provide information to correct for background formaldehyde
exposure during the workday for each participant. However, the published mean exposure
values may be used to provide some idea of the impact of including background exposures on
the study PODs. Comparison of the values listed in Table 4 of Taskinen et al. (1999) allows for
the estimation of the percentage of work time spent performing tasks involving formaldehyde
exposure (Table 5-5, Panel A). For the women in the low exposure category, this percentage is
26%) (mean of measured workplace exposures of 70 ppb times 26%> equals the mean of the TWA
exposure of 18 ppb). Using the same method, the women in the "medium" and "high" exposure
category were performing tasks involving formaldehyde exposure approximately 54%> and 66%>
This document is a draft for review purposes only and does not constitute Agency policy.
5-63 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
of their work time, respectively. Assuming that the women spent the remainder of their work
time at the background concentration of 21.4 ppb (Jurvelin et al., 2001), a more appropriate
estimate of the women's 8-hour TWA formaldehyde exposures would be 34 ppb for the low
category, 86 ppb for the medium category, and 226 ppb for the high category (Table 5-5, Panel
B).
Table 5-5: Adjustment for nonoccupational exposures to formaldehyde.
Panel A: Proportion of workshift corresponding to the exposure group mean
task-level formaldehyde exposure (ppb) and the exposure group daily exposure
index (8 Hr-TWA).
Exposure group (n)
Reported mean
exposure
(ppb, 8 hr-TWA)
Measured task-level
exposures
(PPb)
Estimate of time
during workday for
formaldehyde related
tasks assuming mean
exposure levels.
Mean
Range
Mean
Range
% of
worktime3
Hours per
8 Hr
workshift
Low (119)
18
1-39
70
10-300
26%
2
Medium (77)
76
40-129
140
50-400
54%
4.3
High (39)
219
130-630
330
150-1000
66%
5.3
a: Calculated as mean exposure (ppb 8Hr-TWA) divided by mean task-level exposures for the exposure group.
Panel B: Recalculation of daily exposure index (8 Hr -TWA) where background
formaldehyde exposure is estimated for worktime spent on tasks considered
unrelated to occupational use of formaldehyde.
Exposure
group (n)
Estimate o
exposure dur
relatet
f Formaldehyde
ing formaldehyde-
work tasks
Estimate of f(
exposure from
levels during
>rmaldehyde
background
the workshift
Alternativ
e daily
exposure
index
(ppb, 8
Hr-TWA)
Mean task
level
exposure
(ppb)
% of worktime in
formaldehyde
task
Background
formaldehyde
(PPb)
% of time in
non-
formaldehyd
e-related task
Low (119)
70
26%
21.4
74%
34
Medium
(77)
140
54%
21.4
46%
86
High (39)
330
66%
21.4
34%
226
This document is a draft for review purposes only and does not constitute Agency policy.
5-64 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Candidate RfC derivation for Taskinen et al, 1999:
Critical effect: Decreased FDR.
Point of departure: For decreased FDR, the mid-exposure level is considered a
NOAEL. The mean exposure as an 8-hour TWA for the workday is reported as 76 ppb.
EPA has adjusted this POD to account for potential background formaldehyde exposures
during the workshift (Table 5-5) resulting in an adjusted POD of 86 ppb. No further
duration adjustment is made to this POD to account for background levels of
formaldehyde exposure outside of the workplace.
Application of study-specific Uncertainty Factors (UFs):
Interspecies UF = 1: No interspecies adjustment is needed as this is a human study.
LOAEL-to-NOAEL UF = 1: Selection of an NOAEL as the POD.
Subchronic to chronic UF = 1: The study design represents a study population with a
range of exposure durations, including chronic exposures. By drawing the study
population from full-time employees and members of the wood-working union,
there is an expectation that the study population reflects the demographic of that
group as a whole. Although specific summary information is not published for
this study group (e.g., average length of employment), the lack of this reporting in
itself does not seem to justify an UF for subchronic-to-chronic exposure given the
overall study design. As a study adequate for assessing reproductive effects in a
chronically exposed cohort, no further adjustment was considered needed.
Human variability UF = 10: The study population included women employed in the
wood-working industry who were healthy enough to be gainfully employed.
Additionally, study inclusion criteria ensured that all study participants had at
least one pregnancy resulting in a live birth during the study period (1985-1995).
Therefore, these women were reproductively successful. The authors judged that
selective participation did not influence potential confounders such as irregular
menstruation or earlier miscarriages, which could impact the time to pregnancy
results. Susceptible populations were not addressed and, in fact, the women in the
study may be considered healthier than the general population in terms of
reproductive health. Therefore, an uncertainty factor of 10 for human variability
was applied.
This document is a draft for review purposes only and does not constitute Agency policy.
5-65 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
mr NOAEL 86 ppb
(UFa xUFl xUFS xUFh) (lxlxlxlO) ' PP
UFa = 1 (interspecies UF)
UFl= 1 (LOAEL-to-NOAEL UF)
UFS = 1 (subchronic to chronic UF)
UFh= 10 (human variability UF)
5.1.2.3. Evaluation of the Study-Specific Candidate RfCs
Seven studies were selected as key studies for consideration in RfC derivation (Section
5.1.2, Table 5-4). Candidate RfCs from these studies address various health effects including:
sensory irritation, respiratory effects, asthma, increased allergic sensitization, and decreased
fecundity (Table 5-6).
Three of the seven studies address sensory irritation of the eye, nose, and throat (Liu
et al., 1991; Ritchie and Lehnen, 1987; Hanrahan et al., 1984). The PODs for sensory irritation
range from 50 to 95 ppb for a health effect that is considered minimally adverse.
Two alternatives are presented for the human variability uncertainty factor in RfC derivation
based on these SI studies. Alternative A (UFH=3) results in cRfCs from 9.5 to 23 ppb.
Alternative B (UFH=1) results in cRfCs from 32 to 70 ppb.
A cRfC of 9 ppb is derived for decreased FDR in an occupational study of women in the
wood-working industry (Taskinen et al., 1999). This endpoint is supported by four other
epidemiologic studies and is considered a potential health concern for occupationally exposed
women (John et al., 1994; Taskinen et al., 1994; Seitz and Baron, 1990; Axelsson et al., 1984).
However, there is some uncertainty regarding the influence of peak exposures in the work place
on the apparent exposure-response relationship based on average workday exposures calculated
for study participants. It is unknown if the observed decreased FDR can be attributed to the
average exposures from which the cRfC is derived or if it is a result of the measured exposures
(as high as 1,000 ppb). If this were the case the cRfC of 9 ppb, based on the average time-
weighted exposures, would be protective for decreased fecundity.
Three studies identify adverse health effects in residential populations including children:
increased incidence of asthma, decreased pulmonary function, increase in respiratory symptoms,
and increased allergic sensitization (Rumchev et al., 2002; Garrett et al., 1999; Krzyzanowski
et al., 1999). Asthma, allergic sensitization, pulmonary function, and symptoms of respiratory
disease are not only clinically related, but etiologically related, and it is reasonable that they are
This document is a draft for review purposes only and does not constitute Agency policy.
5-66 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
considered together from a public health perspective. These health effects are observed below
the exposure levels that result in sensory irritation and the resulting cRfCs are correspondingly
lower, in a range between 2.8 and 11 ppb, depending on the study, endpoint considered, and the
application of alternative uncertainty factors for human variability (Table 5-6).
These three studies of related health effects: asthma, allergic sensitization, pulmonary
function, and symptoms of respiratory disease in children from in-home exposure to
formaldehyde (Rumchev et al., 2002; Garrett et al., 1999; Krzyzanowski et al., 1999) were
chosen as the basis for the derivation of the RfC. These co-critical studies are mutually
supportive and provide similar cRfCs. Therefore, the RfC is taken as the mean of the cRfCs of
the cRfCs of the three co-critical studies. For two of these studies (Rumchev et al., 2002; Garrett
et al., 1999), EPA is providing alternatives for the application of the UF addressing human
variability. These alternatives result in a threefold difference in cRfCs for each study when
considering the critical effects of childhood asthma and allergic sensitization (Table 5-6).
Alternative A, described above for each study, acknowledges that evaluation of these effects in
children does address some aspects of human variability, but there remains the potential for
additional inter-individual variability within the studied population, thus a UF of 3 is warranted.
Alternative B, described above for each study, also acknowledges that these studies address
human variability and susceptible populations. However in alternative B it is judged that since
children are a sensitive lifestage for these effects (asthma and atopy), and are likely the most
sensitive population, an UF of 1 may be applied. It is acknowledged that some degree of inter-
individual variability may remain.
This document is a draft for review purposes only and does not constitute Agency policy.
5-67 DRAFT—DO NOT CITE OR QUOTE
-------
3 5
o 5.
^ sZ
Co SS
II
Otq O
m a
S <^2"
,rj ^
^ S
o
•3
3?
w
§•
ci
si
s
s
>}
<£
Table 5-6: Summary of reference concentration (RfC) derivation from critical study and supporting studies.
§•
G\
00
o
H
Study
Application of
cRfC
(ppb)
Endpoint
Study
Homes
Children
POD (ppb)
study-specific UF
size
UFl
UFS
UFh
Respiratory effects / asthma and sensitization
Reduction of PEFR
in children (10%)
Krzyzanowski
etal. (1990)
208
Yes
Yes
BMCLjo = 17
1
1
3
5.6
Asthma incidence
Alternative A
Rumchev et al.
192
Yes
Yes
NOAEL = 33
1
3
3
3.3
(2002)
Alternative B
1
11
Increased asthma;
Alternative A
allergic sensitization
Garrett et al.
148
Yes
Yes
I.OAF.I. = 28
3
1
3
2.8
(1999)
Alternative B
1
9.3
Sensory Irritation
Eye irritation.
Alternative A
burning eyes
Ritchie and
2,007
Yes
Yes
NOAEL = 50
1
1
3
17
Lelinen (1987)
Alternative B
1
50
Alternative A
Hanrahan et al.
61
Yes
Some
BMCL10 = 70
1
1
3
23
(1984)
teenagers
Alternative B
1
70
Alternative A
Liu et al.
1,394
Yes
Yes
I.OAF.I. = 95
3
1
3
9.5
(1991)
Alternative B
1
32
Reproductive / Developmental
Decreased
Taskinen et al..
fecundability
1999
602
No
No
NOAEL= 86
1
1
10
8.6
density ratio (FDR)
Notes: 1: The final RfC will be rounded to one significant digit per EPA policy. Since the Candidate RfC is an interim calculation, two-significant digits are retained as common
practice in mathematics {i.e. one significant diget more that the final result, to avoid rounding errors compounding across multiple mathematical manipulations}.
o
o
-------
Alternative A: Application of a UF of 3 for human variability
Co-critical studies: Rumchev et al. (2002); Krzyzanowski et al. (1999); Garrett et al.
(1999)
Critical endpoints: Asthma, allergic sensitization, pulmonary function, and symptoms of
respiratory disease in children.
Candidate RfCs:
cRfC = 5.6ppb - decreased PEFR (Krzyzanowski et al., 1999)
cRfC = 3.3ppb - increased physician-diagnosed asthma (Rumchev et al., 2002)
cRfC = 2.8ppb - increased asthma, atopy and respiratory symptoms (Garrett et al., 1999)
5.6 ppb + 3.3 ppb + 2.8 ppb 11.7 ppb
RfC: RfC = — — = 4 ppb
Alternative B: Application of a UF of 1 for human variability
(UFh = 3 remains for Krzyzanowski et al., 1999)
Co-critical studies: Rumchev et al. (2002); Krzyzanowski et al. (1999); Garrett et al.
(1999)
Critical endpoints: Asthma, allergic sensitization, pulmonary function, and symptoms of
respiratory disease in children.
Candidate RfCs:
cRfC = 5.6ppb - decreased PEFR (Krzyzanowski et al., 1999)
cRfC = 11 ppb - increased physician diagnosed asthma (Rumchev et al., 2002)
cRfC = 9.3 ppb - increased asthma, atopy and respiratory symptoms (Garrett et al., 1999)
5.6 ppb + 11 ppb + 9.3 ppb 25.9 ppb n
RfC: RfC = — — = 9 ppb
5.1.3. Database Uncertainties in the RfC Derivation
The database of available laboratory animal studies, human clinical and epidemiological
studies, and supporting mechanistic information for formaldehyde is substantial. Many of the
health effects are well studied in animals and humans, especially those endpoints related to
sensory irritation and respiratory effects at the POE, such as respiratory tract pathology, asthma
This document is a draft for re\>iew purposes only and does not constitute Agency policy.
5-69 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
and reduced pulmonary function. This is reflected in the number and high quality of human
studies presented in Table 5-4 and supporting data summarized in Chapter 4.
The data also indicate effects in other health effect categories, specifically neurotoxic
effects, reproductive toxicity, and developmental effects (Section 5.1.2). These are areas where
additional research are needed to reduce uncertainty and better characterize the potential for
health effects and the concentrations at which they might occur in humans.
The existing database strongly supports formaldehyde's potential for causing both
reproductive and developmental toxicity. There is, however, no assessment of these endpoints
from a satisfactory two-generation toxicity study to fully evaluate the effect of formaldehyde
exposure on reproductive and developmental endpoints. Data are adequate to derive a cRfC of 9
ppb for decreased fecundability density ratio (FDR) from a human occupational study (Taskinen
et al., 1999). This study also reports an increase in spontaneous abortions, although there is
uncertainty on the exposure levels of concern for this endpoint; spontaneous abortions may also
contribute to the decreased FDR on which one of the cRfCs is based. The greatest uncertainty in
the cRfC for decreased FDR is the use of a time-weighted exposure metric which does not
address possible contributions of peak exposure levels to the observed health effect. As such, it
is possible that this cRfC is lower than is needed for protection against decreased FDR. The
cRfC for decreased FDR does suggest that the RfC derived from the better studied respiratory
effects would be protective of that reproductive/developmental endpoint, but there remain
uncertainties as to the full range of potential reproductive and developmental effects. No data
exist to sufficiently inform the exposure-response relationship for other reproductive and
developmental endpoints as they relate to RfC derivation (Section 5.1.2.6). For example, male
reproductive effects and structural and behavioral developmental effects (including postnatal
development) are not addressed by a study of decreased FDR. This is a database deficiency. A
survey of the currently available data indicates observed effect levels of 5,000-10,000 ppb for
male reproductive endpoints and 400 ppb and above for growth retardation and structural
anomalies in animal studies. However, these studies employed only one treatment level,
precluding the ability to establish a dose-response relationship, thus limiting the strength of the
studies for use in RfC derivation.
Similarly, there is evidence that formaldehyde can cause neurotoxic effects. There is a
deficit of studies with appropriate exposure scenarios to support derivation of an RfC reflecting
the potential for observed neurotoxicity due to formaldehyde exposure. None of the available
human studies that evaluated neurological effects were adequate for use in quantitative risk
assessment, although they did identify neurological effects of concern, including changes in
memory and concentration (e.g., Bach et al. [1990]; Kilburn et al. [1987, 1985]) and increased
This document is a draft for review purposes only and does not constitute Agency policy.
5-70 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
risk of mortality from amyotrophic lateral sclerosis (ALS) with increasing duration of exposure
to formaldehyde (Weisskopf et al., 2009). The human and animal data indicate the potential for
serious neurological and behavioral effects from short-term formaldehyde exposure (Section
5.1.2.6). Limited studies in humans, as well as controlled studies in established animal models,
confirm the neurotoxic effects of formaldehyde at exposure levels of 100-170 ppb (Malek et al.,
2003a, c; Bach et al., 1990) (Table 5-1). For example, an adverse effect level of 100 ppb for
impaired learning is reported for short-term exposures (2 hours/day for 10 days) in rats (Malek et
al., 2003a). For this effect, appropriate duration adjustment for extrapolation of a 2-hour
repeated exposure over a limited number of days is uncertain. Given the nature of these health
effects, and the potential for children to be exposed in the home to levels as high as 100 ppb (the
level at which effects were seen in animals following a single exposure), this is a significant data
gap. Studies are inadequate to determine whether exposure to levels of formaldehyde at or
below those that impact children's respiratory health and sensitization will cause neurotoxicity in
humans, including endpoints such as impaired learning and memory.
Approaches to the application of a database uncertainty factor:
Options EPA is considering include:
(1) Provide an RfC derived from studies of respiratory and allergenic responses and protective
of sensory irritation effects with a database uncertainty factor of one given significant data on
formaldehyde, but noting that further research reproductive, developmental and neurotoxic
effects would be valuable.
(2) Provide an RfC with a database uncertainty factor of one, with this RfC explicitly
identified as being protective of the well-studied effects.
(3) Apply a database UF of 3 to the RfC derived from studies of respiratory and allergenic
responses to reflect the potential that reproductive, developmental, or neurotoxic effects might
occur at lower doses:
(3) Provide both an RfC identified as protective of the better-studied effects and an RfC with
a database uncertainty factor of 3 incorporated to account for limits to the data on
reproductive, developmental and neurotoxic effects.
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
5-71 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
It is unclear what uncertainty factors are appropriate to account for human variability and
deficiencies in the overall database. For this reason, several alternatives have been presented.
5.1.4. Uncertainties in the RfC Derivation
By design, the RfC is an estimate of an exposure level at which it is unlikely there would
be deleterious effects to the human population (including sensitive subgroups) during a lifetime
of exposure. Although the RfC is derived from the best available studies, there are a number of
uncertainties that underlie the RfC. Some of these uncertainties are addressed quantitatively by
applying UFs on a study-specific a basis for RfCs based on animal studies, less-than-chronic
exposures, use of a LOAEL as the POD, and to address human variability for the relevant
endpoint (Section 5.1.3). This section elaborates on some of the sources of uncertainty in the
final RfC.
As the RfC is derived from human studies, the majority in a residential setting, study
aspects that are often a great source of uncertainty are of no concern (e.g., use of animal studies,
study of a worker population). The uncertainties discussed below apply specifically to the
database of formaldehyde studies and the process to derive the RfC.
Point of departure
Most of the studies considered for RfC derivation did not provide enough data to support
BMD modeling. Rather, the PODs for most studies were LOAELs or NOAELs that have a
number of shortcomings relative to a POD obtained from BMD modeling (i.e., a BMC or BMD):
• LOAELs and NOAELs are a reflection of the particular exposure/dose levels used in a
study, contributing some inaccuracy to the POD determination.
• LOAELs and NOAELs are often determined based on statistical significance and, thus,
reflect the number of study subjects or test animals. Studies are typically dissimilar in
detection ability and statistical power, with smaller studies tending to identify higher
exposure levels as NOAELs compared with larger but otherwise similarly designed
studies.
• Different LOAELs and NOAELs represent different response rates, so direct qualitative
and quantitative comparisons are not possible.
PODs identified from BMD models overcome some of the deficiencies associated with
LOAELs and NOAELs. Benchmark models were used for two inhalation data sets, Hanrahan et
al. (1984) and Krzyzanowski et al. (1990).
This document is a draft for review purposes only and does not constitute Agency policy.
5-72 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
It should also be noted, however, that even for BMCs/BMDs there is often uncertainty, in
particular for continuous responses, about what response level to select as the BMR, i.e., where
to define the cut-off point between a level of change that is not adverse and one that is adverse.
In addition, BMD models currently in use are purely mathematical models and are not intended
to accurately reflect the biology of the effect being modeled.
Another source of uncertainty in the POD is the adjustment for continuous exposure.
RfCs are meant to apply to continuous (24 hour/day) exposures. Exposure patterns in human
and laboratory animal inhalation studies are typically not continuous and assumptions must be
made in converting reported exposure levels to equivalent continuous exposures. Similarly,
there are uncertainties about potential dose rate effects, in particular the effect of peak exposures
in occupational studies.
Extrapolation from laboratory animal data to humans
Because the inhalation database for formaldehyde contains many human studies for a
variety of health effects, it was not necessary to rely on animal data for the endpoints from which
to derive the RfC. Thus, unlike for most RfCs, this is not a source of uncertainty in the RfC for
formaldehyde.
Human variation
Heterogeneity among humans is another uncertainty associated with extending results
observed in a limited human study population or laboratory animal experiment to a larger, more
diverse human population.
For three of the studies used to derive the RfC, a value of 3 was used for the human
variability UF (rather than the default value of 10) because the studies had an apparent over-
representation of populations expected to have increased susceptibility (Section 5.5.3.1):
¦ The residential study by Ritchie and Lehnen (1987) evaluated eye, nose, and throat
irritation in a large number of subjects, including children and the elderly. As a result of
the study's participation criteria, individuals with greater sensitivity were potentially
over-represented.
¦ Thirty percent of the subjects in the residential study by Krzyzanowski et al. (1990) were
children who are more sensitive to formaldehyde-associated decreases in PEFR than
adults. The cRfC determination for this study focused on the results in the children,
among whom asthmatics were over-represented (roughly three times) compared with the
national average of 9.4% in 2008 (Bloom et al., 2009).
This document is a draft for review purposes only and does not constitute Agency policy.
5.73 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
¦ Garrett et al. (1999) conducted a cross-sectional survey of allergy and asthma-like
symptoms in children with or without a doctor's diagnosis of asthma. The study was
designed to include a high proportion of asthmatic children, a sensitive population for the
effects being studied.
EPA notes, however, that, while a human variability UF of 3 rather than 10 was used to
account for certain special attributes of these studies/effects, there is still uncertainty about how
much of the overall population heterogeneity is actually reflected even in these relatively diverse
residential studies.
Subchronic-to-chronic extrapolation
RfCs are intended to apply to chronic lifetime exposures. If a study is subchronic
(typically less than 10% of a lifetime), a UF for subchronic-to-chronic extrapolation is generally
applied to the cRfC for that study. For the key human residential and occupational studies used
to derive the RfC in this assessment, the average durations of exposure in the households or
workplaces under study are unknown. In this assessment, these studies were considered chronic
in nature and no subchronic-to-chronic UF was applied. However, there is uncertainty about
whether or not the responses observed fully reflected the potential effects of chronic exposure,
especially in children, where, for example, impacts on the developing respiratory and immune
systems could be predisposing the children to further adverse effects later in life.
5.1.5. Previous Inhalation Assessment
There is no previous EPA RfC assessment for formaldehyde with which to compare and
contrast the RfC developed in this assessment.
5.2. QUANTITATIVE CANCER ASSESSMENT BASED ON THE NATIONAL
CANCER INSTITUTE COHORT STUDY
For quantitative assessment of cancer risk, it is generally preferable to use good-quality
epidemiologic data, when available, over laboratory animal data. The follow-up studies by
Hauptmann et al. (2004) and Beane Freeman et al. (2009) of the large National Cancer Institute
(NCI) retrospective cohort mortality study of U.S. workers involved in the production or use of
formaldehyde, with quantitative exposure estimates for the individual workers, present an
opportunity to perform quantitative cancer risk assessments of nasopharyngeal cancer (NPC) and
lymphohematopoietic cancers (Hodgkin lymphoma and leukemia) based on human data.
Although other upper respiratory tract cancers were also identified as being causally associated
with formaldehyde exposure in the weight-of-evidence analysis in section 4.5, NPC was the only
This document is a draft for review purposes only and does not constitute Agency policy.
5.74 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
upper respiratory tract cancer with exposure-response data adequate for the derivation of unit
risk estimates in the Hauptmann et al. (2004) follow-up study of solid tumors. Similarly, the
weight-of evidence analysis in section 4.5 concluded that there were causal relationships
between formaldehyde exposure and all lymphohematopoietic cancers as a group as well as
leukemias as a group (with the strongest evidence for myeloid leukemia); however, from the
Beane Freeman et al. (2009) follow-up study of lymphohematopoietic malignancies, only all
leukemias combined and Hodgkin lymphoma were judged to have exposure-response data
adequate for the derivation of unit risk estimates (see section 5.2.3.1 below).
5.2.1. Choice of Epidemiology Study
Several follow-up studies of formaldehyde exposure in industrial workers have recently
become available. These studies are discussed in more detail in chapter 4 and the appendix
(Human Health) and are reviewed only briefly here. Hauptmann et al. (2004) and
Beane Freeman et al. (2009) presented follow-ups of the NCI study (originally described by
Blair et al. [1986]) of workers at 10 U.S. plants producing or using formaldehyde. Marsh et al.
(2007, 2002) focused on pharyngeal cancer and, in particular, NPC mortality in sequential
follow-up analyses of the Marsh et al. (1996) cohort study, which examined 1 of the 10 plants
studied by NCI. Pinkerton et al. (2004) presented a follow-up of the National Institute for
Occupational Safety and Health (NIOSH) study of workers exposed to formaldehyde in three
U.S. garment plants (originally described by Stayner et al. [1988]). Coggon et al. (2003)
presented an extended follow-up of a study of workers in six British factories where
formaldehyde was produced or used (originally described by Acheson et al. [1984] and
previously followed up by Gardner et al. [1993]).
The analyses presented here are based on the NPC (Hauptmann et al., 2004) and
lymphohematopoietic cancer (Beane Freeman et al., 2009) results from the NCI follow-up
studies. The NCI cohort study is the largest of the three independent studies and is the only one
with sufficient individual exposure data for exposure-response modeling. In addition, the NCI
study is the only one of the three studies that used internal comparisons rather than standardized
mortality ratios (SMRs), thus minimizing the impact of the healthy worker effect, which can
attenuate observed effect estimates. The NCI cohort consists of 25,619 workers (88% male)
employed in any of the 10 plants prior to 1966. A follow-up through 1994 presented exposure-
response analyses for nine NPC deaths as well as analyses of deaths from other solid cancers
(Hauptmann et al., 2004). The most recent follow-up (through 2004) analyzed 319 deaths
attributed to lymphohematopoietic malignancy from a total of 13,951 deaths (Beane Freeman et
al., 2009). The results for solid cancers from this recent follow-up had not yet been published at
This document is a draft for review purposes only and does not constitute Agency policy.
5.75 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
the time of this draft assessment. A detailed exposure assessment was conducted for each
worker in the NCI cohort, based on exposure estimates for different jobs held and tasks
performed (Stewart et al., 1986). Exposure estimates were made using several different
metrics—peak exposure, average intensity, cumulative exposure, and duration of exposure.
Respirator use and exposures to formaldehyde-containing particulates and other chemicals were
also considered. For the NPCs, significant trends were observed for the cumulative and peak
exposure metrics (Hauptmann et al., 2004). For the lymphohematopoietic cancers, significant
trends were observed primarily for all lymphohematopoietic cancers and for Hodgkin lymphoma
with the peak exposure metric (Beane Freeman et al., 2009).
The NIOSH follow-up study (Pinkerton et al., 2004) analyzed mortality data (2,206
deaths; 59 from lymphatic and hematopoietic cancers) from their cohort of 11,098 workers (82%
female). Leukemia and aleukemia were elevated for workers with >10 years of exposure and for
workers with >20 years since first exposure. However, since no historical exposure level data
were available for this cohort, individual worker exposures could not be estimated and exposure-
response modeling was not conducted. The British cohort updated by Coggon et al. (2003)
consisted of 14,014 male workers, and the follow-up included 5,185 deaths (83 from
lymphohematopoietic cancers). In this cohort, lung cancer mortality was statistically
significantly increased, especially in workers in the high-exposure category; however, actual
exposure estimates were not available for exposure-response modeling (worker exposures were
categorized as nil/background, low, moderate, or high, depending on the job considered to have
had the highest exposure). Lymphohematopoietic cancers were not elevated in the British
cohort, although, as discussed above, the results were based on external comparisons against
national mortality statistics. Neither the NIOSH nor the British study reported increased risks of
NPC, although only 1 case (0.96) was expected in the NIOSH cohort (Pinkerton et al., 2003) and
only 2.0 cases were expected in the British cohort (Coggon et al., 2003).
5.2.2. Nasopharyngeal Cancer
5.2.2.1. Exposure-Response Modeling of the National Cancer Institute Cohort
A detailed exposure assessment was conducted for the NCI cohort, and quantitative
exposure estimates were generated for each worker (Stewart et al., 1986). Formaldehyde
exposure estimates, including 8-hour time-weighted average (TWA) exposures and level and
frequency of peak exposures, were derived for each job, work area, and calendar year
combination. A peak was defined as a short-duration exposure (typically <15 minutes) above
the TWA. Cumulative exposures (in ppm x years) were estimated by multiplying the time a
worker spent in a specific job by the TWA exposure for that job and summing over all the jobs
This document is a draft for review purposes only and does not constitute Agency policy.
5-76 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
held by the worker. Duration was the total time spent in jobs with formaldehyde exposure, and
average intensity was the ratio of cumulative exposure to duration. Formaldehyde exposures
after 1980 were not taken into account in the follow-up study, but this was considered to have a
minimal impact on the results (see section 5.2.2.4).
The results of NCI's internal analyses for NPC, using the peak exposure, average
intensity, cumulative exposure, and duration of exposure metrics, are presented in Table 5-7.
The relative risks (RRs) were estimated using log-linear Poisson regression models stratified by
calendar year, age, sex, and race and adjusted for pay category (salary/wage/unknown). The
NCI investigators used the low-exposure category as the reference category to "minimize the
impact of any unmeasured confounding variables since nonexposed workers may differ from
exposed workers with respect to socioeconomic characteristics" (Hauptmann et al., 2004). A 15-
year lag interval was used in estimating exposures in order to account from a minimal latency
period for the development of solid cancers, including NPCs.
As can be seen in Table 5-7, peak exposure is the exposure metric that provides the
strongest exposure-response relationship with NPC. However, it is not clear how to extrapolate
RR estimates based on these peak exposure estimates to meaningful estimates of lifetime extra
risk of cancer from environmental exposures, where the risk is usually considered to be from
continuous lifetime exposures to low environmental levels. In addition, peak exposure is a more
subjective measure than the other metrics, it is not based on actual measurements, and it is a
categorical rather than continuous measure. Furthermore, the "true" exposure metric best
describing the biologically relevant delivered dose of formaldehyde is unknown. The
cumulative exposure metric provides a good fit to the data (p trend = 0.029 for all person-years),
and, since this is generally the preferred metric for quantitative risk assessment for
environmental exposure to carcinogens, cumulative exposure was chosen as the exposure metric
for the risk estimate calculations for NPC in this assessment.
The nonexposed person-years were included in the primary cancer risk analyses
presented here in order to be more inclusive of all the exposure-response data. Such data are
typically included in exposure-response modeling. Furthermore, the data were stratified by pay
category, which should alleviate some concerns about the nonexposed workers having different
socioeconomic characteristics. Final results for the exposed person-years only are presented for
comparison.
This document is a draft for review purposes only and does not constitute Agency policy.
5.77 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-7. Relative risk estimates for mortality from nasopharyngeal
2 malignancies (ICD-8 code 147) by level of formaldehyde exposure for
3 different exposure metrics
4
Relative risk (number of deaths)
p trendb
p trend0
Peak exposure (ppm)
0
>0 to <2.0a
2.0 to <4.0
>4.0
1.00d (2)
-(0)
-(0)
1.83 (7)
0.044
<0.001
Average inl
ensity (ppm)
0
>0 to <0.5
0.5 to <1.0
>1.0
1.00d (2)
-(0)
0.38 (1)
1.67 (6)
0.126
0.066
Cumulative exposure (ppm x years)
0
>0 to <1.5
1.5 to <5.5
>5.5
2.40 (2)
1.00 (3)
1.19(1)
4.14 (3)
0.029
0.025
Duration of exposure (years)
0
>0 to <5
5 to <15
>15
1.77 (2)
1.00 (4)
0.83 (1)
4.18 (2)
0.206
0.147
5
6 ""Reference category for all categories.
7 bLikelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
8 peak exposure metric) among all (nonexposed and exposed) person-years.
9 likelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
10 peak exposure metric) among exposed person-years only.
11 dReference category due to no cases in the low-exposure category.
12
13 Source: Hauptmann et al. (2004).
14
15
16 As described above, Hauptmann et al. (2004) investigated the relationship between
17 formaldehyde exposure and NPC mortality using log-linear Poisson regression models. They
18 also conducted log-linear trend tests using the general model RR = epx, where (3 represents the
19 regression coefficient for exposure and X is exposure as a continuous variable. The trend
20 models were stratified by calendar year, age, sex, and race and adjusted for pay category.
21 Dr. Hauptmann provided EPA with the (3 estimates (and their standard errors) from the trend
22 tests for NPC and the cumulative exposure metric for all person-years and for exposed person-
23 years only (personal communication from Michael Hauptmann, NCI, to Jennifer Jinot, EPA,
24 March 29, 2004). These estimates are presented in Table 5-8.
25
26
27
28
This document is a draft for review purposes only and does not constitute Agency policy.
5-78 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-8. Regression coefficients from NCI log-linear trend test models for
2 NPC mortality from cumulative exposure to formaldehyde3
3
Person-years
P
(per ppm x year)
Standard error
(per ppm x year)
All
0.05183
0.01915
Exposed only
0.05318
0.01914
4
5 "Models stratified by calendar year, age, sex, and race and adjusted for pay category; cumulative exposures
6 calculated using a 15-year lag interval.
7
8 Source: Personal communication from Michael Hauptmann to Jenifer Jinot (March 29, 2004).
9
10
11 5.2.2.2. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer Mortality
12 The regression coefficients presented in Table 5-8 were used to predict the extra risk of
13 NPC mortality from environmental exposure to formaldehyde.
14
15 Extra ri sk = (Rx-Ro)/(1 -Ro),
16
17 where Rx is the lifetime risk in the exposed population and Ro is the lifetime risk in an
18 unexposed population (i.e., the background risk). Extra risk estimates were calculated using the
19 |3 regression coefficients and a life-table program that accounts for competing causes of death. 9
20 U.S. age-specific 1999 all-cause mortality rates for all race and gender groups combined
21 (National Center for Health Statistics [NCHS], 2002) were used to specify the all-cause
22 background mortality rates in the life-table program. NCHS 1996-2000 age-specific
23 background mortality rates for NPC were provided by Dr. Eisner of NCI's Surveillance,
24 Epidemiology and End Results (SEER) program (personal communication from Milton Eisner,
25 SEER, to Jennifer Jinot, EPA, December 19, 2003). Risks were computed up to age 85 because
26 cause-specific mortality (and incidence) rates for ages above 85 years are less reliable.
27 Conversions between occupational formaldehyde exposures and continuous environmental
28 exposures were made to account for differences in the number of days exposed per year (240
29 versus 365) and in the amount of air inhaled per day (10 versus 20 m3). An adjustment was also
30 made for the 15-year lag period. The reported standard errors for the regression coefficients
9
This program is an adaptation of the approach that was previously used in BEIRIV, "Health Risks of Radon and
Other Internally Deposited Alpha Emitters." National Academy Press, Washington, DC, 1988, pp. 131-134. The
same methodology was also used more recently in EPA's 1,3-butadiene health risk assessment (U.S. EPA, 2002). A
spreadsheet illustrating the life table used for the extra risk calculation for the derivation of the LECooos for NPC
incidence (see section 5.2.2.3) is presented in Appendix C.
This document is a draft for review purposes only and does not constitute Agency policy.
5.79 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
were used to compute the one-sided 95% upper confidence limits (UCLs) for the extra risks
based on a normal approximation.
Point estimates and one-sided 95% UCLs for the extra risk of NPC mortality associated
with varying levels of continuous exposure to formaldehyde are presented in Table 5-9. The
model predicts extra risk estimates that are fairly linear for exposures below about 0.001 to
0.01 ppm but not for exposures above 0.01 ppm.
Table 5-9. Extra risk estimates for NPC mortality from various levels of
continuous exposure to formaldehyde
Exposure concentration
(ppm)
Extra risk
95% UCL on extra
risk
0.0001
1.69 x 10~7
2.71 x 10~7
0.001
1.69 x 10"6
2.73 x 10~6
0.01
1.76 x 10~5
2.90 x 10~5
0.1
2.63 x 10"*
5.75 x 10~4
1
6.22 x 10"1
9.00 x 10"1
10
9.82 x 10"1
9.85 x 10"1
Consistent with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
the same data and methodology were also used to estimate the exposure level (effective
concentration [ECX]) and the associated (one-sided) 95% lower confidence limit (LECX)
corresponding to an extra risk of 0.05% (x = 0.0005). Although EPA guidelines emphasize the
use of exposure levels associated with a 10% extra risk level for the POD for low-dose
extrapolation, that would not be appropriate in this instance. A 10% extra risk level is very high
for responses generally observed in epidemiology studies; thus, a 1% extra risk level is typically
used for epidemiologic data to avoid upward extrapolation. For NPC, however, even the 1%
level of risk is associated with RR estimates that are substantially higher than those observed in
the epidemiology study. Hence, even a 1% extra risk level would be an upward extrapolation.
Based on the life-table program, the RR estimate for an extra risk of 1% for NPC mortality is 46.
Even O.P/o yields an RR estimate on the high end of the observable range of the epidemiology
study (RR = 5.5). A 0.05% extra risk level yields an RR estimate of 3.27, which better
corresponds to the RRs in the range of the data. Thus, 0.05% extra risk was selected for
determination of the POD, and, consistent with EPA's Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a), the LEC value corresponding to that risk level was used as the
POD. While this may appear to be an inordinately low response level, it must be recognized that
NPC has a very low background mortality rate (e.g., lifetime background risk is about 0.00022);
This document is a draft for review purposes only and does not constitute Agency policy.
5-80 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
therefore, a 1% extra risk (i.e., 0.01) would be a huge increase relative to the background risk.
This is consistent with the fact that, even with a large cohort followed for a long time, only nine
NPC deaths were observed in the NCI follow-up through 1994.10
Because formaldehyde is a mutagenic carcinogen and the weight of evidence suggests
that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic MOA
(section 4.5), a linear low-dose extrapolation was performed in accordance with EPA's
carcinogen risk assessment guidelines (U.S. EPA, 2005a). The ECooos, LECooos, and inhalation
unit risk estimates for NPC mortality are presented in Table 5-10.
Table 5-10. EC0oos? LECooos? and inhalation unit risk estimates for NPC
mortality from formaldehyde exposure based on the Hauptmann et al. (2004)
log-linear trend analyses for cumulative exposure
ECooos
LECooos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.15
0.093
5.4 x 10~3
Exposed only
0.15
0.091
5.5 x 10~3
aUnit risk = 0.0005/LEC0oos-
5.2.2.3. Prediction of Lifetime Extra Risk of Nasopharyngeal Cancer Incidence
EPA cancer risk estimates are typically derived to represent a plausible upper bound on
increased risk of cancer incidence, as from experimental animal incidence data. Cancer data
from epidemiology studies are more often mortality data, as is the case in the NCI study. For
cancers with low survival rates, mortality-based estimates are reasonable approximations of
cancer incidence risk. However, for NPC, the survival rate is substantial (51% at 5 years in the
1990s in the U.S., according to Lee and Ko [2005]), and incidence-based risks are preferred
because EPA is concerned with cancer occurrence, not just cancer mortality.
Therefore, an additional calculation was done using the same regression coefficients
provided by Dr. Hauptmann (Table 5-8) but with age-specific NPC incidence rates for 1996-
2000 from SEER in place of the NPC mortality rates in the life-table program. SEER collects
cancer incidence data from a variety of geographical areas in the U.S. The incidence data used
here are from SEER 12, a registry covering about 14% of the U.S. population, which was the
most current SEER registry at the time this analysis was done. SEER 1996-2000 age-specific
10 Ten NPCs were reported on death certificates and included in NCI's SMR analysis, but one of these cases was
apparently misclassified on the death certificate, so only nine cases were used to estimate the RRs in the internal
comparison analysis, as discussed by Hauptmann et al. (2004).
This document is a draft for review purposes only and does not constitute Agency policy.
5_81 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
background incidence rates for NPC were provided by Dr. Eisner of NCI's SEER program
(personal communication from Milton Eisner, SEER, to Jennifer Jinot, EPA, December 18,
2003). The incidence-based calculation relies on the reasonable assumptions that NPC incidence
and mortality have the same exposure-response relationship for formaldehyde exposure and that
the incidence data are for first occurrences of NPC or that relapses provide a negligible
contribution. The calculation also relies on the fact that NPC incidence rates are small compared
with the all-cause mortality rates.
The resulting ECooos, LECooos, and inhalation unit risk estimates for NPC incidence are
presented in Table 5-11. The unit risk estimate for cancer incidence is twofold higher than the
corresponding mortality-based estimate, for all person-years. This sizeable discrepancy can be
attributed to the high survival rates for NPC.
Table 5-11. ECooos, LECooos, and inhalation unit risk estimates for NPC
incidence from formaldehyde exposure based on the Hauptmann et al. (2004)
trend analyses for cumulative exposure
ECooos
LECooos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.074
0.046
1.1 x 10~2
Exposed only
0.072
0.045
1.1 x 10~2
"Unit risk = 0.0005/LEC0oos-
The preferred estimate for the inhalation cancer unit risk for NPC is the estimate of
1.1 x 10 2 per ppm derived using incidence rates for the cause-specific background rates, for all
person-years. The results from the exposed person-years are essentially identical.
Because NPC is a rare cancer, with a relatively low number of cases occurring per year in
the U.S., a rough calculation was done to assure that the unit risk estimate derived for NPC
incidence is not implausible in comparison to actual case numbers. For example, assuming an
average constant lifetime formaldehyde exposure level of 5 ppb for the U.S. population, the
inhalation unit risk estimate for NPC equates to a lifetime extra risk estimate of 5.5 x 10 5.
Assuming an average lifetime of 75 years (this is not EPA's default average lifetime of 70 years
but rather a value more representative of actual demographic data) and a U.S. population of
300,000,000, this lifetime extra risk estimate suggests a crude upper-bound estimate of 220
incident cases of NPC attributable to formaldehyde exposure per year. Alternatively, assuming
an average constant lifetime formaldehyde exposure level of 20 ppb, the calculation suggests a
crude upper-bound estimate of 880 incident cases of NPC per year. Both upper bound estimates,
This document is a draft for review purposes only and does not constitute Agency policy.
5-82 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
using different assumed lifetime exposure levels, are well below the estimated 2100 total
incident NPC cases per year calculated from a published NPC incidence rate for the U.S. of
0.7/100,000 person-years (Lee and Ko, 2005).11
5.2.2.4. Sources of Uncertainty
The two major sources of uncertainty in quantitative cancer risk estimates are generally
interspecies extrapolation and high-to-low dose extrapolation. The risk estimates derived from
the Hauptmann et al. (2004) analyses of the NCI cohort are not subject to interspecies
uncertainty since they are based on human data. However, substantial uncertainty remains in the
extrapolation from occupational exposures to lower environmental exposures. Although the
actual exposure-response relationship at low exposure levels is unknown, the linear low-dose
extrapolation that was used is warranted by the strong support for formaldehyde carcinogenicity
having a mutagenic MOA (section 4.5). The linear low-dose extrapolation from the 95% lower
bound on the exposure level associated with the extra risk level serving as the benchmark
response is generally considered to provide a plausible upper bound on the risk at lower
exposure levels. Actual low-dose risks may be lower to an unknown extent.
Other sources of uncertainty emanate from the epidemiologic study and its analysis
(Hauptmann et al., 2004), including the retrospective estimation of formaldehyde exposures in
the cohort, the modeling of the epidemiologic exposure-response data, the appropriate exposure
metric for exposure-response analysis, and potential confounding or modifying factors.
The same team of investigators (Stewart et al., 1986) conducted a detailed retrospective
exposure assessment to estimate the individual worker exposures. Formaldehyde exposures
were estimated for specific jobs/tasks based on monitoring data, discussions with workers and
plant managers, and assessment by industrial hygienists. Individual worker estimates were
derived for a variety of exposure metrics based on work histories. This exposure assessment was
a major undertaking, involving over 100 person-months. Hauptmann et al. (2004) suggested that
employment of such a detailed exposure assessment would tend to minimize exposure
misclassification for average and cumulative exposure and duration of exposure but that peak
exposure estimates could be more susceptible to misclassification because they were not based
on actual measurements. In addition, the follow-up study did not take into account exposures
after 1980. Hauptmann et al. (2003) stated that any underestimation of (total) exposure resulting
11 With the application of age-dependent adjustment factors (see Section 5.4.4), the lifetime unit risk estimate for
NPC would increase by a factor of 1.66, and the crude upper-bound estimates of the incident cases per year
attributable to formaldehyde exposure would similarly increase by a factor of 1.66. The resulting adjusted estimates
of 365 and 1460 for 5 ppb and 20 ppb exposure levels, respectively, are still well below the estimated total number
of incident cases per year in the U. S.
This document is a draft for review purposes only and does not constitute Agency policy.
5-83 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
from the 1980 cutoff "would be small because only 3.7% of all person-years were contributed by
workers who were 65 years or younger and in exposed jobs in 1980" and because exposure
levels were believed to have been much lower after 1980 than in earlier years.
As discussed in Chapter 4 and the appendix (Human Health), Marsh et al. (1996) also
estimated individual worker exposures at 1 of the 10 plants (Wallingford, Connecticut) studied
by the NCI team, and 5 of the 9 NPC deaths were from that plant. The Marsh et al. (1996)
exposure estimates were about 10-fold lower than those derived by the NCI team for the workers
at the Wallingford plant. Marsh et al. (2002) hypothesized that "the NCI used data from several
facilities to estimate exposures in a single facility." However, the NCI investigators maintained
that they estimated exposures for each plant separately. While the exact reasons for such a large
discrepancy are unclear, some differences in the assessment procedures which could have
resulted in substantial differences in the estimates are apparent. First, according to Marsh et al.
(1996), 91.7% of the white male Wallingford plant workers were specified as being exposed to
formaldehyde in the NCI study, while only 83.3% were considered to have been exposed in the
Marsh et al. (1996) analysis (it should be noted that these two cohorts of the Wallingford plant
are not identical). Second, the NCI investigators (Stewart et al., 1987, 1986) did their own
exposure monitoring at all the plants, including the Wallingford facility, in order to standardize
the data provided by the plants as well as to fill data gaps for certain jobs. There is no indication
that Marsh et al. (1996) made any additional measurements themselves. Third, although the
Marsh et al. (2002, 1996) papers are not entirely consistent on this point, those investigators
apparently assumed that the job-specific exposures at the plant were essentially constant over the
history of the plant, whereas the NCI team, based on interviews with plant personnel
knowledgeable about equipment and process changes, assumed that past exposures were higher.
In any event, despite the discrepancies in the absolute exposure values, the relative
exposures for both the Marsh et al. (2002, 1996) and NCI studies, as reflected in the exposure-
response relationships, are less subject to misclassification and are considered to be reliable.
The Wallingford plant is just 1 of the 10 plants in the NCI study (representing 4,389 of the
25,619 workers in the NCI cohort), but if the Marsh et al. (1996) exposure estimates, which are
roughly 10-fold lower than the NCI estimates, are closer to the actual exposures for those
workers, then the true potency of formaldehyde could be greater than that suggested by the unit
risk estimates calculated above based on the NCI data. Furthermore, if the NCI exposure values
were significantly overestimated across all 10 plants, then the actual potency could be higher
still.
With respect to the exposure-response model, the log-linear model used by Hauptmann et
al. (2003) for their trend tests (i.e., RR = epx) is a commonly used model for epidemiologic data
This document is a draft for review purposes only and does not constitute Agency policy.
5-84 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
with exposure as a continuous variable. However, the actual exposure-response relationship is
unknown. Moreover, even if the correct exposure-response model were known, there would be
substantial uncertainty in estimating the model parameters because there are only nine NPC
deaths to model. Furthermore, Beane Freeman et al. (2009) reported that in the follow-up
through 2004 it was discovered that 1,006 deaths that occurred during the 1980 to 1994
follow-up period had not been included in the analyses of the 1994 follow-up study (Hauptmann
et al., 2004, 2003), for reasons that have not been identified. Because NPC is such a rare cancer,
it is not expected that many, if any, NPC deaths were among the 1,006 excluded deaths;
however, it is unknown how inclusion of the 1,006 deaths would have altered the overall
exposure-response relationship and, hence, the regression coefficient. Additionally, a 15-year
lag was used for all the NCI solid cancer models. The actual minimum latency is unknown;
however, the investigators reported that lag intervals between 2 and 20 years yielded similar
results.
Another potentially significant source of uncertainty is associated with the exposure
metrics. With the log-linear model used for modeling the occupational data, the peak exposure
metric gave the strongest exposure-response relationship between formaldehyde exposure and
increased risk of NPCs. However, it is unclear how to extrapolate RR estimates based on peak
exposure estimates to meaningful estimates of lifetime extra risk of cancer from environmental
exposure (i.e., extra risk from lifetime continuous low-level environmental exposures). The
cumulative exposure metric also yielded a statistically significant exposure-response relationship
and was used for the primary cancer risk calculations in this assessment. The "true" exposure
metric best describing the toxicologically relevant dose of formaldehyde for nasopharyngeal
carcinogenesis is unknown. If a peak-exposure type of metric is the best representative of the
toxicologically relevant dose, this suggests that there are dose-rate effects in the exposure-
response relationship for formaldehyde and NPC. If this is the case, the unit risk estimates
presented here, which are based on a linear low-dose extrapolation, may overestimate the true
risks to an unknown extent.
Hauptmann et al. (2004) gave a lot of consideration to potential confounding and
modifying factors in their analyses. The important factors of age, race, sex, calendar year, and
pay category were taken into account in their Poisson regression and trend analyses.
Furthermore, they used the low-exposure person-years, rather than the unexposed person-years,
as their referent group in an effort to minimize any potential confounding effects resulting from
differences in socioeconomic or other characteristics between exposed and unexposed workers.
When the slope estimate (i.e., regression coefficient) for the exposed person-years only was used
This document is a draft for review purposes only and does not constitute Agency policy.
5-85 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
in the analyses presented here, the unit risk estimate was essentially identical to that calculated
from the slope estimate for all person-years (see Tables 5-12 and 5-13).
In addition, these investigators evaluated routine respirator use, exposure to
formaldehyde-containing particulates, durations of exposure to 11 other chemicals/substances in
the plants (antioxidants, asbestos, carbon black, dyes and pigments, hexamethylenetetramine,
melamine, phenol, plasticizers, urea, wood dust, and benzene), and duration of employment as a
chemist or laboratory technician. Only 133 workers ever routinely used a respirator (Hauptmann
et al., 2003). Hauptmann et al. (2004) reported that RR estimates for NPC changed when
adjusted for duration of melamine exposure, although trend tests remained significant for
cumulative formaldehyde exposure (p = 0.006). The investigators suggested that the association
with melamine may be spurious, and the regression coefficients (i.e., P estimates) used in this
assessment were not adjusted for melamine. RR estimates reportedly did not change
substantially when adjusted for exposure to any of the other 10 chemicals/substances. None of
the workers who died of NPC was identified as being exposed to wood dust. On the other hand,
each of the seven formaldehyde-exposed workers who died of NPC was also exposed to
particulates, and neither of the two workers who died of NPC but were not exposed to
formaldehyde was exposed to particulates. However, for those workers exposed to particulates,
NPC risk increased with increasing formaldehyde exposure, suggesting a formaldehyde-
associated effect. Nonetheless, because of the correspondence between formaldehyde and
particulate exposures within the workers who died of NPC, there is uncertainty as to whether or
not particulates were acting as a modifying factor. Adjusting for duration of time spent working
as a chemist or laboratory technician did not substantially alter the results (Hauptmann et al.,
2004).
Adjusting for plant may result in overadjustment because plant is highly correlated with
exposure. Moreover, Hauptmann et al. (2004) adjusted for important plant-related factors by
adjusting for the 11 chemicals/substances. Nonetheless, these investigators conducted analyses
adjusted for plant to address potential unmeasured confounders associated with plant, and they
reported that the association with NPC remained. As noted above, five of the nine NPC deaths
were from the Wallingford plant also studied by Marsh et al. (2006, 2002). Marsh et al. (2007)
hypothesized that the excess NPCs in the Wallingford plant could be due to external employment
in metal-working industries, but we found no evidence to support this supposition (see section
4.1.1.1).
Although smoking data were not available for the cohort, smoking is unlikely to explain
the excesses in NPCs because there was no consistent increase for tobacco-related diseases,
This document is a draft for review purposes only and does not constitute Agency policy.
5-86 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
including lung cancer, across the same exposure metrics. No information was available on
Epstein-Barr virus, a major risk factor for NPC, in the cohort.
Despite inevitable uncertainties, it is important not to lose sight of the strengths of the
NCI study. In addition to the use of internal analyses and the extensive exposure assessment and
consideration of potential confounding or modifying variables, the NCI study has a large cohort
that has been followed for a long time. The cohort included 25,619 subjects, 75% of whom
entered before 1960, contributing a total of 865,708 person-years (730,312 for the exposed
workers) to the 1994 follow-up. Duration of follow-up in 1994 ranged up to 58 years, with a
median of 35 years. Duration of exposure ranged up to 46 years, with a median of 2 years.
Additional uncertainties are not so much inherent in the exposure-response modeling or
in the epidemiologic data themselves but rather stem from the process of obtaining more general
EPA risk estimates from these specific results. EPA cancer risk estimates typically represent a
plausible upper bound on increased risk of cancer incidence in the general population for all
tissue sites potentially affected by an agent. For experimental animal studies, this is
accomplished by using tumor incidence data and summing across all the tumor sites that
demonstrate significantly increased incidences, generally using data from the most sensitive sex
and species. However, in estimating comparable risks from the NCI epidemiologic data, certain
limitations are encountered. First, the NCI study is a retrospective mortality study, and cancer
incidence data are unavailable for the cohort. Second, these occupational epidemiology data
represent a worker cohort that is generally healthier than the general population
(e.g., SMRs < 1) (see Table 2 of Hauptmann et al. [2004]).
The first limitation was addressed quantitatively in the calculation of cancer incidence
risk estimates from the mortality results, and, even though there are assumptions made in using
incidence data this way, the incidence-based estimates are believed to be better estimates of
cancer incidence risk than the mortality-based estimates. With respect to the second limitation,
the healthy worker effect is often an issue in occupational epidemiology studies, and it is
difficult to know to what extent there is a healthy worker effect with respect to the development
of NPC in this study. As discussed above, Hauptmann et al. (2004) sought to minimize potential
confounding effects resulting from differences in socioeconomic or other characteristics between
exposed and unexposed workers by using the low-exposure person-years, rather than the
unexposed person-years, as their referent group. Nonetheless, when the slope estimates for the
exposed person-years only were used in the analyses in this assessment, unit risk estimates
essentially identical to those calculated from the slope estimates for all person-years were
obtained (Tables 5-12 and 5-13). In terms of representing the general population, the NCI cohort
This document is a draft for review purposes only and does not constitute Agency policy.
5-87 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
was somewhat diverse, but the workers were predominantly white males (81%) then white
females (12%), black males (7%), and black females (<1%), and they were all adults.
Finally, NPC is just one of the upper respiratory tract cancers concluded to be causally
associated with formaldehyde exposure (section 4.5). These upper respiratory tract cancers are
rare cancers and are difficult to detect in cohort studies. Thus, although NPC was the only such
cancer with an exposure-response relationship amenable to the derivation of a unit risk estimate,
additional, unquantified risk may exist for the other upper respiratory tract cancers. If there was
a strong exposure-response relationship between these cancers and formaldehyde exposure, a
more apparent association in the Hauptmann et al. (2004) study might have been expected, as
was seen for NPC, despite the rare nature of these cancers. Thus, the exposure-response
relationship for these other upper respiratory tract cancers is likely modest, at best, and, because
these are rare cancers, the contribution of the risk for these cancers to the total cancer risk from
formaldehyde exposure is not expected to be large. Nonetheless, with such rare cancers, there is
uncertainty regarding the extent to which the estimate based on NPC may underestimate the risk
for all upper respiratory tract cancers.
In summary, the inhalation cancer unit risk estimate of 1.1 x 10 2 per ppm for NPC is
based on human data from a high-quality epidemiologic study with individual exposure
estimates for each worker. A major uncertainty is the appropriate model/exposure metric for
extrapolation to environmental exposures.
5.2.3. Lymphohematopoietic Cancer
5.2.3.1. Exposure-Response Modeling of the National Cancer Institute Cohort
The results of NCI's internal analyses for lymphohematopoietic cancers using the peak
exposure, average intensity, and cumulative exposure metrics from the follow-up through 2004
are reported by Beane Freeman et al. (2009). There was reportedly no evidence of associations
with duration of exposure, and those results were not presented. For the peak exposure metric,
statistically significant log-linear trends were observed for all lymphohematopoietic cancers,
Hodgkin lymphoma, and leukemia (the latter only when the unexposed person-years were
included). There was also evidence for potential associations with myeloid leukemia
specifically, especially when risks were viewed over time, and with multiple myeloma. Using
the average exposure metric, there was a significant trend for Hodgkin lymphoma. With the
cumulative exposure metric, there were no statistically significant trends; however, the Hodgkin
lymphoma trend results were of borderline significance (p trends = 0.06 and 0.08 with and
without the unexposed person-years, respectively), as were the leukemia trend results (p trends =
0.08 and 0.12 with and without the unexposed person-years, respectively). As discussed above
This document is a draft for review purposes only and does not constitute Agency policy.
5-88 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
with NPC, it is not clear how to extrapolate RR estimates based on the peak exposure estimates
to meaningful estimates of lifetime extra risk of cancer from environmental exposures. The
average exposure metric is also problematic because it suggests that duration of exposure is not
important (e.g., exposure to a given exposure level for 1 year conveys the same amount of risk
as exposure to the same level for 70 years). Cumulative exposure is generally the preferred
metric for quantitative risk assessment for environmental exposure to carcinogens, and, because
the Hodgkin lymphoma and leukemia trend results were of borderline statistical significance
using the cumulative exposure metric and the elevations in risk with that metric were consistent
with significant elevations observed with the peak exposure (for Hodgkin lymphoma and
leukemia) and average exposure (for Hodgkin lymphoma) metrics (Table 5-12), a determination
was made to calculate unit risk estimates for Hodgkin lymphoma and leukemia based on
cumulative exposure. There is also support for associations between formaldehyde exposure and
both Hodgkin lymphoma and leukemia from other studies (section 4.5.2). No other
lymphohematopoietic cancer responses provided adequate exposure-response data with the
cumulative formaldehyde exposure metric in the NCI cohort from which to derive unit risk
estimates.
As for the NPC results discussed in section 5.2.2, the RR estimates in Table 5-12 were
derived using log-linear Poisson regression models stratified by calendar year, age, sex, and race
and adjusted for pay category (salary/wage/unknown). The NCI investigators used the low-
exposure category as the reference category to "minimize the impact of any unmeasured
confounding variables since nonexposed workers may differ from exposed workers with respect
to socioeconomic characteristics" (Hauptmann et al., 2004). A 2-year lag interval was used to
determine exposures in order to account for a minimal latency period for lymphohematopoietic
cancers.
Dr. Beane Freeman provided EPA with the regression coefficient estimates for Hodgkin
lymphoma and leukemia mortality from the log-linear trend test models for cumulative exposure
(i.e., RR = epx, with exposure [X] as a continuous variable) used in the NCI analyses (personal
communication from Laura Beane Freeman, NCI, to John Whalan, EPA, August 26, 2009).
These estimates are presented in Table 5-13. As with the NPC calculations in section 5.2.2, the
nonexposed person-years were included in the primary unit risk estimate derivations in order to
be more inclusive of all the exposure-response data. Final results for the exposed person-years
only are presented for comparison.
This document is a draft for review purposes only and does not constitute Agency policy.
5-89 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-12. Relative risk estimates for mortality from Hodgkin lymphoma
2 (ICD-8 code 201) and leukemia (ICD-8 codes 204-207) by level of
3 formaldehyde exposure for different exposure metrics
4
Cancer type
Relative risk (number of deaths)
p trendb
p trend0
Peak exposure (ppm)
0
>0 to <2.0a
2.0 to <4.0
>4.0
Hodgkin
lymphoma
0.67 (2)
1.0(6)
3.30 (8)
3.96(11)
0.004
0.01
Leukemia
0.59 (7)
1.0 (41)
0.98 (27)
1.42 (48)
0.02
0.12
Average intensity (ppm)
0
>0 to <0.5
0.5 to <1.0
>1.0
Hodgkin
lymphoma
0.53 (2)
1.0 (10)
3.62 (9)
2.48 (6)
0.03
0.05
Leukemia
0.54 (7)
1.0 (67)
1.13 (25)
1.10(24)
0.50
>0.5
0
Cumulative exposure (ppm x years)
0
>0 to <1.5
1.5 to <5.5
>5.5
Hodgkin
lymphoma
0.42 (2)
1.0 (14)
1.71 (7)
1.30 (4)
0.06
0.08
Leukemia
0.53 (7)
1.0 (63)
0.96 (24)
1.11 (29)
0.08
0.12
5
6 "Reference category for all categories.
7 bLikelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
8 peak exposure metric) among all (nonexposed and exposed) person-years.
9 likelihood ratio test (1 degree of freedom) of zero slope for formaldehyde exposure (continuous variable, except for
10 peak exposure metric) among exposed person-years only.
11
12 Source: Beane Freeman et al. (2009).
13
14
15 Table 5-13. Regression coefficients for Hodgkin lymphoma and leukemia
16 mortality from NCI trend test models3
17
Cancer type
Person-years
P
(per ppm x year)
Standard error
(per ppm x year)
Hodgkin lymphoma
All
0.02959
0.01307
Exposed only
0.02879
0.01333
Leukemia
All
0.01246
0.006421
Exposed only
0.01131
0.00661
18
19 aModels were stratified by calendar year, age, sex, and race and adjusted for pay category; exposures included a
20 2-year lag interval.
21
22 Source: Personal communication from Laura Beane Freeman to John Whalan (August 26, 2009).
23
This document is a draft for review purposes only and does not constitute Agency policy.
5-90 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
5.2.3.2. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and Leukemia Mortality
Extra risk estimates for Hodgkin lymphoma and leukemia mortality were calculated
using the same general methodology described above for the NPC mortality estimates (section
5.2.2.2), with the following exceptions. U.S. age-specific 2006 all-cause mortality rates (NCHS,
2009) and NCHS age-specific 2002-2006 background mortality rates for Hodgkin lymphoma
and leukemia (http://seer.cancer.gov/csr/1975_2006/) for all race and gender groups combined
were used in the life-table programs. In addition, a 2-year lag period was used instead of a
15-year lag period.
The resulting point estimates and one-sided 95% UCLs for the extra risk of Hodgkin
lymphoma mortality associated with varying levels of continuous exposure to formaldehyde are
presented in Table 5-14. The results for leukemia are shown in Table 5-15. In both cases, the
models predict extra risk estimates that are fairly linear for exposures below about 0.01-0.1 ppm
but not for exposures above 0.1 ppm.
Table 5-14. Extra risk estimates for Hodgkin lymphoma mortality from
various levels of continuous exposure to formaldehyde
Exposure concentration
(ppm)
Extra risk
95% UCL on extra risk
0.0001
2.04 x 10~7
3.53 x 10~7
0.001
2.05 x 10~6
3.55 x 10~6
0.01
2.10 x 10~5
3.71 x 10~5
0.1
2.79 x 10~4
6.17 x 10~4
1
1.63 x 10"1
8.36 x 10"1
10
9.89 x 10"1
9.90 x 10"1
Table 5-15. Extra risk estimates for leukemia mortality from various levels
of continuous exposure to formaldehyde
Exposure concentration
(ppm)
Extra risk
95% UCL on extra risk
0.0001
1.64 x 10~6
3.02 x 10~6
0.001
1.64 x 10~5
3.03 x 10~5
0.01
1.66 x 10~4
3.10 x 10~4
0.1
1.87 x 10~3
3.90 x 10~3
1
8.07 x 10~2
5.19 x 10"1
10
9.80 x 10"1
9.89 x 10"1
This document is a draft for review purposes only and does not constitute Agency policy.
5-91 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
As discussed in section 5.2.2.2 above, 1% extra risk levels are typically used as the basis
for the POD for low-dose extrapolation from epidemiologic data. As for NPC, however,
Hodgkin lymphoma has a very low background mortality rate (e.g., lifetime background risk is
about 0.00038), and the 1% level of risk is associated with RR estimates that are substantially
higher than those observed in the epidemiology study. Hence, a 1% extra risk level would be an
upward extrapolation. Based on the life-table program, the RR estimate associated with an extra
risk of 1% for Hodgkin lymphoma mortality is 27. Even 0.1% yields an RR estimate at the
higher end of what was observed in the epidemiology study (RR = 3.6) (note that our primary
analyses include the nonexposed workers, and thus the 0-exposure group becomes the referent
group and the RR estimates presented for Hodgkin lymphoma and cumulative exposure in Table
5-12 would be adjusted upward [about 2.4-fold] relative to the 0-exposure group). A 0.05%
extra risk level yields an RR estimate of 2.3, which better corresponds to the RRs at the lower
end of the observable range. Thus, 0.05% extra risk was selected for determination of the POD
for Hodgkin lymphoma, and, consistent with EPA's Guidelines for Carcinogen Risk Assessment
(U.S. EPA, 2005a), the LEC value corresponding to that risk level was used as the POD.
For leukemia, although the background mortality rates are higher (0.0065), the 1% extra
risk level typically used as the basis for the POD for epidemiologic data still corresponds to an
RR estimate (2.5) that would be above the highest categorical result reported, even after
adjusting the RR estimates upward relative to the 0-exposure group (see above paragraph). A
0.5% extra risk level yields an RR estimate of 1.8, which better corresponds to the RRs in the
range of the data. Thus, the LEC value corresponding to 0.5% extra risk was selected for the
POD for leukemia.
Because formaldehyde is a mutagenic carcinogen and the weight of evidence suggests
that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic MOA
(section 4.5), a linear low-dose extrapolation was performed, also in accordance with EPA's
Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a). The ECooos, LECooos, and
inhalation unit risk estimates for Hodgkin lymphoma mortality are presented in Table 5-16, and
the ECoos, LECoos, and inhalation unit risk estimates for leukemia mortality are presented in
Table 5-17.
This document is a draft for review purposes only and does not constitute Agency policy.
5-92 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-16. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin
2 lymphoma mortality from formaldehyde exposure based on Beane Freeman
3 et al. (2009) log-linear trend analyses for cumulative exposure
4
ECooos
LECooos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.151
0.0875
5.7 x 10~3
Exposed only
0.155
0.0881
5.7 x 10~3
5
6 aUnit risk = 0.0005/LEC0oos-
7
8
9 Table 5-17. ECoos, LECoos? and inhalation unit risk estimates for leukemia
10 mortality from formaldehyde exposure based on Beane Freeman et al. (2009)
11 log-linear trend analyses for cumulative exposure
12
ECoos
LECoos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.224
0.121
4.1 x 10~2
Exposed only
0.246
0.126
4.0 x 10~2
13
14 aUnit risk = 0.005/LEC005.
15
16
17 5.2.3.3. Prediction of Lifetime Extra Risks for Hodgkin Lymphoma and Leukemia Incidence
18 As for NPC, both Hodgkin lymphoma and leukemia have substantial survival rates
19 (84.7% at 5 years for Hodgkin lymphoma [http://seer.cancer.gov/statfacts/html/hodg.html] and
20 53.1% at 5 years for leukemia [http://seer.cancer.gov/statfacts/html/leuks.html], based on 1999-
21 2005 SEER data); thus, it is preferable to derive incidence estimates. Unit risk estimates for
22 Hodgkin lymphoma and for leukemia incidence were calculated as described above for the NPC
23 incidence estimates (section 5.2.2.3). Age-specific background incidence rates for 2002-2006
24 for Hodgkin lymphoma and for leukemia from SEER17, a registry covering about 26% of the
25 U.S. population, were obtained from the SEER Web site (http://seer.cancer.gov/csr/1975_2006/).
26 The incidence-based calculation relies on the assumptions that Hodgkin lymphoma (and
27 leukemia) incidence and mortality have the same exposure-response relationship for
28 formaldehyde exposure and that the incidence data are for first occurrences of Hodgkin
29 lymphoma (and leukemia) or that relapses provide a negligible contribution. The first
30 assumption is more uncertain for leukemia because it is a grouping of subtypes with different
31 survival rates (see section 5.2.3.4 for further discussion). The calculation also relies on the fact
32 that Hodgkin lymphoma (and leukemia) incidence rates are small compared with the all-cause
This document is a draft for review purposes only and does not constitute Agency policy.
5.93 DRAFT—DO NOT CITE OR QUOTE
-------
1 mortality rates. The resulting ECooos, LECooos, and inhalation unit risk estimates for Hodgkin
2 lymphoma incidence are presented in Table 5-18, and the ECoos, LECoos, and inhalation unit risk
3 estimates for leukemia incidence are presented in Table 5-19. The unit risk estimate for Hodgkin
4 lymphoma incidence is about threefold higher than the corresponding mortality-based estimate,
5 for all person-years. This sizeable discrepancy can be attributed to the high survival rates for
6 Hodgkin lymphoma. For leukemia, the incidence unit risk estimate is about 40% higher than the
7 mortality-based estimate. This difference is lower than the twofold difference seen with NPC
8 estimates, despite comparable survival rates, probably because of different age distributions of
9 the mortality and incidence rates.
10
11 Table 5-18. ECooos, LECooos, and inhalation unit risk estimates for Hodgkin
12 lymphoma incidence from formaldehyde exposure, based on Beane Freeman
13 et al. (2009) log-linear trend analyses for cumulative exposure
14
ECooos
LECooos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.0515
0.0298
1.7 x 10~2
Exposed only
0.0529
0.0301
1.7 x 10~2
15
16 aUnit risk = 0.0005/LEC00o5
17
18
19 Table 5-19. ECoos, LECoos, and inhalation unit risk estimates for leukemia
20 incidence from formaldehyde exposure based on Beane Freeman et al. (2009)
21 log-linear trend analyses for cumulative exposure
22
ECoos
LECoos
Unit risk3
Person-years
(ppm)
(ppm)
(ppm1)
All
0.162
0.0875
5.7 x 10~2
Exposed only
0.178
0.0909
5.5 x 10~2
23
24 aUnit risk = 0.005/LEC0o5.
25
26
27 The preferred estimate for the inhalation cancer unit risk for Hodgkin lymphoma is the
28 estimate of 1.7 x 10 2 per ppm derived using incidence rates for the cause-specific background
29 rates, for all person-years. Similarly, the preferred estimate for leukemia is the estimate of
30 5.7 x 10~2 per ppm derived using incidence rates, for all person-years. In both cases, the results
31 from the exposed person-years only are essentially identical.
This document is a draft for review purposes only and does not constitute Agency policy.
5.94 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Because Hodgkin lymphoma is a rare cancer, with a relatively low number of cases
occurring per year in the U.S. (according to SEER statistics, an estimated 8,510 people were
diagnosed with Hodgkin lymphoma in the U.S. in 2009
[http://seer.cancer.gov/statfacts/html/hodg.html]), a rough calculation was done to assure that the
unit risk estimate derived for Hodgkin lymphoma incidence is not implausible in comparison to
actual case numbers. For example, assuming an average constant lifetime formaldehyde
exposure level of 5 ppb for the U.S. population, the inhalation unit risk estimate for Hodgkin
lymphoma equates to a lifetime extra risk estimate of 8.5 x 10 5, Assuming an average lifetime
of 75 years (this is not EPA's default average lifetime of 70 years but rather a value more
representative of actual demographic data) and a U.S. population of 300,000,000, this lifetime
extra risk estimate suggests a crude upper-bound estimate of 340 incident cases of Hodgkin
lymphoma attributable to formaldehyde exposure per year. Alternatively, assuming an average
constant lifetime formaldehyde exposure level of 20 ppb, the calculation suggests a crude upper-
bound estimate of 1,360 incident cases of Hodgkin lymphoma per year. Both upper bound
estimates, using different assumed lifetime exposure levels, are well below the estimated 8,510
total incident Hodgkin lymphoma cases diagnosed per year in the U.S.12
5.2.3.4. Sources of Uncertainty
By and large, the sources of uncertainty discussed above (section 5.2.2.4) for the NPC
risk estimates, such as high-to-low dose extrapolation, retrospective exposure estimation,
exposure metric/model uncertainties, and application of data from a "healthy" worker cohort to
the more diverse general population also apply to the Hodgkin lymphoma and leukemia risk
estimates. The Hodgkin lymphoma risk estimates are based on 27 deaths, which is more than
were available for the NPC risk estimates, but 27 is still a small number for exposure-response
modeling. The leukemia risk estimates are based on 123 deaths, so there is less uncertainty with
the parameter estimation from the exposure-response modeling for that cancer type, although
uncertainties still exist about the general model form. A 2-year lag interval was used for
12 With the application of age-dependent adjustment factors (see Section 5.4.4), the lifetime unit risk estimate for
Hodgkin lymphoma would increase by a factor of 1.66, and the crude upper-bound estimates of the incident cases
per year attributable to formaldehyde exposure would similarly increase by a factor of 1.66. The resulting adjusted
estimates of 564 and 2260 for 5 ppb and 20 ppb exposure levels, respectively, are still well below the estimated total
number of incident cases per year in the U.S.. Similar calculations for leukemia yield even lower relative upper-
bound estimates of cases attributable to formaldehyde exposure, in comparison to estimated total incidenct cases,
because, although the unit risk estimate for leukemia is about 3.3 times the unit risk estimate for Hodgkin lymphoma,
the total estimated number of incident leukemia cases in the U.S. is 5.3 times the estimate for Hodgkin lymphoma
(an estimated 44,790 cases diagnosed in the U.S. for 2009, according to SEER
[http://seer.cancer.gov/statfacts/html/leuks.html]).
This document is a draft for review purposes only and does not constitute Agency policy.
5.95 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
lymphohematopoietic cancers versus the 15-year lag for NPC. Beane Freeman et al. (2009)
evaluated lag intervals between 2 and 25 years and reported that lag intervals of about 18 years
provided the best fit to the lymphohematopoietic cancer data but did not change the risk
estimates; thus, they retained the 2-year lag interval that was used in the previous follow-up
(Hauptmann et al., 2003). The most appropriate lag intervals for Hodgkin lymphoma and
leukemia are unknown, but alternate lags are unlikely to have a large impact on the results.
The same potential confounding or modifying factors that were investigated for NPC and
the other solid cancers, as discussed in section 5.2.2.4 above, were evaluated for the
lymphohematopoietic cancers. Beane Freeman et al. (2009) reported that controlling for
duration of exposure to the 11 other substances that they considered (see section 5.2.2.4) or for
working as a chemist or laboratory technician "did not meaningfully change results"; results
were not shown. The investigators also reported that excluding the 586 individuals with exposure
to benzene, a known leukemogen, did not change the RR estimates for myeloid or lymphoid
leukemia in the highest peak exposure category. Furthermore, Beane Freeman et al. (2009)
found no evidence of heterogeneity of RR estimates for lymphohematopoietic cancers by race,
sex, or pay category, and adjusting for plant reportedly did not substantively change results.
A further uncertainty is which lymphohematopoietic cancer types are linked to
formaldehyde exposure. As discussed in section 4.5.2, lymphohematopoietic cancers are a
diverse group of cancers with different etiologies, and the epidemiologic database suggests
associations with multiple different subtypes of these cancers. Section 4.5 concludes that
formaldehyde is causally associated with all lymphohematopoietic cancers as a group and with
leukemias as a group (with the strongest evidence for myeloid leukemia). However, at present,
exactly which subtypes are etiologically linked to formaldehyde exposure is unknown. Cancer
risk estimates were derived for Hodgkin lymphoma and leukemia because, in addition to support
for an association between these lymphohematopoietic cancer subtypes and formaldehyde
exposure with other exposure metrics and from other studies, these had the strongest associations
with cumulative exposure in the Beane Freeman et al. (2009) update of the large, high-quality
NCI study. However, it is unknown whether these two subtypes best represent the total
lymphohematopoietic cancer risk.
In addition, leukemia itself is a grouping of diverse (e.g., acute lymphocytic, chronic
lymphocytic, acute myeloid, chronic myeloid) subtypes, and using this grouping injects
additional uncertainty into the derivation of cancer incidence estimates. One of the assumptions
that the incidence-based calculation relies on is that the cancer incidence and mortality have the
same exposure-response relationship for formaldehyde exposure. This assumption may be
problematic for the leukemia incidence estimates if not all of the leukemia subtypes represented
This document is a draft for review purposes only and does not constitute Agency policy.
5-96 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
in the grouping are associated with formaldehyde exposure to the same extent. This is because
different leukemia subtypes have different survival rates, so if a subtype with a relatively high
survival rate is included in the background incidence rates while not actually being associated
with formaldehyde exposure or being associated to a lesser extent than other subtypes, then the
incidence risk will be overestimated. The mortality risk calculations are not similarly affected
by including subtypes that may not actually be associated with formaldehyde exposure because
background mortality for the subtypes is already taken into account in the regression coefficient.
Figure 5-10 shows the mortality versus incidence rates for all leukemia and the two main
subtypes, myeloid leukemia and lymphoid leukemia. This figure does not show the acute versus
chronic myeloid and leukemia subtypes or the monocytic or other leukemia subtypes; however,
it serves to illustrate the impact of using rates for groupings that contain subtypes with different
survival rates. For example, if lymphoid leukemia is the predominant subtype associated with
formaldehyde exposure, then using the leukemia grouping for the incidence rates may
underestimate the cancer incidence risk because the incidence rates for leukemia (relative to the
mortality rates) are diluted with inclusion of the incidence rates for myeloid leukemia, which has
a smaller incidence-to-mortality ratio (i.e., poorer survival). On the other hand, if myeloid
leukemia is the predominant subtype associated with formaldehyde exposure, then using the
leukemia grouping for the incidence rates may overestimate cancer incidence risk. If incidence
risks are being overestimated, the effect should be minimal because the incidence risk estimates
for leukemia calculated in section 5.2.3.3 are not that much greater (about 40%) than the
mortality-only estimates.
Finally, as for the NPC risk estimates, when the slope estimates for the exposed person-
years only were used for the Hodgkin lymphoma and leukemia risk calculations, unit risk
estimates similar to those calculated from the slope estimates for all person-years were obtained
(Tables 5-18 and 5-19); thus, the impacts of including the unexposed person-years are minimal.
As discussed in section 5.2.2.4, despite inevitable uncertainties, it is important not to lose
sight of the strengths of the NCI study. In addition to the use of internal analyses and extensive
exposure assessment and consideration of potential confounding or modifying variables, the NCI
study has a large cohort that has been followed for a long time. With the additional follow-up
through 2004, reflected in the lymphohematopoietic cancer results of Beane Freeman et al.
(2009), the median duration of follow-up was 42 years, and the 25,619 cohort members had
accrued 998,106 person-years of follow-up. Over half of the cohort was deceased, and there was
a substantial number of lymphohematopoietic deaths (319 total; 286 in the exposed workers).
This document is a draft for review purposes only and does not constitute Agency policy.
5.97 DRAFT—DO NOT CITE OR QUOTE
-------
^ ^2
Co SS
II
Oq o
0
1
s
I
§•
^n
3?
W
§•
cj
s
s
>!
fc
$•
e5
VO
00
o
80.0
L 40.0
age (years)
— all leuk moil
all leuk inc
— myeloid mort
^—myeloid inc
¦ — lymphoid mort
— lymphoid inc
Figure 5-10. Age-specific mortality and incidence rates for myeloid, lymphoid, and all leukemia.
O
O
-------
1
2
3
4
5
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
In summary, the inhalation cancer incidence unit risk estimates of 1.7 x 10 2 per ppm for
Hodgkin lymphoma and 5.7 x 10 2 per ppm for leukemia are based on human data from a high-
quality epidemiologic study with individual exposure estimates for each worker. The major
source of uncertainty in both risk estimates is the extrapolation to environmental exposures.
5.2.4. Conclusions on Cancer Unit Risk Estimates Based on Human Data
In this assessment, a (plausible upper bound) lifetime extra cancer unit risk of 5.4 x 10 3
per ppm of continuous formaldehyde exposure was estimated using a life-table program and
linear low-dose extrapolation of the excess NPC mortality and log-linear modeling results (for
cumulative exposure) reported in a high-quality occupational epidemiologic study (based on nine
NPC deaths). Applying the same regression coefficient and life-table program to background
NPC incidence rates yielded a lifetime extra cancer unit risk estimate of 1.1 x 10~2 per ppm
(8.8 x 10~6 per |ig/m3).
Using similar methods and data for Hodgkin lymphoma (27 deaths) and leukemia
(123 deaths) mortality based on the cumulative exposure metric, from a further follow-up of the
same cohort study, (plausible upper bound) lifetime extra cancer risk estimates of 1.7 x 10 2 per
ppm (1.4 x 10~5 per |ig/m3) and 5.7 x 10~2 per ppm (4.6 x 10~5 per |ig/m3) for Hodgkin
lymphoma incidence and leukemia incidence, respectively, were derived.
To estimate the total cancer risk from formaldehyde exposure, risk estimates for these
three cancer types (NPC, Hodgkin lymphoma, and leukemia) were combined, although, as
discussed above, these three cancer types may not fully reflect the total cancer risk for all
cancers thought to be causally associated with formaldehyde exposure. For an approximate
estimate of the combined (upper bound) risk, risk estimates were combined assuming a normal
distribution. For comparability, risk estimates for formaldehyde were combined at a common
level of 0.1 ppm. This level was selected because it is close to the PODs (LECooss) used above
for leukemia mortality (0.121 ppm) and leukemia incidence (0.0875 ppm), and leukemia is the
predominant cancer type in terms of extra risk. Note that unit risk estimates for the different
cancer types calculated at 0.1 ppm will differ slightly from those reported above (sections 5.2.2
and 5.2.3) because they are calculated at a level other than the PODs used in the above
calculations. To derive the combined risk, maximum likelihood estimates (MLEs) of risk and
their 95% upper bounds (UCLs) were calculated for each cancer type using the same methods
and life-table programs employed in sections 5.2.2 and 5.2.3. The standard errors (SEs) were
then estimated from the risk estimates using the equation: UCL = MLE + 1.645 x SE. The
variances can then be calculated from the SEs according to the equation: Variance = SE2. The
sum of the variances then provides an estimate of the variance for the sum of the MLEs, and the
This document is a draft for review purposes only and does not constitute Agency policy.
5.99 DRAFT—DO NOT CITE OR QUOTE
-------
1 95% upper bound on the sum of the MLEs can be estimated by applying the above equations in
2 reverse. Tables 5-20 and 5-21 provide a summary of the results of these calculations for the
3 combined cancer mortality and incidence risks, respectively.
4
5 Table 5-20. Calculation of combined cancer mortality unit risk estimate at
6 0.1 ppm
7
Cancer type
MLE of risk
95% upper bound
on risk
SE
Variance
NPC
2.63 x 10~4
5.75 x 10~4
1.90 x 10~4
3.60 x 10~8
Hodgkin
lymphoma
2.79 x 10~4
6.17 x 10~4
2.05 x 10~4
4.22 x 10~8
Leukemia
1.87 x 10~3
3.90 x 10~3
1.23 x 10~3
1.52 x 10~6
Sum
2.41 x 1(T3
5.09 x 10~3
1.60 x 10~6
Combined risk
4.49 x 10~3
1.27 x 10~3
Combined unit
riska
(per ppm)
4.49 x 10~2
8
9 aUnit risk = 95% upper bound on combined risk/0.1 ppm.
10
11
12 Table 5-21. Calculation of combined cancer incidence unit risk estimate at
13 0.1 ppm
14
Cancer type
MLE of risk
95% upper bound
on risk
SE
Variance
NPC
7.56 x 10~4
1.62 x 10~3
5.25 x 10~4
2.76 x 10~7
Hodgkin
lymphoma
1.10 x 10~3
2.35 x 10~3
7.60 x 10~4
5.77 x 10~7
Leukemia
2.84 x 10~3
5.89 x 10~3
1.85 x 10~3
3.44 x 10~6
Sum
4.70 x 10~3
9.86 x 10~3
4.29 x 10~6
Combined risk
8.10 x 10~3
2.07 x 10~3
Combined unit
riska
(per ppm)
8.10 x 10~2
15
16 "Unit risk = 95% upper bound on combined risk/0.1 ppm.
This document is a draft for review purposes only and does not constitute Agency policy.
5-100 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
As can be seen from the results in Table 5-20, the upper bound risk estimates for cancer
mortality for the individual cancer types at 0.1 ppm are within 10% of the values that would be
obtained from the unit risk estimates derived in sections 5.2.2 and 5.2.3 (Tables 5-10, 5-16, and
5-17). Furthermore, the combined unit risk estimate for mortality for the three cancer types
(4.5 x 10 2 per ppm) is appropriately bounded by the mortality unit risk estimate for leukemia
(4.1 x 10~2 per ppm), which has the highest individual mortality unit risk estimate, and by the
sum (5.2 x 10~2 per ppm) of the individual unit risk estimates presented in sections 5.2.2 and
5.2.3. Similarly, the combined risk calculated at 0.1 ppm is necessarily bounded by the sum of
the MLEs and the sum of the 95% upper bounds for the individual risks calculated at 0.1 ppm.
Thus, the value of 4.5 x 10~2 per ppm (3.7 x 10~5 per |ig/m3) calculated at 0.1 ppm for the
combined unit risk is a reasonable estimate for the total cancer mortality unit risk (based on the
three cancer types considered).
As can be seen from the results in Table 5-21, the upper bound risk estimates for cancer
incidence for the individual cancer types at 0.1 ppm are within 33% of the values that would be
obtained from the unit risk estimates derived in sections 5.2.2 and 5.2.3 (Tables 5-13, 5-20, and
5-21). Furthermore, the combined (incidence) unit risk estimate for the three cancer types
(8.1 x 10 2 per ppm) is appropriately bounded by the unit risk estimate for leukemia
(5.7x 10"2 per ppm), which has the highest individual unit risk estimate, and by the sum (8.6 x
10~2 per ppm) of the individual unit risk estimates presented in sections 5.2.2 and 5.2.3.
Similarly, the combined risk calculated at 0.1 ppm is necessarily bounded by the sum of the
MLEs and the sum of the 95% upper bounds for the individual risks calculated at 0.1 ppm.
Thus, the value of
8.1 x 10~2 per ppm (6.6 x 10~5 per |ig/m3) calculated at 0.1 ppm for the combined unit risk is a
reasonable estimate for the total cancer unit risk (based on the three cancer types considered).
As documented in section 4.5, formaldehyde is a mutagenic carcinogen and the weight of
evidence suggests that formaldehyde carcinogenicity can be attributed, at least in part, to a
mutagenic MOA. Therefore, since there are no chemical-specific data to evaluate susceptibility
of different life stages, increased early-life susceptibility should be assumed, and, if there is
early-life exposure, the age-dependent adjustment factors (ADAFs) should be applied in
accordance with EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b). See section 5.4.4 below for more details on the
application of the ADAFs.
This document is a draft for review purposes only and does not constitute Agency policy.
5-101 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5.3. DOSE-RESPONSE MODELING OF RISK OF SQUAMOUS CELL CARCINOMA
IN THE RESPIRATORY TRACT USING ANIMAL DATA
In the previous section, dose-response analyses based on human data for
lymphohematopoietic cancer and NPC were presented. The dose-response analyses of cancer
risk presented in this section are based on nasal tumor data from laboratory bioassays using F344
rats. Because the analyses involved are extensive, most of the details are provided in the
appendices.
An increased incidence of nasal squamous cell carcinoma (SCC) was seen in two long-
term bioassays using F344 rats (Monticello et al., 1996; Kerns et al., 1983). Although other
studies in laboratory animals exist, these two studies, when combined, provide the most robust
data for analyses. These inhalation data on nasal SCC tumor incidence were used to estimate
human respiratory cancer risk in the nose and were also extrapolated to the entire respiratory
tract; in other words, a site concordance between rat and human is not assumed. This is
reasonable because the respiratory and transitional epithelial cell types considered to be at risk
of SCC in the upper respiratory tract are also prevalent in the lower human respiratory tract, and
there is greater penetration of formaldehyde flux posteriorly in the nose and in the rest of the
human respiratory tract relative to that of the rat. These considerations are strengthened by the
findings of DNA-protein cross-links (DPXs) in the proximal portions of the rhesus monkey
lower respiratory tract (Casanova et al., 1991). In addition, some epidemiologic studies
(Gardner et al., 1993; Blair et al., 1990, 1986) reported an increase in lung cancer associated
with formaldehyde exposure, while others (Collins et al., 1997; Stayner et al., 1988) reported no
such increases.
EPA's cancer guidelines (U.S. EPA, 2005a) suggest using a BBDR model for
extrapolation when data permit. A BBDR model for formaldehyde was developed by scientists
at the CUT Centers for Health Research (see Appendix D) (Conolly et al., 2004, 2003, 2000;
Kimbell et al., 2001a, b; Overton et al., 2001; CUT, 1999), which interfaced several models to
combine the extensive mechanistic information available in studies involving the F344 rat and
rhesus monkey and time-to-tumor incidence data in long-term bioassays, as shown by the
schematic in Figure 5-11. This mechanistic information included formaldehyde and DPX
dosimetry in the rat, monkey, and human airways and cell proliferation data in the rat nasal
lining. This document presents extensive evaluation of the underlying models and data and of
the alternative parametrizations of the models that were also explored for the purpose of the
current assessment (see Appendix E, Appendix F). A summary of conclusions is presented in
section 5.3.3. In particular, the following conclusions by EPA were critical in determining how
the models could be used to inform the quantitative dose-response assessment:
This document is a draft for review purposes only and does not constitute Agency policy.
5-102 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
• When used to model the dose-response in the range of the available data, the BBDR
models were judged to have the advantage of being more accurate and biologically based
(than purely statistical descriptions such as the multistage-weibull model) and allowing
utilization of various data in an integrated manner.
• Variations to the modeling in Conolly et al. (2003) were examined. Each of these models,
including the modeling in Conolly et al., was judged to be just as biologically plausible
as the other, described the rat tumor incidence equally well, was based on different
characterizations of the same empirical cell kinetic data, and was based on the same
empirical data on DPX measurements. However, the added human risk over baseline
levels estimated by these models (including the original model) ranged from negative to
large positive values at environmental exposure concentrations.
• When used for the purpose of extrapolating risk, the BBDR models did not appear to
reasonably constrain either risk estimates extrapolated to human exposures or risk
estimates for the F344 rat when they were extrapolated below the range of observable
data.
• Because human respiratory cancer risk calculated in Conolly et al. (2004) was
numerically unstable, clonal growth modeling was not found to be a useful approach for
human extrapolation of rodent risk estimates.
• Thus, the biologically based derivation of human risk estimates in Conolly et al. (2004)
cannot be characterized as a plausible upper bound in the face of model uncertainties (a
key conclusion of those authors).
For all these reasons, the BBDR modeling of the rat data
• was employed in this assessment to derive multiple PODs (for SCC in the respiratory
tract) in the range of the observed data and using model-derived internal dose estimates,
• but was not used to extrapolate far below the observed data.,
The inhalation unit risk estimates of SCC in the human respiratory tract were derived by
using multiple methods to model the F344 tumor incidence data as follows: conventional
mutistage Weibull time-to-tumor modeling and variations of the model implemented in Conolly
et al. (2003) that were considered in the process of the evaluation.
PODs were calculated as exposure concentrations corresponding to the 95% statistical
upper bound extra risks of 0.005, 0.01, and 0.05 (0.005 used only with BBDR modeling). The
inhalation unit risk for SCC in the human respiratory tract (upper and lower) derived from the
above animal bioassay data was then calculated by linear extrapolation to the origin from the
This document is a draft for review purposes only and does not constitute Agency policy.
5-103 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
POD. Linear extrapolation is supported in part by the proven genotoxicity of the chemical and
the observation of cytogenetic effects in human occupational exposures (see chapter 4). In
particular, the formation of DPXs on formaldehyde interaction with DNA has been observed at
doses well below those considered cytotoxic. In results obtained in some implementations of the
biologically based models, formaldehyde-induced mutagenicity (modeled as proportional to
DPX concentration) was found to be a critical determinant of its tumorigenicity, both at the low
dose pertaining to human exposure concentrations as well as in the dose range in which
formaldehyde is considered to be cytotoxic.
The human equivalent concentration was calculated by assuming that continuous lifetime
exposure to a given steady-state flux of formaldehyde (expressed in pmol/mm2-hour) leads to
equivalent risk of nasal cancer across species. Risk per respiratory or transitional epithelial cell
with replicative potential in a given region was computed as a function of formaldehyde flux in
the nasal region and extrapolated to the rest of the respiratory tract.
5.3.1. Long-Term Bioassays in Laboratory Animals
This section briefly describes the various animal data and dosimetry information utilized
in the above (but not in all) models, based on which estimates for the inhalation unit risk are
derived later in this chapter.
5.3.1.1. Nasal Tumor Incidence Data
Various bioassays have reported the effects of formaldehyde on rats, mice, and rhesus
monkeys and have been discussed at length earlier in this document. Two of these bioassays
(Monticello et al., 1996; Kerns et al., 1983), when combined, allow for the most robust
characterization of the long-term dose response in a laboratory species. These long-term
bioassays found an increased incidence of nasal SCCs in rats exposed to formaldehyde by the
inhalation route. In these combined data, rats were exposed to 0, 0.7, 2.0, 6.01, 9.93, and
14.96 ppm (0, 0.86, 2.5, 7.4, 12.2, and 18.4 mg/m3) exposure concentrations of formaldehyde.
SCCs were observed only at 6.01 ppm and higher exposure concentrations. Table 5-22 provides
a summary of the tumors from these bioassays, and the time-to-tumor characteristics are as
shown by the data in Figure 5-12 (in section 5.3.3). The focus here is on these two bioassays,
combined, because they provide the most extensive chronic dose-response information. Other
tumor bioassays were also conducted by various researchers and have been detailed in chapter 4.
This document is a draft for review purposes only and does not constitute Agency policy.
5-104 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Table 5-22. Summary of tumor incidence in long-term bioassays on
F344 rats
Formaldehyde
exposure, ppm
Number
of animals
Number
with SCC
Percent
with SCC
0.0
341
0
0
0.7
107
0
0
2
353
0
0
6.01
343
3
0.87
9.93
103
22
21.4
14.96
386
162
42.0
Sources: Combined data from Monticello et al. (1996) and Kerns et al. (1983).
5.3.1.2. Mechanistic Data
The Kerns et al. (1983) and Monticello et al. (1996) tumor studies were accompanied or
followed by additional studies that provided extensive mechanistic information on both
pharmacokinetics and pharmacodynamics. These studies have been summarized elsewhere in
this document and in other reviews (CUT, 1999; Monticello and Morgan, 1997; Morgan, 1997;
Heck et al., 1990). In addition to the tumor incidence data, the following data and mechanistic
information (some of which were model derived) are used in the quantitative models utilized in
this chapter. In all these cases, additional data for the rhesus monkey are also available that
inform the hazard assessment but which have not been explicitly used in deriving the inhalation
unit risk. Rhesus monkey data have been discussed in chapter 4 and chapter 3 (DPX and
formaldehyde dosimetry).
• Formaldehyde interacts with DNA to form DPXs. These cross-links are considered to
induce mutagenic as well as clastogenic effects. Casanova et al. (1994, 1989) carried out
two studies of DPX measurements in F344 rats. In the first study, rats were exposed to
concentrations of 0.3, 0.7, 2, 6, and 10 ppm for 6 hours and DPX measurements were
made over the whole respiratory mucosa of the rat, while, in the second study, the
exposure was to 0.7, 2, 6, or 15 ppm formaldehyde for 3 hours and measurements were
made at "high" and "low" tumor sites. DPX formation was observed at all exposure
concentrations in both studies (0.3 ppm - 15 ppm); the DPX levels were statistically
significantly elevated at concentrations >2 ppm, with the trend also indicating elevated
DPXs at 0.7 ppm. These data were used in the development of a PBPK model for
predicting DPX levels in the nasal lining (see chapters 3 and 4).
This document is a draft for review purposes only and does not constitute Agency policy.
5-105 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
• Male F344 rats were exposed to formaldehyde gas over a range of concentrations (0, 0.7,
2, 6, 10, or 15 ppm) in two phases of a labeling study. The first phase (Monticello et al.,
1991) employed injection labeling with a 2-hour pulse labeling time, and animals were
exposed to formaldehyde for periods of 1, 4, and 9 days and 6 weeks. The second phase
(Monticello et al., 1996) used osmotic mini pumps for labeling with a 120-hour release
time to quantify labeling in animals exposed for 13, 26, 52, and 78 weeks. These data
have been analyzed at length in Appendix E.
• Physical and computer models of airflow in anatomically realistic representations of the
F344 rat and human upper respiratory tract have been constructed (Kimbell et al., 1993,
1997; Kepler et al., 1993; Subramaniam et al., 1998; see chapter 3).
• Regional uptake of formaldehyde has been calculated for the upper respiratory tract of
the rat and human by using the above computer representations and for the lower
respiratory tract of the human by using an idealized representation of the human lower
respiratory tract (Kimbell et al., 2001a; Overton et al., 2001; also see chapter 3 and
further discussion of uncertainties in Appendix F).
5.3.2. The CUT Biologically Based Dose-Response Modeling
The studies mentioned above in 5.3.1.1 and 5.3.1.2 were generated at the CUT Centers
for Health Research and led to the development of a biologically motivated dose-response model
for formaldehyde-induced cancer as represented in a series of papers and in a health assessment
report (CUT model) (Conolly et al., 2004, 2003, 2000; Conolly, 2002; Kimbell et al., 2001a, b;
Overton et al., 2001; CUT, 1999). EPA's cancer guidelines (U.S. EPA, 2005a) suggest using a
BBDR model for extrapolation when data permit since it facilitates the incorporation of MO A in
risk assessment. The CUT modeling and available data were evaluated in a series of peer-
reviewed papers (Klein et al., 2009; Crump et al., 2008; Subramaniam et al., 2008, 2007) and
debated further in the literature (Conolly et al., 2009; Crump et al., 2009). Alternatives to the
parametrization and model structure in the CUT biological modeling (but based on that original
model) are further explored and evaluated in this assessment (Appendix E). Appendix F carries
out a sensitivity analysis of the human risk estimates in Conolly et al. (2004) based on key
uncertainties evaluated in Appendix E. These BBDR models are used in this assessment to
calculate PODs from the dose-response curve for the F344 rat nasal tumor risk. Extrapolation to
human is then carried out by using EPA's baseline ("default") approach (U.S. EPA, 1994) but
using model-derived internal dose metrics. See section 5.3.3 for rationale supporting these
decisions.
This document is a draft for review purposes only and does not constitute Agency policy.
5-106 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
First, the key features of the BBDR modeling in Conolly et al. (2003, 2004) are briefly
described, and the following notation is used throughout this section: N cell = normal cell; I cell
= initiated cell; LI = labeling index and is equal to the number of labeled cells/(number labeled +
unlabeled cells); ULLI = unit length LI equal to the number of labeled cells/length of basement
membrane; ocn = division rate of normal cells (hour-1); |i\ = rate at which an initiated cell is
formed by mutation of a normal cell (per cell division of normal cells).
In Conolly et al. (2003), tumor incidence data in the Kerns et al. (1983) and Monticello et
al. (1996) long-term bioassays were modeled by using an approximation of the two-stage clonal
growth model (Moolgavkar et al., 1988) and allowing formaldehyde to have a direct mutagenic
action. Conolly et al. (2003) combined these data with historical control data on 7,684 animals
obtained from National Toxicology Program (NTP) bioassays. These models are based on the
Moolgavkar, Venzon, and Knudson (MVK) stochastic two-stage model of cancer (Moolgavkar
et al., 1988; Moolgavkar and Knudson, 1981; Moolgavkar and Venzon, 1979), which accounts
for growth of a pool of normal cells, mutation of normal cells to initiated cells, clonal expansion
and death of initiated cells, and mutation of initiated cells to fully malignant cells.
The MVK model for formaldehyde accounted for two MO As as follows that may be
relevant to formaldehyde carcinogenicity:
1. An indirect MOA in which the regenerative cell proliferation in response to formaldehyde
cytotoxicity increases the probability of errors in DNA replication. This MOA was modeled
by using labeling data on normal cells in nasal mucosa of rats exposed to formaldehyde.
2. A possible direct mutagenic MOA, based on information indicating that formaldehyde is
mutagenic (Speit and Merk, 2002; Heck et al., 1990; Grafstrom et al., 1985), was modeled by
using rat data on formaldehyde production of DPXs (Monticello et al., 1996, 1991). In
Conolly et al. (2003), the intracellular dose that induces mutations is considered proportional
to the local DPX dose.
The human model for formaldehyde carcinogenicity (Conolly et al., 2004) is
conceptually very similar to the rat model. The model uses, as input, results from a dosimetry
model for an anatomically realistic representation of the human upper airways and an idealized
representation of the lower airways. However, the model does not incorporate any data on
human responses to formaldehyde exposure.
A novel contribution of the CUT model, described by the schematic in Figure 5-11, is
that cell replication rates and DPX concentrations are driven by local dose, which is
formaldehyde flux to each region of nasal tissue expressed as pmol/mm2-hour. This dosimetry is
predicted by computational fluid dynamics (CFD) modeling using anatomically accurate
This document is a draft for review purposes only and does not constitute Agency policy.
5-107 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
representations of the nasal passages of a single F344 rat or Caucasian male human (see
chapter 3). Such a feature is important in incorporating site-specific toxicity in the case of a
highly reactive gas like formaldehyde for which uptake patterns are spatially localized and
significantly different across species (see chapter 3). In the CUT model, each of these
parameters is characterized by local flux (see Figure 5-11). The inputs to the two-stage cancer
modeling consisted of results from other model predictions as well as empirical data as follows:
• Regional uptake of formaldehyde in the respiratory tract was predicted by using CFD
modeling in the F344 rat and human (Kimbell et al., 2001a, b; Overton et al., 2001;
Subramaniam et al., 1998).
• Normal cell replication rates were inferred from LI data on rats exposed to formaldehyde
(Monticello etal., 1996, 1991, 1990).
• Concentrations of DPXs linked to the regional flux of formaldehyde were predicted by a
PBPK model (Conolly et al., 2000) calibrated to fit the DPX data in F344 rat and rhesus
monkey (Casanova et al., 1994, 1991) and subsequently scaled up to humans. The DPX
concentration levels were incorporated into the two-stage clonal expansion model by
defining mutation rate of normal and initiated cells as the same linear function of DPX.
That is,
l^N = [ll = (J-Nbasal + KMU X DPX (5-1)
where |i\ is the rate at which an initiated cell is formed by mutation of a normal cell (per
cell division of normal cells), and likewise |ii is the rate at which a malignant cell is
formed by mutation of an initiated cell (per cell division of initiated cells). The unknown
constants (j,Nbasai (the baseline rate) and KMU were estimated by fitting model predictions
to the tumor bioassay data.
The rat model in Conolly et al. (2003) involved six unknown statistical parameters that
were estimated by fitting the model to the rat formaldehyde bioassay data shown in Table 5-22
(Monticello et al., 1996; Kerns et al., 1983) plus data from several thousand control animals
from all the rat bioassays conducted by the NTP. These NTP bioassays were conducted from
1976 through 1999 and included 7,684 animals with an incidence of 13 SCCs (i.e., 0.17%
incidence). The resulting model predicts the probability of a nasal SCC in the F344 rat as a
function of age and exposure to formaldehyde. The fit to the tumor incidence data is shown in
Figure 5-12 (in section 5.3.3.). (For later reference in Appendix E, this figure compares the fit to
This document is a draft for review purposes only and does not constitute Agency policy.
5-108 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
the data obtained by the modeling in Conolly et al. [2003] with that obtained by the
reimplementation of this model in Subramaniam et al. [2007].)
Subsequent to the BBDR model for modeling rat cancer, Conolly et al. (2004) developed
a corresponding model for humans for the purpose of extrapolating the risk to humans estimated
by the rat model. Also, rather than considering only nasal tumors, the model is used to predict
the risk of all human respiratory tumors. The human model for formaldehyde carcinogenicity
(Conolly et al., 2004) is conceptually very similar to the rat model and follows the schematic in
Figure 5-11. The following points need to be noted:
• The model does not incorporate any data on human responses to formaldehyde exposure.
• The model is based on an anatomically realistic representation of the human nasal
passages (in a single individual) and an idealized representation of the lower respiratory
tract. Local formaldehyde flux to respiratory tissue is estimated by a CFD model for
humans (Kimbell et al., 2001a; Overton et al., 2001; Subramaniam et al., 1998).
• Rates of cell division and cell death are, with a minor modification, assumed to be the
same in humans as in rats.
• The concentration of formaldehyde-induced DPXs in humans is estimated by scaling up
from values obtained from experiments in the F344 rat and rhesus monkey (Conolly et
al., 2000, and also discussed further in section 3.6.6 of this document). The statistical
parameters for the human model are either estimated by fitting the model to the human
background data, assumed to have the same value as that obtained in the rat model, or, in
one case, fixed at a value suggested by the epidemiologic literature. The human value for
KMU in eq 1 is obtained by assuming that the ratio KMU/|ib;lsai is invariant across
species.
Some further clarification pertaining to the structure and calibration of the models in
Conolly et al. (2004, 2003) that are key to understanding model assumptions is provided in
Appendix D.
This document is a draft for review purposes only and does not constitute Agency policy.
5-109 DRAFT—DO NOT CITE OR QUOTE
-------
^ ^2
Co Si
f>Q O
SS
^5
0
1
S
I
CSj
§•
Zn
§
W
%¦
C>
si
s
s
>3
%
o
o
Figure 5-11. Schematic of integration of pharmacokinetic and pharmacodynamic components in the CUT
model.
Note: (3 = death rate; (.i = mutation rate per cell division; an, N(t), |i\ are informed (partially or fully) by empirical data;
other parameters are estimated by fitting to tumor incidence data.
-------
1
2
3
4
5
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
5.3.2.1. Major Results of the CUT Modeling Effort
Based on the biologically based modeling of the rat SCC data, CUT (1999) and Conolly
et al. (2004, 2003) presented the following major conclusions. The evaluation of the strength of
these conclusions is summarized in section 5.3.3., and as addressed in that section, this current
assessment is not in agreement with these conclusions.
• The putative, directly mutagenic action of formaldehyde "does not play a significant role
in the tumor response in the rat (and also in the human), [and such a conclusion] should
be robust for any potentially mutagenic effect of formaldehyde with a time course similar
to that of DPX."
• Respiratory cancer risks associated with inhaled formaldehyde are de minimis (10 6 or
less) at relevant human exposure levels. This was based on using an upper bound on the
model estimate for the directly mutagenic action of formaldehyde.
• Therefore, exposure standards protective of effects of formaldehyde-induced cytotoxicity
should be sufficient to protect from its potential carcinogenic effects.
• The human risk estimates in Conolly et al. (2004) were judged by the authors to be
conservative in the face of model uncertainties because the model included a hockey-
stick model for normal cell replication rates when the cell replication dose-response
curve as averaged by the authors had a J shape, used overall respiratory tract cancer
incidence data in humans, and evaluated the model at the statistical upper bound of the
proportionality parameter relating DPXs to the probability of mutation.
• The dose-response assessment in Conolly et al. (2004) did not explicitly evaluate the risk
of lymphohematopoietic cancers. However, Conolly et al. (2004) argued that
formaldehyde was unlikely to cause the cancers reported in Hauptmann et al. (2003).
Their reasoning was based on the steepness of the dose-response curve predicted in
Conolly et al. (2004) for respiratory cancer at exposures of 1 ppm and above and the
conclusions in Heck and Casanova (2004).
5.3.3. This Assessment's Conclusions from Evaluation of Dose-Response Models of DPX,
Cell-Replication and Genomics Data, and of BBDR Models for Risk Estimation
The CUT modeling and alternative approaches that were developed based on the
conceptual framework in that modeling were extensively evaluated for this assessment and are
presented in Appendices D, E (BBDR modeling of the rat data), and F (sensitivity analysis of
BBDR model results for human risk). In particular, Table E-l in Appendix E and Table F-l in
Appendix F list all the uncertainties and assumptions that were examined and summarize the
results of the evaluation. The quantitative and qualitative characterization of the cell replication
data from Monticello et al. (1996, 1991) are presented in Appendix E. The most significant
This document is a draft for review purposes only and does not constitute Agency policy.
5_111 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
conclusions resulting from these various analyses, focusing on the ones that have maximal
impact on the dose-response assessment, are presented below.
Description of time-to-tumor data
The overall approach and use of data in Conolly et al. (2004, 2003) have substantial
advantages to offer in describing the dose response observed in animal bioassays. The authors'
model provides a good statistical description of the time-to-tumor data. The fit to the data was
found to be superior to that obtained by using multistage Weibull time-to-tumor modeling of the
tumor incidence data (comparison based on visual inspection [see Figure 5-12 in this section and
Figures 5-17, 5-18, 5-19 in section 5.3.4]).
.1?
is
0.0020
0.0015
0.0010
0.0005
0.0000
0 ppm
200
400 600 800 1000
0.20-1
0.15
0.10
0.05
0.00
6 ppm
/ /
/ /
0
200
400 600 800 1000
1.0-,
O.i
0.6-
0.4-
0.2-
0.0-
10 ppm
/ /
/
200 400 600 800 1000
Age (days)
1.0
0.8
0.6
0.4
0.2
0.0
15 ppm
0 200 400 600 800 1000
Age (days)
- Kaplan-Meier Conolly fit
¦ Our fit Subramaniam et al. fit
Figure 5-12. Fit to the rat tumor incidence data using the model and
assumptions in Conolly et al. (2003).
Note: Fitting was performed on data of Kerns et al. (1983) and Monticello et al. (1996) combined
with ALL NTP historical controls under the assumption that all SCCs are fatal. Figure compares
the fit obtained by Conolly et al. (2003) with the reproduction of these results under identical
conditions, inputs, and assumptions by Subramaniam et al. (2007). There were minor residual
differences among the implementations; see the appendix in Subramaniam et al. (2007) for
explanation.
Source: Subramaniam et al. (2007). Reprint permission required.
This document is a draft for review purposes only and does not constitute Agency policy.
5_112 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Integration of various relevant data
The model framework integrates various pharmacokinetic and pharmacodynamic
components (regional formaldehyde flux, DPX, cell-replication, and tumor incidence data)
within a single conceptual framework and thus facilitates description of the dose response that
utilizes the extensive mechanistic information available for formaldehyde.
Regional dosimetry
Regional (site-specific) dosimetry in the upper respiratory tract is considered important
for understanding the tumorigenicity of a reactive chemical like formaldehyde. The regional
dosimetry models discussed in chapter 3 compute local formaldehyde flux to the tissue and are
based on anatomically realistic constructions of the nasal airways in each species. The other
relevant mechanistic data, DPX and cell replication, are expressed as a function of this local
formaldehyde flux.
Confidence in dosimetry
Model predictions of formaldehyde flux to the respiratory lining have not been verified
experimentally, and such verification would present formidable experimental challenges.
Overall, the formaldehyde dosimetry modeling utilized in the CUT modeling presents a
reasonable level of confidence, as detailed in chapter 3, section 3.6, by virtue of agreement
among multiple model predictions (models that predict airflow profiles as well as a PBPK model
for DPX, which uses the calculated formaldehyde flux as input) and various kinds of available
data. These data comprise airflow profiles in physical casts of the nasal cavity of an F344 rat
(Kimbell et al., 2001a), a human (Subramaniam et al., 1998), and a rhesus monkey (Kepler et al.,
1998); DPX data (see discussion of Cohen-Hubal et al. [1997] in chapter 3); and qualitative
concordance between uptake patterns and cell proliferation (Morgan et al., 1997; Monticello et
al., 1996). The CFD models of formaldehyde flux represent only an individual of each species.
However, considerable interindividual differences are to be expected in the regional dosimetry,
particularly in the human (Garcia et al., 2009: Subramaniam et al., 2008). This is discussed
briefly in Chapter 3 (section 3.6) and further in Appendices B and F.
Control tumor data
In developing their model, Conolly et al. (2004, 2003) included control rats from all NTP
cancer bioassays—a total of 7,684 rats. The inclusion of all NTP historical control animals does
not appear to be supportable and substantially alters dose-response predictions (Crump et al.,
2009, 2008; Subramaniam et al., 2008, 2007). There are legitimate questions regarding
This document is a draft for review purposes only and does not constitute Agency policy.
5-113 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
comparability of results in rats from different stocks, studied at different times, in different
laboratories, and by different routes of exposure and evaluated by using somewhat different
pathological procedures. If historical controls are used from only those inhalation studies that
present a low potential for genetic and time-related variations in tumor incidence and survival of
animals or if only concurrent controls are used, the model for extrapolation of risk to humans
(the human BBDR model) becomes numerically unstable. In such a model, it is not possible to
bound human risk by using the extrapolation approach applied in the CUT model. When the
included NTP control data were restricted to all NTP inhalation controls, the upper bound human
risk estimate obtained by Conolly et al. (2004) was increased by 50-fold (Crump et al., 2008).
Cell replication dose response
As discussed in chapter 4, characterization of the uncertainties and variability in the cell
replication dose response is crucial to understanding formaldehyde carcinogenicity. Analyses of
the dose response for cell replication presented in Appendix E demonstrate the following:
• Sustained exposure to formaldehyde affects cell division rates (compared to baseline
levels). This effect is seen over a continuum of formaldehyde flux to the nasal lining that
includes flux levels below those thought to be cytotoxic.
• Given the qualitative and quantitative uncertainties in the data and in their interpretation,
a variety of cell replication dose-response models are plausible as reasonable
characterization of the data. Cell replication response differs substantially among nasal
sites and over time during the course of the bioassay. In consideration of these
differences, the shape of the cell replication low-dose response could be alternatively
described as monotonic or non-monotonic. For example, rather different statistical
descriptions of the data result depending on whether
i. different sites and exposure times were modeled separately;
ii. all exposure times were pooled to model the response at each site;
iii. the labeling index was time-weighted and averaged over all sites;
iv. flux and labeling index were weighted by the number of cells at a given site;
v. the short exposure durations in Monticello et al. (1991) were examined separately. In
addition, transient increases in cell turnover at sub-cytotoxic doses are seen in other
experiments in rats exposed to formaldehyde (see chapter 4).
• At higher, cytolethal formaldehyde flux levels, regenerative hyperplasia-induced cell
proliferation clearly takes over.
This document is a draft for review purposes only and does not constitute Agency policy.
5_114 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Genotoxicity
Chapter 4 provides multiple lines of evidence to characterize formaldehyde as a
genotoxicant. Of particular note is the observation of cytogenetic effects at human occupational
exposures and the formation of DPXs upon formaldehyde interaction with DNA at doses well
below those considered cytotoxic. As noted earlier, DPX formation was detected in rats at
exposures ranging from 0.3 ppm to 15 ppm. These DPX levels are seen to be statistically
significantly increased over baseline levels at 2 ppm and above. The DPX measured at 0.7 ppm
shows a trend that is consistent with an increase at this dose (see chapter 3); while not
statistically significant, it is critical to consider "trend" when analyzing low-dose data.
Inferences on MO A from modeling the data
The highly curvilinear nature of dose responses associated with DPX formation, LI data,
and tumor response as well as mechanistic interpretation of these observed data has provided
grounds for arguments in the literature that formaldehyde tumorigenicity (at exposures >6 ppm)
should be uncoupled from its potential carcinogenicity in the low-dose region. Furthermore,
researchers have argued that any potential low-dose risk is due to its mutagenicity, that this
mutagenic potential is too weak to be of significance, and that the high-dose risk is entirely due
to cell proliferation induced by regenerative hyperplasia in response to cell injury at cytotoxic
doses (i.e., without a relevant role for the direct mutagenic action of formaldehyde). Conolly et
al. (2004, 2003) represented a quantitative expression of this point of view. However, alternative
parametrizations of the model in Conolly et al. (2004, 2003) have shown that the mutagenic
component can be important in explaining the high-dose effect and that the risk at low dose due
this mutagenicity can be significant (Subramaniam et al., 2007; Appendix E). Accordingly, the
dose-response assessment in this document does not treat formaldehyde as a threshold
carcinogen.
Of further relevance to mode-of-action considerations, analyses detailed in Appendix E
indicate that the chronic rat nasal time-to-tumor incidence data can be quantitatively explained,
and with equal force, by invoking any of the following multiple sets of plausible events induced
by formaldehyde. The role of spontaneously occurring mutations and increased cell turnover
rates in response to various baseline insults to the nasal lining are common to all these scenarios
(and are not separately mentioned).
a. Mutations occur over a dose continuum that includes sub-cytotoxic and cytotoxic levels of
exposure. The only cell proliferation induced by formaldehyde is a regenerative response at
cytotoxic concentrations.
This document is a draft for review purposes only and does not constitute Agency policy.
5-115 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
b. Cytotoxicity-induced regenerative cell proliferation occurs, but there is no significant
formaldehyde-induced mutational effect. This latter scenario expresses arguments that
formaldehyde-induced mutagenicity may be too weak to be of significance to its
tumorigenicity.
c. Both mutations and cell proliferation are induced by formaldehyde only at cytotoxic levels.
d. Mutational events occur and cell replication is altered over a continuum that includes low
and high levels of exposure.
i. At high dose, the effect on cell replication is regenerative.
ii. At lower doses, the data indicate that both monotonic and non-monotonic dose-
response curves for cell replication are plausible.
iii. With respect to the previous argument, the following result was very instructive. The
models were exercized with normal cell replication rates considered to be less than
(non-monotonic) or equal to (threshold) baseline rates over a segment of the low-dose
range in conjunction with the chronic time-to-tumor data for the F344 rat. Such a
scenario did not necessarily lead to lower than baseline or threshold in formaldehyde
respiratory cancer risk in the rat in that low-dose range. This is partly because there
are no data to inform how formaldehyde-induced mutation might alter cell replication
and apoptotic rates (in particular if the mutation is to be construed as an initiating
event in the carcinogenesis).
e. Formaldehyde-induced mutagenic action acts only in concert with baseline cell turnover at
low dose.
Kinetics of initiated cells
Modeling results are hypersensitive to the division and death rate of initiated cells that
cannot be further inferred by the available empirical cell labeling data (Conolly et al., 2009;
Crump et al., 2009, 2008). Several plausible alternate model structures for describing initiated
cell kinetics, none of which degrade the agreement of the model with the underlying data used to
construct the model originally, led to low-dose risk estimates in the rodent that varied by many
orders of magnitude, including negative values (see Figures E-5A,B and E-6A,B in Appendix E).
Extremely small perturbations in the division rate (and, likewise, of death rates) of initiated cells
in the model lead to human risk estimates ranging anywhere from negative values to +0.01 at
0.01 ppm (see Crump et al. 2008 and Appendix F, Figure F-5). These perturbations were small
compared with the normal variation in the division rates of normal cells.
The sensitivity analyses on the basis of which these conclusions were reached have been
criticized as resulting in implausible risk estimates (given the epidemiologic data) as a
This document is a draft for review purposes only and does not constitute Agency policy.
5-116 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
consequence of implementing model variations that are not biologically reasonable (Conolly et
al. 2009). This criticism was rebutted by Crump et al. (2009) on biological and epidemiological
grounds. These debates are discussed fully in Appendix F.
In addition, there are major qualitative uncertainties in extrapolating normal cell
replication rates from the rat to human (Table F-l in Appendix F, and Subramaniam et al.
[2008]). Subramaniam et al. (2008) concluded that several inferences that arise from the
assumptions in the CUT model on initiated cell replication and death rates cannot be supported
by available biological information.
Risk extrapolation
Use of the modeling approach in Conolly et al. (2004) or the variations examined were
determined not to be informative for extrapolation from animal to human at any exposure
concentration because of extreme sensitivity, including numerical instability, to uncertain model
assumptions. In the face of model uncertainties, the biologically based derivation of human risk
estimates of 10~6 or less at exposures of 0.1 ppm and below in Conolly et al. (2004) or CUT
(1999) cannot be characterized as a plausible upper bound.
The use of this model for extrapolation of risk from high to low exposures in the rodent
followed by a conventional (default) approach to extrapolate the low-dose animal risk to the low-
dose human risk was next evaluated. This avenue was also found not to be informative because
the models do not adequately constrain risk in the rodent. For example, various model
representations as shown in Figure E-6A,B in Appendix E were used to evaluate added MLE risk
at the 10 5 level (Figure F-5A,B in Appendix E) in the F344 rat. Human exposures were then
calculated that would result in equivalent lifetime risk by using formaldehyde flux estimated in
each species as the dosimeter and conventional extrapolation methods (U.S. EPA, 1994b). A 25-
fold difference was found between the different models in the equivalent exposure concentration
so derived. Therefore, the CUT model or its variations were also not used in this assessment as a
biologically based or biologically motivated means of extrapolating outside the observed dose-
response in the F344 rat. Model uncertainty was substantially higher than the statistical
uncertainty arising out of a given model specification.
Thus, in view of all the above considerations and in accordance with EPA's cancer
guidelines (U.S. EPA, 2005a), the derivation in this document of unit risk for human respiratory
cancer from animal bioassay data is based on a linear extrapolation to the origin from a POD on
the dose-response curve. Low-dose linearity was exhibited by the risk estimates from most of
the models that were examined in the sensitivity analysis (see discussion surrounding Figure E-
5A,B in Appendix E).
This document is a draft for review purposes only and does not constitute Agency policy.
5-117 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
BBDR modeling for deriving an "integrated" POD
The CUT BBDR modeling approach provides a good fit to the time-to-tumor data and
therefore allows for an appropriate determination of a POD while at the same time incorporating
a large amount of mechanistic information in an integrated manner and allowing the use of
model-derived internal dose estimates. Thus, use of this model provides an alternative to
developing separate PODs based on several of the underlying components of the data, such as
DPX, flux, and labeling data. Accordingly, the model is used in this assessment to derive a POD
from a dose response, based on the nasal cancers in rats. Uncertainties in the derivation of the
POD were represented by using the variations of the CUT model examined in this chapter.
These POD calculations as well as others are detailed below.
Genomics data
The genomics data of Thomas et al. (2007) and Andersen et al. (2008) provide additional
insight into formaldehyde's biological effects and the steep dose-response curve for
tumorigenesis. However, there are various limitations in the interpretation of these genomics
data and their relevance for the pathways contributing to the disease process in humans. In
particular, the data from these studies, as analyzed, do not inform the critical MOA questions
pertaining to formaldehyde carcinogenicity. These insights have been elaborated in a separate
section in chapter 4, and the difficulties in the use and interpretation of the quantitative modeling
of these data, as presented in these studies, are detailed at length in Appendix G.
5.3.4. Benchmark Dose Approaches to Rat Nasal Tumor Data
This section describes various BMD analyses to determine PODs for low-dose
extrapolation of SCC risk in the human respiratory tract (upper and lower).
5.3.4.1. Benchmark Dose Derived from BBDR Rat Model and Flux as Dosimeter
5.3.4.1.1. Response for benchmark dose. Typically, the BMD is calculated at the 5 or 10%
response level. However, it appears appropriate to consider the benchmark response (BMR) at
lower levels in exceptional cases that are supported by empirical data. In the case of the
combined Kerns et al. (1983) and Monticello et al. (1996) bioassays, the lowest observed tumor
response of SCC was below the 1% level (at 0.85%) (see Table 5-22). Additionally, the BBDR
modeling incorporates precursor response in the form of LI data. Therefore, it was determined
that it would also be appropriate to evaluate the POD at the 0.5% level while still staying in the
neighborhood of the experimentally observed response.
This document is a draft for review purposes only and does not constitute Agency policy.
5-118 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
The various data presented earlier in this chapter point to highly curvilinear dose
responses for formaldehyde-induced tumor incidence as well as DPX and cell replication. This
is also borne out by dose-response information based on gene array data (Thomas et al. 2007;
Andersen et al. 2008). Cytotoxicity-driven regenerative replication and epithelial degeneration
play a critical role in the steeply rising nature of the tumor dose-response. These observations
raise the concern that cancer potency derived by straight-line extrapolation from the low end of
observed tumor data (roughly at the 1% response) has the potential to be a significant
overestimate for a reasonable upper bound. Thus, a pertinent question is what is a low-dose
linear dose-response modeling of the data that is statistically consistent with the uncertainties in
the observed time-to-tumor data. To address this question, the risk estimate based on the linear
extrapolation (from a POD to the origin) is compared with that predicted at the low-dose end by
the Multistage Weibull model fitted to the observed time-to-tumor data. The unit risk based on
this model is obtained by calculating ql*, the 95% statistical upper bound on the coefficient
associated with the linear term in the multistage model polynomial. This model fits the data
reasonably well and reflects the highly curvilinear shape of the dose-response because of its
mathematical flexibility while also allowing for low-dose linearity. In particular, it has been
noted that even in cases where the first term (ql) in the multistage model is zero, the upper
bound (ql*) is linear with dose (Subramaniam et al., 2006; Guess et al., 1977). Thus, for
comparison the following estimates of unit risk are also presented:
1) Unit risk that is based on ql *, which is derived from fitting the multistage Weibull model to
the observed data.
2) Unit risk based on low-dose linear extrapolation from a POD at the 0.5% level.
5.3.4.1.2. Dose metric. The dose metric used for the extrapolation was the average wall mass
flux of formaldehyde (expressed in pmol/mm2-hour to the entire surface of the airway lining but
excluding tissue lined by non-mucus-coated squamous tissue, which was considered to not
absorb formaldehyde). The use of flux as a dosimeter is similar to the calculation of a regional
gas dose ratio (RGDR) as proportional to minute volume divided by the surface area in the given
species and is thus in line with EPA's prescription for calculating a dosimetric adjustment factor
(DAF) for category 1 gases, whose effects are presumed to be at the POE (U.S. EPA, 1994b)
(i.e., ratio of average flux over the same respiratory region in each species = ratio of the quantity
[minute volume/surface area of the region] between the two species). This lends support to an
interspecies extrapolation based on the equivalence of formaldehyde flux as a determinant of
risk.
This document is a draft for review purposes only and does not constitute Agency policy.
5-119 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
The spatial distribution of formaldehyde over the nasal lining was characterized by
partitioning the nasal surface by formaldehyde flux to the tissue, resulting in 20 "flux bins"
(Figure 5-13). Each bin is comprised of elements (not necessarily contiguous) of the nasal
surface that receive a particular interval of formaldehyde flux per ppm of exposure
concentration (Kimbell et al., 2001a). The spatial coordinates of elements comprising a
particular flux bin are fixed for all exposure concentrations, with formaldehyde flux in a bin
scaling linearly with exposure concentration (ppm). The number of cells at risk varies
across the bins, as shown in Figure 5-14.
O ^
2 E .
LU £ 1500
E £
^ o
fc £ 1000
Q- a.
x
il 500
10
Flux Bin No.
3e+7
w 2e+7
"35
o
1 e+7 -
5 10 15
Flux Bin No.
20
Figure 5-13. Spatial distribution of
formaldehyde over the nasal lining, as
characterized by partitioning the nasal
surface by formaldehyde flux to the
tissue per ppm of exposure
concentration, resulting in 20 flux bins.
Figure 5-14. Distribution of cells at
risk across flux bins in the F344 rat
nasal lining.
Source: Subramaniam et al. (2008).
Source: Subramaniam et al. (2008).
5.3.4.1.3. Extrapolation to humans. For linear extrapolation from the 0.5 and 1% levels, two
alternative versions of the biologically based model in Conolly et al. (2003) for the F344 rat
were used. In both cases, only the historical control data from NTP inhalation studies data were
added to the concurrent controls and weekly averaged DPX concentrations as calculated by a
variant of the PBPK model in Conolly et al. (2000) [described in Subramaniam et al. (2007)]
were used. Both models provided good fits to the tumor incidence data, similar to the fit shown
in Figure 5-12. Neither model could be considered better than the other on the basis of model
description of tumor incidence data.
This document is a draft for review purposes only and does not constitute Agency policy.
5-120 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
In Model 1 the normal cell replication dose response was described by the same
hockey-stick-shaped curve used in Conolly et al. (2003). The form of the dose-response
curves for initiated cell kinetics (division and death) was also the same as that considered by
Conolly et al. (2003).
Model 2 was an alternative to the Conolly et al. (2003) model, and is considered in
the sensitivity analysis described in Appendix E. The dose response for normal cell
replication was monotone increasing and did not exhibit a threshold in dose. This was
obtained by fitting the 13-week cell replication data. The cell replication dose response for
initiated cells was a sigmoidal-shaped curve, increasing monotonically with flux from a
background value up to an asymptotic value. The baseline cell-replication for initiated cells
was constrained to not be less than that for normal cells. Initiated cell death rate was
considered proportional to initiated cell birth rate.
Models 1 & 2 predicted monotonic dose-response curves.
The sequence of steps in arriving at a unit risk for SCC in human nasal airways from
a given BBDR modeling of the F344 rat nasal tumor incidence data is outlined below.
Extrapolation to the lower respiratory tract is described later.
1. Calculate the MLE risk and 95% upper confidence bound on risk at various exposure
concentrations (dRAxin ppm) by exercising the two BBDR models. Here, the POD is
defined as dRAT for which the 95% upper bound added risk is either 0.005 or 0.01. These
values approximate the 95% lower bounds on the BMD corresponding to the added risks
(i.e., the BMDLrat)-
2. Using CFD modeling simulations in Kimbell et al. (2001a, b), calculate the average flux
over the entire rat nose at resting breathing rates corresponding to dRAT- Here, the
subscript "i" is over flux bins and N is the number of cells at risk in a given bin.
AvgI''lux(dR jr ) = x
i ppm
2>.
(5-2)
1RAT
3. The experiment exposure was for periods of 5 days/week, 6 hours/day. Therefore,
calculate the average daily exposure, obtained by making a 5/7 x 6/24 duration
adjustment; that is, 5/7 x 6/24 x AvgFlux(dRAT)-
4. Now assume that lifetime exposure to similar levels of average formaldehyde flux to
cells at risk leads to similar lifetime risk (MLE or upper bound, respectively) of tumor
This document is a draft for review purposes only and does not constitute Agency policy.
5-121 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
incidence across animal species. Also, in calculating human equivalent concentrations,
EPA has traditionally assumed chronic animal laboratory exposure scenarios to be
equivalent to human lifetime exposures (U.S. EPA, 1994b).
5. Since a CFD model for a human is available (Subramaniam et al., 1998), it is possible to
determine the average wall mass flux in this particular human nose for any specific
breathing scenario. Likewise, a computational model to determine mass flux at any
specific lung depth was available in the form of the single-path model of Overton et al.
(2001); however, risk in the lower respiratory tract will be addressed later. From the
human CFD simulations in Kimbell et al. (2001a, b), the human airborne exposure
concentration level that would yield the same average wall mass flux in the human nose
as [(5/7) x (6/24) x AvgFlux(dRAT)] is then calculated. In other words, given a risk-
specific dose in the rat, the equivalent human exposure concentration is given by
d.
= (5 / 7) x (6 / 24) x AvgFlux(dRAT ) x
.¦ PPm
(5-3)
HUMAN
6. To use this equivalent human exposure concentration, make the following assumption:
when humans are exposed to the above concentration of formaldehyde (dHuMAN)
throughout the course of a lifetime, the added risks are anticipated to be similar to those
experienced by the animal in the chronic bioassay.
7. Let/denote the ratio of the average flux per ppm of exposure concentration in the two
species:
i PPm
/=
(5-4)
, ppm
2>.
Now, the olfactory epithelium comprises a substantial fraction of nasal tissue in the rat.
Because the olfactory region in the rat projects directly in the path of main airstreams
This document is a draft for review purposes only and does not constitute Agency policy.
5-122 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
(Kimbell et al., 1997), a sizable flux of formaldehyde is delivered to this region in the
rat. Tumors were not observed in the olfactory tissue of the rat. Therefore, since effects
observed in the rat are being extrapolated to the human, cells from olfactory tissue are
excluded in calculating average flux in the rat in the eq 4. For the human, on the other
hand, both volumetric flow (2.5%, Subramaniam et al. [1998]) and surface area (-5%,
Kelly et al. [2000]) for the olfactory region are relatively small, so inclusion of this
region is not likely to make a difference of much significance in the calculation of
average flux in the human. Since data on formaldehyde flux delivered to the human
olfactory region were not readily available, the olfactory region was not excluded for the
human. The average human flux calculated here uses a working level classification for
the activity profile where an individual spent equal amounts of time in a day at resting
and light and moderate activity levels, corresponding to minute volumes of 7.5, 9, and
25 L/minute, respectively. This resulted in the following ratio13:
/= 444[rat]/956.4[1_] = 0.46 (5-5)
8. The airborne exposure concentrations dmjMAN corresponding to a given MLE and upper
bound lifetime added risk levels are the human BMDHuman and BMDLHuman,
respectively. These are shown in Figure 5-15. (The rather sudden increase by -0.0015
in the upper confidence bound on risk for model 1 for exposure exceeding -0.41 ppm
could not be explained. This jump was verified by repeated calculations that used
different initial simulation conditions and convergence criteria.)
Next, the human lower respiratory tract is also considered to be potentially at risk.
Therefore, the above calculations of BMD and BMDL need to be augmented to include the
lower respiratory tract for humans. This calculation was facilitated by dosimetry
calculations of formaldehyde wall mass flux to various depths in the lung by using a single
path model. Refer to Overton et al. (2001) for details on their modeling. The calculations
for including the lower respiratory tract in determining an overall BMD and BMDL
involved the following steps:
a. As given by eq 5-3, calculate dHUMAN for various MLE risk levels. This gives a dose-
response relationship for lifetime risk of SCC in the human nose due to continuous
exposure to airborne formaldehyde.
13 This is to be contrasted with a corresponding value of 0.71 in Schlosser et al. (2003) who used only resting
inspiratory rates.
This document is a draft for review purposes only and does not constitute Agency policy.
5-123 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
0
0.4 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.48 0.49
human exposure cone (ppm)
Figure 5-15. MLE and upper bound (UB) added risk of SCC in the human
nose for two BBDR models.
Note: Airborne exposure concentrations cIhuman corresponding to a given MLE
and upper bound lifetime added risk levels are the human BMDHuman and
BMDLruman, respectively.
b. Express this dose-response relationship in terms of average flux over the entire human
nasal lining.
c. Next, express this dose-response relationship, calculated here for the entire nose, as risk
per nasal cell versus average flux.
d. Now, if the respiratory and transitional cell types in the human lung and nose are equally
susceptible to formaldehyde-induced cancer risk (as is also assumed in Conolly et al.
[2004]), then it appears reasonable to assume that MLE risk per cell at a given value of
formaldehyde flux is the same in the lung as in the nose.
e. The number of cells and the average flux in a given flux bin in the lung are known
(Overton et al., 2001). Thus, at a given air concentration, the MLE risk due to cells in
the various flux bins of the lung is obtained.
f. One important feature of Overton et al. (2001) was that their flux bins mapped
physically with lung depth. Therefore, in addition to extrapolating risk to the entire
human lung, it was also relatively easy to calculate risk as a function of airway
generation in the lung (corresponding to different lung depths).
This document is a draft for review purposes only and does not constitute Agency policy.
5-124 DRAFT—DO NOT CITE OR QUOTE
-------
1 g. Because of high formaldehyde reactivity and solubility at the POE, the MLE value risk
2 to the lower respiratory tract (as determined above in steps 1-5) was a small fraction of
3 risk to the upper respiratory tract. Therefore, it sufficed to assume that the relative
4 increase in upper bound risk for the combined upper respiratory tract + lower respiratory
5 tract compared to that for only the upper respiratory tract would be the same as the
6 corresponding relative increase in the value of the MLE risk. The upper bound risk to
7 the entire respiratory tract and consequently the BMDL value corresponding to a given
8 response were thus determined.
9
10 These calculations indicated that including the risk of SCC in the lower respiratory
11 tract resulted in at most a 3% increase in the added risk at the lower end of the human
12 exposure range in Figure 5-16 (i.e., at 0.42 ppm) and about a 1.5% increase at the higher end
13 of the range in that plot. Therefore, because of the steepness in the dose-response curve in
14 this exposure range and much lower risk in the lung at any exposure concentration,
15 including the lower respiratory tract did not appreciably alter the human BMDs and BMDLs
16 at the 0.5 and 1% response levels.
17 Unit risks of SCC in the human respiratory tract extrapolated in this manner are reported
18 in Table 5-23.
19
20 Table 5-23. BMD modeling of unit risk of SCC in the human respiratory
21 tract
22
Extra risk level
Benchmark levels (ppm)
Unit risk3
(per ppm)
BMD
BMDL
0.005
0.415-0.450
0.410-0.435
1.2 x 10~2
0.010
0.430-0.460
2.2 x 10~2
23
24 "Obtained from the mean of the two BMDLs.
25
26 Note: Findings are based on nasal tumors in rats and formaldehyde flux to tissue as dosimeter, using dose-
27 response curves for the F344 rat predicted by clonal growth modeling. Two chronic bioassays (Monticello
28 et al., 1996; Kerns et al., 1983) were combined, and control animals from the historical NTP inhalation
29 bioassays were added to the control animals in these bioassays.
30
31
32 5.3.4.2. Comparison with Other Benchmark Dose Modeling Efforts
33 The CUT assessment (Schlosser et al., 2003; CUT, 1999) also presented, as their less
34 preferred option, a benchmark approach on the data set obtained by combining two chronic
35 bioassays with similar protocols (Monticello et al., 1996; Kerns et al., 1983) along with data
This document is a draft for review purposes only and does not constitute Agency policy.
5-125 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
from 94 animals that had not been previously examined. These authors used two measures of
response—tumor incidence and cell proliferation—and, in each case, they used two
dosimeters—DPX and formaldehyde flux to the nasal lining.
The extrapolation to human was carried out by using a hybrid CFD and pharmacokinetic
model. The CFD model (Kimbell et al., 2001a, b; Kepler et al., 1998; Subramaniam et al., 1998)
enabled calculation of site-specific flux in the nose of the rat, monkey, and human species for
inhaled formaldehyde concentrations, and the PBPK model (Conolly et al., 2000) linked this flux
to predicted DPX levels. The models were constructed for anatomically realistic representations
of a single individual in each species. The CFD and PBPK modeling and uncertainties in these
estimates have been reviewed in the Modeling the Toxicokinetics of Formaldehyde and DPX
section of chapter 3.
5.3.4.2.1. Benchmark dose using administered concentration. Schlosser et al. (2003) fit
multistage, Weibull, polynomial, and log-probit quantal models to the tumor data and exercised
the models (except the log-probit) with and without requiring that the fits pass through the
origin. The log-probit fit passed through the origin (see Figure 5-16). A fifth degree polynomial
was used in the multistage model. The best fit was obtained with the polynomial and Weibull
models for the tumor incidence data with a non-zero intercept (threshold) on the dose axis. Fits
passing through the origin did not pass the statistical goodness-of-fit criteria (p > 0.01) for
models other than the log-probit. The dose response near the lowest dose was fairly steep, with
the LEDioS and LEDoiS nearly the same for each model, at least to one significant figure. In
terms of administered concentration, the LEDs ranged from 3.8 to 6.4 ppm.
5.3.4.2.2. Benchmark dose derived with internal dose (flux and DPX) as dose metrics in
Schlosser et al. (2003).
Schlosser et al. (2003) used CFD simulations (Kimbell et al., 2001a, b) of mass flux of
formaldehyde delivered across the nasal lining. The dose metric used by Schlosser et al. (2003)
for the extrapolation was the average flux of formaldehyde, expressed in pmol/cm2-minute, to
the entire surface of the airway lining (excluding tissue lined by non-mucus-coated squamous
tissue, which was considered not to absorb formaldehyde).
This document is a draft for review purposes only and does not constitute Agency policy.
5-126 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Probit Model with 0.95 Confidence Level
0.8
Probit
./i
0.6
0.4
0.2
1j, j _—ir
BMD.U bMD.
0
0 2 4 6 8 10 12 14 16
dose
15:19 02/04 2004
Figure 5-16. Replot of log-probit fit of the combined Kerns et al. (1983) and
Monticello et al. (1996) data on tumor incidence showing BMCio and
BMCL10.
Source: Adapted from Schlosser et al. (2003).
In the CFD model, flux in any region is linearly related to the airborne exposure
concentration (i.e., flux = f x Cair [ppm], where f is a constant of proportionality and Cair is the
exposure concentration). The ratio of f (rat)/f (human) was determined as given by eq 5-4. This
ratio was equal to 0.71 and differed from the value of 0.46 used in this document (as presented in
eq 4-5) because Schlosser et al. (2003) used resting inspiratory rates. In the next level of
dosimetric complexity, Schlosser et al. (2003) used DPX as the relevant dosimeter based on
values predicted by PBPK models developed by Conolly et al. (2000). This expressed the local
dose as pmol of formaldehyde equivalents covalently bound to DNA per unit volume of nasal
tissue. Human CFD and PBPK models were exercised to determine the airborne concentration
of formaldehyde that yields average DPX levels equal to those in the rat at the BMC. This
airborne concentration was then the HEC. The human benchmark extrapolations in Schlosser et
al. (2003) using flux and DPX are shown in Table 5-25, located at the end of section 5.4.
The assumption in using DPX data was that lifetime exposure to the same DPX
concentration for a given duration each day leads to equivalent risk across species. Table 5-25
shows their human benchmark calculations for a continuous environmental exposure. These
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
5-127 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
were exposures that resulted in the same steady-state DPX concentrations as the weekly TWA
DPX values in rats at the rat benchmark exposure concentrations.
5.3.4.2.3. Cell proliferation in CUT benchmark modeling.
Schlosser et al. (2003) also used cell proliferation as representing the adverse response,
and the BMDs calculated with these data did not differ appreciably from their other benchmark
estimates. The use of cell proliferation as an end point is considered to have the advantage that it
represents an early step contributing to carcinogenesis. In this document, a BMD is not
calculated based solely on cell replication as a response. Instead, cell replication rates are used
as input to the clonal growth model and a BMD based on a fit to the tumor response using that
model is considered a better choice since it integrates cell replication along with other relevant
data, such as the number of cells at risk and DPXs.
5.3.4.3. Kaplan-Meier Adjustment
In the simplest consideration of the impact of competing risks on the nasal tumor
incidence, tumor incidences were adjusted for early deaths according to Kaplan-Meier (KM)
survival estimates (KS Crump Group, 2001). This procedure allows for the possibility that some
tumors may otherwise have developed in the animals that died early due to other causes. All the
animals in the study were considered except those that were kept past termination of exposure.
A comparison of the adjusted incidence data is presented below in Table 5-24. While the
adjustments have been provided in Table 5-24, it needs to be noted that the data allow for a full
time-to-tumor analysis as presented below.
Table 5-24. Formaldehyde-induced rat tumor incidences
Observed tumors/
Exposure level (ppm)
KM adjusted incidence
number at risk"
0.0
0.0
0/242
0.7
0.0
0/70
2.0
0.0
0/254
6.0
0.02
3/120a
10.0
0.61
22/3 6a
15.0
0.83
157/190a
aKM adjusted. Numbers not indicated by footnote were not amenable to KM
adjustment because there were no tumors; these numbers at risk reflect all animals
surviving 1 year on study.
Source: Monticello et al. (1996); Kerns et al. (1983).
This document is a draft for review purposes only and does not constitute Agency policy.
5-128 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
5.3.4.4. EPA Time-to-Tumor Statistical Modeling
Instead of using the KM adjustment, EPA has used the multistage Weibull time-to-tumor
model (Portier et al., 1986; Krewski et al., 1983) in other assessments (e.g., ethylene oxide,
1,3-butadiene, chloroprene). This is a dose-response model that includes the exact time of
observation of the tumors and therefore gives appropriate weight to the amount of time each
animal was on study without a tumor and acknowledges earlier tumor incidence with increasing
dose level. The data used in this analysis were obtained from the appendix in Conolly et al.
(2003) with one crucial modification. These data combined the nasal squamous carcinoma data
of Kerns et al. (1983) and Monticello et al. (1996) along with results from an additional
94 animals not previously examined in the Monticello et al. (1996) study. Animals in some
exposure groups were held up to 6 months following the 24-month exposure period; these
animals were deleted from the analysis for the following reason: there were no tumors among
these animals, and inclusion of them would have required estimating an equivalent TWA
exposure over the entire study period for these animals (40 in 2 ppm group, 39 in 6 ppm group, 3
in 15 ppm group), whereas the other animals would be represented by their actual exposure
concentrations.
Due to earlier tumor occurrence with increasing exposure level and increased mortality
with increasing exposure level, methods that can reflect the influence of competing risks and
intercurrent mortality on site-specific tumor incidence rates are preferred. EPA has generally
used the multistage Weibull model because it incorporates the time at which death with tumor
occurred, giving appropriate weight to the amount of time each animal was on study without a
tumor; the model has the following form: P(d) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk) x (t -
t0)z], whereP(d) represents the lifetime risk (probability) of cancer at dose d (i.e., human
equivalent exposure in this case); parameters q, > 0, for i = 0, 1, ..., k; l is the time at which the
tumor was observed; and z is a parameter estimated in fitting the model, which characterizes the
change in response with age. The parameter t0 represents the time between when a potentially
fatal tumor becomes observable and when it causes death.
A further consideration is the distinction between tumor types as being either fatal or
incidental in order to adjust for competing risks. Incidental tumors are those tumors thought not
to have caused the death of an animal (such as those observed during interim or terminal
sacrifices), while fatal tumors are thought to have resulted in animal death. For these data, nasal
tumors observed with early deaths were considered to be fatal.
The dose-response analyses (Figures 5-17, 5-18, 5-19) were conducted by using the
computer software program TOXRISK, version 5.3 (ICF, Fairfax, VA), which is based on
Weibull models drawn from Krewski et al. (1983). Parameters were estimated by using the
This document is a draft for review purposes only and does not constitute Agency policy.
5-129 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
method of MLE. Specific multistage Weibull models were selected for the individual tumor
types for each sex, based on the values of the log likelihoods according to the strategy used by
EPA (U.S. EPA, 2002b). If twice the difference in log-likelihoods was less than a %2 with
degrees of freedom equal to the difference in the number of stages included in the models being
compared, the models were considered comparable, and the most parsimonious model (i.e., the
lowest-stage model) was selected contingent on visual fits of the data as follows. For incidental
tumors, plots of model fits compared with Hoel-Walburg estimates of cumulative incidence were
also examined for goodness of fit in the lower exposure region of the observed data (Gart et al.,
1986) (Figure 5-18). For fatal tumors, plots of model fits were compared with KM estimates of
cumulative incidence. If a model with one more stage fitted the low-dose data better than the
most parsimonious model, then the model with one higher stage was selected.
Due to the sharp increase in responses between 6 and 10 ppm, no adequate fit was
achieved. Data for the highest dose were dropped in an effort to focus the fitting process for this
empirical model on the low-dose region. The model that then provided the best overall fit
included five stages but with coefficients for the lower stages estimated to be zero (see
Figures 5-17, 5-18, 5-19). The parameter t0 was estimated to be zero, consistent with rapidly
fatal tumors. On the other hand, an alternate run treating all tumors as incidental to the death of
the affected animals yielded BMCLs and BMCs within 10% of these estimates (Figure 5-18);
thus, tumor context is not a sensitive consideration for these data.
For the same reasons as discussed in section 5.3.3 (the concluding discussion of the
BBDR modeling), a linear low-dose extrapolation approach was used to estimate human
carcinogenic risk associated with formaldehyde exposure. PODs for estimating low-dose risk
were identified at doses at the lower end of the observed data, corresponding to 1% extra risk,
defined as the extra risk over the background tumor rate [P(d) - P(0)]/[1 - P(0)]. PODs
corresponding to 10% extra risk are also provided to facilitate comparison with other chemicals.
Rat benchmark levels obtained by analysis of the tumor data are shown in Table 5-25. PODs
were converted to continuous human-equivalent exposure levels by multiplying by
(5 days/7 days) x (6 hours/24 hours), or 0.178, and by multiplying by the ratio of fluxes
developed in section 5.3.6.1.3. The lifetime continuous inhalation unit risk for humans is
defined as the slope of the line from the lower 95% bound on the exposure at the POD,
calculated by dividing the BMR level (1%) by the corresponding BMCLoi. This 95% UCL
represents a plausible upper bound on the true risk.
This document is a draft for review purposes only and does not constitute Agency policy.
5-130 DRAFT—DO NOT CITE OR QUOTE
-------
0.9
0.8
0.7
¦E 0.6
T3
C
o
&05
Q)
o 0.4
r
o
£ 0.3
£
0.2
0.1
0
0 2 4 6 8 10 12 14 16
Formaldehyde exposure (ppm)
1
2 Figure 5-17. EPA Multistage Weibull modeling: nasal tumor dose response.
3
4 Note: Time-to-tumor modeling of Kerns et al. (1983) and Monticello et al. (1996)
5 data compared with incidences adjusted by using KM estimates evaluated at
6 104 weeks.
7
8 Source: Adapted from Schlosser et al. (2003).
9
~ Kaplan-Meier adjusted
incidence, from Schlosser
et al. (2003)
Multistage-Weibull estimate
at 104 weeks, omitting high
dose from fit
This document is a draft for re\'iew purposes only and does not constitute Agency policy.
5-131 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
15:36 02/26/2004
a.
Incidental Graph
hcho5.ttd - nasal squamous cell carcinomas
Model: Five Stage Weib
Dose (ppm)=0
Dose (ppm)=2
Dose (ppm)=6
Dose (ppm)=10
Hoel Walburg (6)
Hoel Walburg (10)
60
Time (wks)
Figure 5-18. Multistage Weibull model fit.
Note: Data of Kerns et al. (1983) and Monticello et al. (1996) compared with
Hoel-Walburg estimates of tumor incidences occurring at interim and terminal
sacrifices.
15:36 02/26/2004
Fatal Graph
hcho5.ttd - nasal squamous cell carcinomas
Model: Five Stage Weib
1
0.8
0.6
0.4
0.2
Dose (ppm)=0 7
Dose (ppm)=2
Dose (ppm)=6
Dose (ppm)=10
Kaplan Meier (6)
Kaplan Meier (10)
20
40
60
Time (wks)
80
100
Figure 5-19. Multistage Weibull model fit of tumor incidence data compared
with KM estimates of spontaneous tumor incidence.
Source: Developed from data reported in Kerns et al. (1983) and Monticello et al.
(1996).
This document is a draft for review purposes only and does not constitute Agency policy.
5-132 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
The extrapolation to humans in terms of using formaldehyde flux to tissue as the dose
metric is shown in Table 5-25, where unit risk in terms of ql*, the statistical upper bound on the
coefficient of the term linear in dose in the multistage model, is also presented, ql* is presented
even though this is no longer done, as per current EPA practice (see section 5.3.6 for discussion).
These results are to be compared with the preferred benchmark estimates obtained in
Table 5-23 by using the results of biologically based models. In summary, the unit risks
obtained by various methods, including the results in Schlosser et al. (2003), fall within a rather
tight range. In particular, ql * was obtained to within a factor of two of other values even though
ql itself was zero. The large difference between ql and ql* aptly reflects the large uncertainty
in the low-dose response.
5.4. CONCLUSIONS FROM I II I QUANTITATIVE ASSESSMENT OF CANCER RISK
FROM FORMALDEHYDE EXPOSURE BY INHALATION
5.4.1. Inhalation Unit Risk Estimates Based on Human Data
As described in section 5.2, a (plausible upper bound) lifetime extra cancer unit risk of
1.1 x 10 2 per ppm (8.8 x 10 6 per |ig/m3) of continuous formaldehyde exposure was estimated
for NPC incidence using the log-linear modeling results (for NPC mortality from cumulative
exposure) from a high-quality occupational epidemiologic study in a life-table analysis to obtain
a POD and then applying linear low-dose extrapolation from the POD. Using similar methods
and data from the same study for Hodgkin lymphoma and leukemia mortality from cumulative
formaldehyde exposure, (plausible upper bound) lifetime extra cancer risk estimates of 1.7 x 10 2
per ppm (1.4 x 10~5 per |ig/m3) for Hodgkin lymphoma incidence and 5.7 x 10~2 per ppm
(4.6 x 10 5 per |ig/m3) for leukemia incidence were derived. Sources of uncertainty in these
estimates are discussed in sections 5.2.2.4 and 5.2.3.4. For the incidence risk for these three
cancer types combined, a total (upper bound) cancer unit risk estimate of 8.1 x 10 2 per ppm
(6.6 x 10~5 per |ig/m3) was obtained (section 5.2.4).
5.4.2. Inhalation Unit Risk Estimates Based on Rodent Data
As described in section 5.3, the unit risk derived for SCC in the upper and lower
respiratory tract (combined) based on linear extrapolation from PODs from several plausible
models, including purely statistical modeling (nose only, quantal and time-to-tumor modeling)
and biologically based modeling (entire respiratory tract), resulted in a narrow range of
1.2 x 10 2 to 2.2 x 10 2 per ppm. Risk to the lower respiratory tract was numerically
insignificant compared to the nasal cancer risk.
This document is a draft for review purposes only and does not constitute Agency policy.
5-133 DRAFT—DO NOT CITE OR QUOTE
-------
a 2
o S.
a ^3
Grj St
li
>!
5^
W
§•
rs
sj
3
5
>!
6
s ^
a ^
>!
Table 5-25. Human benchmark extrapolations of nasal tumors in rats by using formaldehyde flux and DPX
Model
Source
Rat benchmark levels
(ppm)
Extrapolated human benchmark levels (ppm)
Unit risk" (ppm) 1
1%
5%
10%
Dose metricb
1%
5%
10%
1%
5%
10%
Weibulfd
(with
threshold)
Schlosser et
al. (2003)
ED
LED
5.91
5.58
6.12
5.94
6.40
6.22
Fluxe
ED
0.75
0.78
0.82
LED
0.71
0.76
0.79
1.4 xl0~2
6.6 x 10~2
1.3 x 10"1
DPXf
ED
0.76
0.79
0.84
LED
0.71
0.76
0.81
1.4 xl0~2
6.6 x 10~2
1.2 x 10"1
Multistage
Weibull (time-
to-tumor)°'d'8
EPA (this
assessment)
ED
LED
4.28
3.57
5.93
5.52
6.84
6.41
Fluxh
ED
0.35
0.49
0.57
LED
0.30
0.46
0.53
3.4 xKT2
1.1 x KT1
1.9 x 10"1
ql* = 2.2 x 10~2
BBDR models
(Table 5-23)
EPA (this
assessment)
See Table 5-23 and associated text
at 1%: 2.2 xl0~2
at 0.5%: 1.2 x 10~2
Crq o
§ 4
<3 ^
O
Note 1: Combined tumor incidence data from Kerns et al. (1983) and Monticello et al. (1996) were used for response.
"Slope of straight line extrapolation from the POD of the dose-response curve at the 1, 5, and 10% extra risk level.
''Flux: CFD modeling. DPX: CFD + PBPK modeling.
cp Value for Weibull model fit = 0.90. For the time-to-tumor modeling, goodness-of-fitp value was not provided by software package; therefore, fit was
judged by comparing fitted curve to KM survival estimates (see Figure 5-19).
dFor Weibull model, Schlosser et al. (2003) obtained best fit with a positive intercept on dose axis. For multistage Weibull model, curves pass through origin.
eHuman benchmark levels extrapolated using flux were multiplied by fHCHoWfHCHo-Human (=0.71) for interspecies extrapolation and multiplied by (6/24) x
(5/7) to adjust for continuous exposure.
fHuman benchmark levels using DPX were continuous environmental exposures that would result in steady-state DPX levels in humans equal to the weekly
TWA DPX levels in rats at the rat BMCs for 6 hours/day and 5 days/week.
£ 8P(d,t) = 1 - exp[-(q0 + qid + q2d2 + ... + qkdk)* tz]. q0, qi, q2, q3, q4 were all taken to be zero. q5 = 2.9 x 10 22. z = 8.1.
hHuman benchmark levels extrapolated using flux were multiplied by fHCHo WfHCHo-Human = 0.46 for interspecies extrapolation and multiplied by (6/24) x (5/7)
to adjust for continuous exposure (see section 5.3.6.2).
O
o
o
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
5.4.3. Summary of Inhalation Unit Risk Estimates
The epidemiologic and rodent inhalation data indicate multiple sites of concern. Unit risk
estimates calculated separately from these data are presented in Table 5-26.
As can be seen in the summary table (Table 5-26), the unit risk estimate based on human
data for NPC is in the range of the estimates calculated for respiratory tract cancer from the
rodent nasal cancer data. The unit risk estimate for Hodgkin lymphoma is also in the same
range, while the unit risk estimate for leukemia and the total cancer unit risk estimate are up to
fourfold higher.
Table 5-26. Summary of inhalation unit risk estimates
Cancer type3
Dose metric
Unit risk estimate
(ppm1)
Based on epidemiologic data
Nasopharyngeal
Cumulative exposure
0.011
Hodgkin lymphoma
Cumulative exposure
0.017
Leukemia
Cumulative exposure
0.057
Total cancer riskb
Cumulative exposure
0.081
Based on experimental animal data
SCC of the respiratory
tract
Local dose (flux) of
formaldehyde in pmol/mm2-
hour
0.011-0.022
aThe unit risk estimates are all for cancer incidence.
bThe total cancer unit risk estimate is an estimate of the upper bound on the sum of risk estimates calculated
for the 3 individual cancer types (nasopharyngeal cancer, Hodgkin lymphoma, and leukemia); it is not the
sum of the individual (upper bound) unit risk estimates (see Section 5.2.4).
As noted in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), when
high-quality human data are available, they are generally preferred over laboratory animal data
for quantitative risk assessment. Thus, the preferred (plausible upper bound) unit risk estimate in
this assessment is the value of 8.1 x io~2 per ppm (6.6 x 10~5 per |ig/m3) based on human data
for NPC, Hodgkin lymphoma, and leukemia.
As documented in section 4.5, formaldehyde is a mutagenic carcinogen and the weight of
evidence suggests that formaldehyde carcinogenicity can be attributed, at least in part, to a
mutagenic MOA. Therefore, since there are no adequate chemical-specific data to evaluate the
This document is a draft for review purposes only and does not constitute Agency policy.
5-135 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
susceptibilities of different life stages by the inhalation route of exposure14, increased early-life
susceptibility should be assumed, and, if there is early-life exposure, the ADAFs should be
applied, in accordance with EPA's Supplemental Guidance for Assessing Susceptibility from
Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). See section 5.4.4 below for more
details on the application of the ADAFs.
5.4.4. Application of Age-Dependent Adjustment Factors (ADAFs)
When there is sufficient weight of evidence to conclude that a mutagenic MOA is
operative in a chemical's carcinogenicity and there are inadequate chemical-specific data to
assess age-specific susceptibility, as is the case for formaldehyde (by inhalation exposure; see
section 5.4.3), U.S. EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b) recommends the application of default ADAFs to
adjust for potential increased susceptibility from early-life exposure (see U.S. EPA [2005b] for
detailed information on the general application of these adjustment factors). In brief, U.S. EPA
(2005b) establishes ADAFs for three specific age groups: 10 (for <2 years), 3 (for 2 to
<16 years), and 1 (for 16 years and above). For risk assessments based on specific exposure
assessments, the 10-fold and threefold adjustments to the unit risk estimates are to be combined
with age-specific exposure estimates when estimating cancer risks from early-life (<16 years
age) exposure. The ADAFs and their age groups may be revised over time. The most current
information on the application of ADAFs for cancer risk assessment can be found at
www.epa.gov/cancerguidelines.
For inhalation exposures, assuming ppm equivalence across age groups (i.e., equivalent
risk from equivalent exposure levels, independent of body size) and using the preferred unit risk
estimate of 6.6 x 10 5 per |ig/m3 from section 5.4.3, the calculation is fairly straightforward. The
ADAF-adjusted lifetime total cancer unit risk estimate is calculated as shown in Table 5-27:
14 The oral exposure bioassay of Soffritti et al. (1989) provides evidence of increased early-life susceptibility for
carcinogenicity at the portal of entry (i.e., gastrointestinal tract cancers), but it is unclear how to extrapolate the
increased susceptibility quantitatively to portal-of-entry cancers from inhalation exposures. There was no apparent
increased early-life susceptibility for hemolymphoreticular cancers; however, there are unresolved discrepancies
between the Soffritti et al. (1989) and the Soffritti et al. (2002) reportings of the hemolymphoreticular cancer results
for the adult-only exposure component of the study which make interpretation of all of the hemolymphoreticular
cancer results from the Soffritti et al. (1989) paper uncertain (see Section 4.5).
This document is a draft for review purposes only and does not constitute Agency policy.
5-136 DRAFT—DO NOT CITE OR QUOTE
-------
1 Table 5-27. Total cancer risk from exposure to a constant formaldehyde
2 exposure level of 1 jig/m3 from ages 0-70 years
3
Age group
ADAF
Unit risk
(per jig/m3)
Exposure
concentratio
n (jig/m3)
Duration
adjustment
Partial risk
0 to < 2
years
10
6.6 x 10~5
1
2 years/70 years
1.9 x 10~5
2 to < 16
years
3
6.6 x 10~5
1
14 years/70
years
4.0 x 10~5
>16 years
1
6.6 x 10~5
1
54 years/70
years
5.1 x 1(T5
Total risk = 1
1.1 x 10~4
4
5 (Note that the partial risk for each age group is the product of the values in columns 2-5 [e.g.,
6 10 x (6.6 x 10 ') x l x 2/70 = 1.9 x 10 5], and the total risk is the sum of the partial risks. This 70-year risk
7 estimate for a constant exposure of 1 (ig/m3 is equivalent to a lifetime unit risk of
8 1.1 x 10 4 per |ig/nr\ adjusted for early-life susceptibility, assuming a 70-year lifetime and constant exposure
9 across age groups.)
10
11
12 In addition to the uncertainties discussed above for the inhalation unit risk estimate, there
13 are uncertainties in the application of ADAFs to adjust for potential increased early-life
14 susceptibility. The ADAFs are general default factors, and it is uncertain to what extent they
15 reflect increased early-life susceptibility for exposure to formaldehyde, if, in fact, early-life
16 susceptibility is increased as assumed. To some extent, the unit risk estimates for Hodgkin
17 lymphoma and leukemia already reflect some partial risk from early-life exposure because the
18 life-table programs include background rates for childhood cancers. However, the impact of this
19 partial risk is negligible compared to the effect of the ADAFs on the final risk estimate. For
20 example, eliminating the background rates up to age 16 from the life-table programs decreases
21 the lifetime extra risks at the PODs by about 0.5% for leukemia and about 1.2% for Hodgkin
22 lymphoma. The ADAFs, on the other hand, increased the lifetime unit risk estimate by about
23 66%.
24
25 5.4.5 Conclusions: Cancer Inhalation Unit Risk Estimates
26 As presented in section 5.4.3, the preferred (plausible upper bound) cancer unit risk
27 estimate for formaldehyde exposure in this assessment is the total cancer risk estimate of
28 8.1 x 10"2 per ppm (6.6 x 10~5 per jig/m3) based on (adult) human data for NPC, Hodgkin
29 lymphoma, and leukemia.
This document is a draft for review purposes only and does not constitute Agency policy.
5-137 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
In addition, as described in section 5.4.4, because the weight of evidence suggests that
formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic MOA and there
are inadequate chemical-specific data to assess age-specific susceptibility, increased early-life
susceptibility should be assumed and, if there is early-life exposure, ADAFs should be applied,
in accordance with EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b). Consequently, applying the ADAFs to the
preferred unit risk estimate to obtain a full lifetime unit risk estimate yields
0.081/ppm x [(10 x 2 years/70 years) + (3 x 14/70) + (1 x 54/70)]
= 0.13/ppm = 1.1 x 10~4/(ng/m3)
Using the above full lifetime unit risk estimate of 0.13 per ppm, the lifetime chronic
exposure level of formaldehyde corresponding to an increased cancer risk of 10 6 can be
estimated as follows: (10~6)/(0.13/ppm) = 7.7 x 10-6 ppm = 0.008 ppb = 0.009 [^g/m3. Similarly,
the lifetime chronic exposure level of formaldehyde corresponding to an increased cancer risk of
10 4 is 0.8 ppb, or 0.9 (J,g/m3. (Note that for less-than-lifetime exposures scenarios [or for
exposures that vary with age], the adult-based combined estimate of 0.081 per ppm should be
used, but if there is early-life exposure, the ADAFs should be applied in accordance with EPA's
Supplemental Guidance [see section 5.4.4]).
This document is a draft for review purposes only and does not constitute Agency policy.
5-138 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
6. MAJOR CONCLUSIONS IN Till CHARACTERIZATION OF HAZARD AND
DOSE-RESPONSE
6.1 SUMMARY OF HUMAN HAZARD POTENTIAL
6.1.1 Exposure
Formaldehyde (CH20) occurs as a gas at room temperature. It is highly reactive and
dissolves readily in water. Formaldehyde is present in a wide variety of products including
plywood adhesives, abrasive materials, insulation, insecticides and embalming fluids (IPCS,
2002a; Agency for Toxic Substances and Disease Registry [ATSDR], 1999). The major sources
of anthropogenic emissions of formaldehyde are motor vehicle exhaust, power plants,
manufacturing plants that produce or use formaldehydes or substances that contain formaldehyde
(i.e., glues), petroleum refineries, coking operations, incineration, wood burning, and tobacco
smoke (INEG, 2003). Reported outdoor air concentrations of formaldehyde in urban and
suburban areas are near 3 |ig/m3 (~ 3 ppb) (U.S. EPA, 2008) and indoor residential levels are
approximately 10 times higher (Health Canada and Environment Canada, 2001).
Limited U.S. data indicate that the upper end of the formaldehyde concentration range in
drinking water is approximately 10 |ig/L IPCS, 2002a). Formaldehyde is a natural component of
a variety of foodstuffs (IARC, 1995; IPCS, 1989). In addition, foods may also be contaminated
with formaldehyde as a result of fumigation (e.g., grain, seeds), cooking (as a combustion
product), and release from formaldehyde resin-based tableware (IARC, 1995). Limited data
measuring formaldehyde in food indicates that the concentration range is <0.03-14 mg/kg
Health Canada and Environment Canada, 2001). Daily intake of formaldehyde has been
estimated to be between 1.5 and 14 mg/day for an average adult (IPCS, 1989; Fishbein, 1992).
6.1.2 Absorption, Distribution, Metabolism, and Excretion
In water, less than 0.1% of formaldehyde exists unhydrated, with the majority reported to
be in the hydrated form, methylene glycol (CH2(OH)2) (Priha et al., 1996). Formaldehyde is a
reactive molecule that can react with both low molecular weight cellular components (e.g., GSH)
as well as high molecular weight components. It is also a well-known cross-linking agent.
Further, formaldehyde is a product of normal cellular metabolic processes. Endogenous
formaldehyde is a constituent of the one-carbon pool. It is thought that most endogenous
formaldehyde exists in a form that is reversibly bound to nucleophiles (Heck et al., 1990; 1982)
and that the formaldehyde hemithioacetal adduct formed with glutathione could account for 50-
80% of the total formaldehyde normally present in cells (Heck et al., 1982).
This document is a draft for review purposes only and does not constitute Agency policy.
6-1 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Inhaled formaldehyde is efficiently absorbed ("scrubbed") in the upper respiratory tract.
The fraction that is absorbed was determined to be approximately 97% in rats (Morgan et al.,
1986), and 85% and 90% respectively in computer simulations of one rhesus monkey and human
at rest (Kepler et al., 1998; Kimbell et al., 2001). As the inspiratory rate increased,
formaldehyde decreased to about 70% during light exercise and to 58% during heavy exercise
conditions in the human. During heavy exercise, the absorption of formaldehyde in the first six
to eight generations of the tracheobronchial airways is estimated to be comparable to that in the
nasal region (Overton et al., 2001).
Airway geometry is an important determinant of inhaled-formaldehyde dosimetry in the
respiratory tract. There are large differences across species in the anatomy of the upper
respiratory tract and in airflow patterns. Using computer simulation, the regional uptake patterns
of formaldehyde in the upper respiratory tract are observed to be spatially non-homogenous and
to exhibit strong species differences. Airflow patterns are also significantly different as
breathing patterns and activity profiles change, depending on whether breathing is oral or nasal.
The overall information on the disposition of inhaled formaldehyde comes from many
studies using different experimental methods including: [14C] radiolabeling, gas
chromatography-mass spectroscopy (GC-MS), dual isotope labeling (3H, 14C) and high-
performance liquid chromatography (HPLC) studies. In a study of rats following exposure to
radiolabeled formaldehyde, the radioactivity was very high in the nasal mucosa but was also
extensively distributed to various tissues including the bone marrow (Heck et al., 1983). The
elevated 14C in various tissues was thought unlikely to be due to free formaldehyde but instead to
arise from either rapid metabolic incorporation or formation of covalent adducts or incorporation
via carboxylation reactions of the 14C02 formed during metabolism (Heck et al., 1983;
Casanova-Schmitz et al., 1984). Studies using the GC-MS method indicate that exposure to
formaldehyde over a wide range of exposure concentrations and durations does not result in
elevated levels in blood, above those of endogenous formaldehyde levels in rats, rhesus monkeys
and humans (Heck et al., 1985; Casanova et al., 1998). These GC-MS measurements are
consistent with the conclusions that formaldehyde does not appreciably reach the blood, is
rapidly metabolized, interacts with macromolecules when it escapes metabolism, or is otherwise
undetected.
In further studies on the disposition of inhaled formaldehyde, Casanova-Schmitz et al.
(1984) and Casanova-Schmitz and Heck (1983) used dual-isotope labeling of inhaled
formaldehyde as an approach to distinguish between formaldehyde adduct formation and
metabolic incorporation. These were followed by more sensitive experiments using HPLC
measurements in rats and rhesus monkeys exposed to radiolabeled formaldehyde (Casanova et
This document is a draft for review purposes only and does not constitute Agency policy.
6-2 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
al. 1989, 1991). Results from this sets of experiments found that labeling in the nasal mucosa
was due to both covalent binding and metabolic incorporation and labeling of bone marrow
macromolecules was found to be entirely due to metabolic incorporation. Overall, Heck,
Casanova-Schmitz, and their coworkers interpreted the results of these experiments to indicate
that inhaled formaldehyde does not reach distant sites (beyond the portal of entry) at detectable
levels.
Formaldehyde is primarily metabolized by glutathione-dependent formaldehyde
dehydrogenase. In humans this enzyme is referred to using the protein code of ADH3. The
major factor in the disposition of formaldehyde is metabolic clearance by oxidation to formate,
which is either further metabolized to CO2 and water, incorporated into the one-carbon pool,
and/or eliminated in the urine as a sodium salt.
In radiolabeling studies, Heck et al. (1983) determined that the relative contributions of
various excretion pathways in F344 rats following inhalation exposure to formaldehyde were
independent of exposure concentration. Nearly 40% of inhaled [14C]-formaldehyde appeared to
be eliminated via expiration, presumably as C02, while about 17% and 5% was eliminated in the
urine and feces, respectively. Nearly 40% of inhaled [14C]-formaldehyde remained in the
carcass, presumably due to metabolic incorporation. For exposure via the oral route, absorption
of [14C]-formaldehyde (7 mg/kg) in rats resulted in 40% exhaled (as 14CC>2), 10% excreted in
urine, 1% excreted in feces, and much of the remaining 49% retained within the carcass,
presumably due to metabolic incorporation (IARC, 1995; Buss et al., 1964).
Several human and animal studies have attempted to measure the concentration of
formaldehyde in exhaled breath (see section 3.5.2). None of the human studies investigated
whether there is any correlation between exhaled formaldehyde levels and food intake, life stage,
smoking, or health status. Additionally, they were not designed to distinguish between
exogenous (room air) and endogenous (systemic) formaldehyde in exhaled breath. In order to
discern whether endogenous formaldehyde is excreted into the lungs, human subjects must
breathe formaldehyde-free air. Because subjects were breathing room air, which contained 9-10
ppb formaldehyde in two studies and unspecified concentrations in two other studies, there is no
way of knowing whether there was any endogenous formaldehyde in their exhaled breath. This
assessment identifies a critical research need for further studies on the measurement of exhaled
formaldehyde.
The most informative study, performed by Cap et al. (2008), demonstrated that subjects
breathing room air containing 9.6 ±1.5 ppb formaldehyde exhaled a mean formaldehyde
concentration of 2 ppb. This suggests that a substantial portion of inhaled formaldehyde, which
is highly reactive, was retained in the respiratory tract and not exhaled. It is impossible to tell
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
whether any portion of the 2 ppb in exhaled breath was of endogenous origin. In this and other
human studies, there was no adjustment for an artifact in the analytical method that makes it
impossible to distinguish between formaldehyde and reaction products for 1% of exhaled
methanol and ethanol because they have the same mass to charge ratio {m/z = 31). In fact, the
concentration of methanol and ethanol that is misidentified as formaldehyde exceeds the reported
concentrations of exhaled formaldehyde. Thus, it is highly likely that the actual exhaled
formaldehyde concentration in Cap et al. (2008) was significantly lower than 2 ppb, and that
there was little or no endogenous formaldehyde in the exhaled breath. This would be consistent
with an animal study in which Mashford and Jones (1982) detected no exhaled formaldehyde in
rats injected I.P. with 40 mg/kg [14C]-formaldehyde. In summary, there are insufficient data at
this time to confidently establish a concentration of formaldehyde in exhaled breath that can be
attributed to endogenous sources.
6.1.3 Noncancer Health Effects in Humans and Laboratory Animals
A wide variety of human and animal studies provide evidence for health effects in
response to formaldehyde exposure. Some of these health effects are commonly noted at the
portal of entry, as expected for exposure to a reactive gas. In addition, effects on the nervous
and reproductive systems, developmental effects, and immunomodulation have been reported.
The overall weight of evidence (WOE) of human and animal studies for the hazard potential of
formaldehyde is discussed below, along with information on plausible modes of action (MOAs).
6.1.3.1. Sensory Irritation
Formaldehyde, a chemical irritant, binds to protein receptors of the trigeminal nerve,
triggering a burning and painful sensation in humans. This process is distinct from taste and
smell (Nielsen 1991; Cometto-Muniz and Cain, 1992). The trigeminal nerve, which has three
branches (ophthalmic, maxillary and mandibular), not only acts as an afferent nerve relaying
these sensations to the central nervous system, but also has efferent nerve activity (Stedman's
Medical Dictionary: Meggs, 1993). Stimulation of the trigeminal nerve may result in reflex
responses including lacrimation, coughing, and sneezing. Both the reflex responses as well as
sensations such as burning, pain, and itching of the eyes, nose, and throat are considered adverse.
Formaldehyde-induced eye, nose, and throat irritation has been well documented in a
wide range of epidemiologic studies. Common effects of chemically-induced sensory irritation
include lacrimation, burning of the eyes and nose, rhinitis, burning of the throat, and cough
(Feron et al., 2001). Studies examining these endpoints were either controlled chamber studies
with a defined population (e.g., healthy volunteers or sensitive individuals), worker/student
This document is a draft for review purposes only and does not constitute Agency policy.
6-4 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
studies, or general population studies (e.g., residential). Chamber studies, by design, are acute
studies, although some researchers have investigated the outcomes after repeated exposures.
Occupational, student, and residential exposures are generally of longer duration, although there
is variability in exposure level and duration among subjects. The endpoints for assessing
irritation include self-reporting of symptoms (e.g., pain, burning, itching) and objective measures
of irritation (e.g., eye-blink counts, lacrimation).
Eye irritation is the most sensitive of reported effects in human studies. Two different
short-term chamber studies provide similar 10% BMDLs for eye irritation of 560 ppb and 240
ppb for 3 and 5 hour exposures, respectively (Kulle, 1993; Andersen and Molhave, 1983,
modeled by Arts et al., 2006b). Various occupational studies have noted increased eye irritation
for average exposures ranging from 180 ppb to 690 ppb (Horvath et al., 1988, Alexandersson
and Hedenstiera, 1998; Holmstrom and Wilhelmsson, 1988). The results of residential studies,
where in-home formaldehyde levels are used to document exposure, indicate eye irritation may
increase with increasing exposure from 70 to 200 ppb for these chronic exposure scenarios
(Ritchie and Lehnen, 1987, Hanrahan et al., 1984; Liu et al., 1991.)
When a rodent is exposed to an irritant, the inhaled dose and pattern of deposition can be
profoundly affected by reflex bradypnea, a protective reflex observed in rodents but not in
humans. Reflex bradypnea is manifest as markedly decreased activity or prostration, reduced
metabolism, hypothermia (as much as 5°C), significantly reduced respiratory rate and minute
volume, and altered blood and brain chemistry. Reflex bradypnea can occur when the trigeminal
nerve is exposed to a sufficient concentration of an irritant, such as formaldehyde. Because of
their small size, rodents are able to rapidly lower their metabolism and body temperature and
therefore their oxygen demand. The consequence is that their inhaled dose of an irritating
chemical is dramatically lowered. Reflex bradypnea is quantified as the RD50, which is the
concentration of a chemical that results in a 50% decrease in respiratory rate (Tables 4-7 and
4-8). After the irritant exposure is removed, it can take up to two hours for rodents to fully
recover from the effects of reflex bradypnea. Even though humans do not exhibit reflex
bradypnea, involvement of trigeminal nerve stimulation, which is the mechanism for reflex
bradypnea in rodents, may be relevant to MO As for formaldehyde in other species, such as
primates and humans. For example, trigeminal nerve stimulation has been associated with
sensory irritation in humans, highlighting the relevance of this effect.
6.1.3.2. Respiratory Tract Pathology
Formaldehyde-induced respiratory tract pathology includes inflammation, rhinitis, goblet
cell hyperplasia, metaplastic changes, squamous cell hyperplasia, and impaired mucociliary
This document is a draft for review purposes only and does not constitute Agency policy.
6-5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
transport. Formaldehyde binding to the trigeminal nerve triggers the release of neurogenic
mediators of inflammation resulting in tissue edema, lacrimation, mucus production, and
leukocyte infiltration. Therefore, observed pathological changes may be directly related to
neurogenic inflammation from activation of the trigeminal nerve or result, at least in part, from
formaldehyde-induced cell damage to the mucosal tissue. A series of exposures has also been
positively associated with reduced mucociliary clearance, and the induction of histopathologic
lesions in the nose in both human and animal studies assessing formaldehyde-induced changes in
the nasal mucosa suggest that these changes may be, at least in part, a protective or adaptive
response and that increased mucus flow and metaplastic changes would progress in relation to
the concentration and duration of exposure protecting the underlying tissue (Swenberg et al.,
1983).
In rodent studies, formaldehyde-induced histopathological lesions ranging from
inflammation to ulceration, necrosis, and metaplasia have been frequently reported in nasal
turbinates, maxilloturbinates, and in goblet and microvilli cells (e.g., Bhalla et al., 1991;
Monteiro-Riviere and Popp, 1986; Cassee and Feron, 1994; Ionescu et al., 1978; Schreibner et
al.,1979; Monticello et al., 1989). These effects were observed after a variety of exposure
scenarios (e.g., 10 ppm for 4 hrs (Bhalla et al., 1991), 0.5 or 2 ppm for 6 hrs/day for 1 or 4 days
and 6 or 15 ppm for 6 hrs/day for 1 or 2 days (Monteiro-Riviere and Popp, 1986), 3.6 ppm
intermittently for 3 days (Cassee and Feron, 1994), 3% aerosols of formaldehyde for 3 hrs/day
for 50 days (Ionescu et al., 1978)). The progressive pathology of the nasal passages from
formaldehyde inhalation exposure is dependent on increasing concentration and duration of
exposure, as well as from proximal to distal regions of the nasal cavity. For example, some
lesions may be transient (e.g., low-exposure cell proliferation), while others may have a
maximum response and be irreversible (e.g., rhinitis). The nasal epithelium responds with both
adaptive and adverse epithelial changes. As respiratory epithelium transitions to squamous
metaplasia, the effective tissue dose of formaldehyde increases posterior to these lesions. As
epithelial barriers degrade (e.g., squamous metaplasia, keratinization), formaldehyde penetrates
more deeply into the nasal passages. Therefore, the relationship between concentration and
duration of exposure and health outcomes has been difficult to define and, in fact, may be
different for various health effects. Formaldehyde-related histopathological lesions of the nasal
mucosa have been observed at concentrations as low as 2 ppm for chronic exposure and after a
duration as short as 6 hrs at higher concentrations (e.g., 6 ppm) (Table 4-32, table 4-38).
Similar pathology has been reported for workers exposed to formaldehyde, including loss
of cilia, goblet cell hyperplasia, and cuboidal and squamous cell metaplasia and dysplasia, and
these pathology scores were significantly elevated in workers over controls (Holmstrom and
This document is a draft for review purposes only and does not constitute Agency policy.
6-6 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Wilhelmsson, 1988; Edling et al., 1988; and Boysen et al., 1990). Holmstrom and Wilhelmsson
(1988) reported associations between the mean daily exposure of 240 ppb (8hr TWA) and these
changes. Edling et al. (1988) reported that workers experienced a range of exposures (80-900
ppb), with peak exposures of 4000 ppb. Boysen et al. (1990) provided a range of estimated
exposures from 500 ppb to more than 2000 ppb for workers with elevated mean pathology
scores. One controlled chamber study indicated formaldehyde-induced inflammatory changes
which persisted for 18 hours in adults exposed at 400 ppb for only 2 hours (Pazdrak et al.,
1993)).
Short-term formaldehyde exposure also impairs the function of the mucociliary apparatus
which is a critical defensive barrier for the upper respiratory tract. Numerous laboratory animal
studies have reported impaired mucociliary clearance activity associated with formaldehyde
exposures as low as 500 ppb (Table 4-10). Low-concentration or short-term exposures first lead
to an increased rate of ciliary beat, followed by impaired mucus flow, with slowed rate of ciliary
beat and eventual mucostasis (lack of mucus flow) and ciliastasis (lack of ciliary beat) occurring
at higher doses or longer exposure times. These effects have been shown to be both
concentration- and duration-dependent and to occur within 15 minutes after the initial exposure.
Morgan et al. (1983c) suggested that the initial stimulation of ciliary activity may be a defensive
response to the irritant gas, at which time some penetration of formaldehyde to the underlying
epithelial cells may occur. Later effects of mucostasis and ciliastasis may occur as a result of
formaldehyde-induced glycoprotein cross-links, creating a rigid mucus that effectively stops
mucus flow.
Formaldehyde-induced cell proliferation has been demonstrated in nasal epithelium in
animal studies after a range of exposure conditions (e.g. Swenberg et al., 1986; Cassee and
Feron, 1994; Reuzel et al., 1990; Woutersen et al., 1987) (Table 4-43). Formaldehyde-induced
histopathology and mitogenesis may occur as a direct effect of exposure (Tyihak et al., 2001) or
as a secondary effect resulting from adaptive responses and/or compensatory tissue repair that
can occur after formaldehyde exposure (Swenberg, 1983). In a study of Rhesus monkeys
Monticello et al. (1996) noted that increased cell proliferation was seen in locations with
minimal histological changes in the respiratory tract indicating that cell proliferation may be a
more sensitive predictor of more severe health effects due to formaldehyde exposure. Cellular
proliferative responses may initiate lesion formation. A number of studies illustrate that the
duration of repeated exposures may be an important determinant of cell proliferation rates
(Wilmer et al., 1987; Swenberg et al., 1986). Reduced mucociliary clearance and the induction
of histopathologic lesions in the nose effects have been noted in human formaldehyde studies.
This document is a draft for review purposes only and does not constitute Agency policy.
6-7 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Histopathological lesions and biochemical changes have been reported in the lung
following formaldehyde inhalation exposure in experimental animal studies (Kamata et al.,
1996a; Ionescu et al., 1978) following high exposure levels (128.4 or 294.5 ppm formaldehyde).
6.1.3.3. Effects on Pulmonary Function
The potential of formaldehyde exposure to cause pulmonary functional deficits in
humans has been examined on several time scales. The epidemiologic literature includes studies
of acute exposures among naively exposed anatomy graduate students (Kriebel et al., 1993;
2001), anatomy graduate students with several weeks of episodic exposure (Kriebel et al., 1993),
and post-shift versus pre-shift worker pulmonary function among those with regular
occupational exposure (Malaka and Kodama, 1990; Herbert et al., 1994; Alexandersson et al.,
1982; Alexandersson and Hedenstierna, 1989). Depending on whether the exposures are naive
or not, the epidemiologic studies that assessed the pulmonary effects after acute exposures to
formaldehyde are assessing different biological responses, namely, the acute effect alone or the
acute effect(s) in people who may have already been sensitized to different and unknown
degrees.
The observed effects in the previously unexposed anatomy students provide additional
information on acute exposures in two naive populations (Kriebel et al., 1993; 2001), as well as
insight into the possible intermediate stages of sensitization (Kriebel et al., 1993). Kriebel and
colleagues (1993) examined the pre-laboratory and post-laboratory peak expiratory flow (PEF)
in students attending anatomy classes once a week. They found the strongest pulmonary
response when examining the average cross-laboratory decrement in peak expiratory flow in the
first 2 weeks of the study when formaldehyde concentrations collected in the breathing zones
had a geometric average concentration of 0.73 ppm. Overall, the students exhibited a 2%
decrement in PEF, while the students with any history of asthma showed a 7.3% decrement in
PEF. These findings of acute decreases in PEF following students' initial formaldehyde
exposure were corroborated by the Kriebel et al. (2001) study, using a similar study design
applied to a separate class of anatomy students. Similar findings have been reported for low-
level residential formaldehyde exposure including decreased peak expiratory flow rates (PEFRs)
(Krzyzanowski et al., 1990). Workers chronically exposed to formaldehyde have exhibited signs
of reduced lung function consistent with bronchial constriction, inflammation, or chronic
obstructive lung disease. Lung function deficits have been reported both in pre-shift versus post-
shift measurements andjLS a result of chronic exposures (Malaka and Kodama, 1990; Herbert et
al., 1994; Pourmahabadian et al., 2006, Alexandersson et al., 1982; Alexandersson and
Hedenatiena 1989). Decreases in spirometric values, including vital capacity (VC), forced
This document is a draft for review purposes only and does not constitute Agency policy.
6-8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
expiratory volume (FEV), forced vital capacity (FVC) and FEV/FVC have been reported in
humans. Chronic studies also reported increased respiratory symptoms such as cough, increased
phlegm, asthma, chest tightness and chest colds in exposed workers (Malaka et al., 1990; Herbet
et al., 1994; Pourmahabadian et al., 2006, Alexandersson et al., 1982; Alexandersson and
Hedenatiena 1989). Similar findings have been reported following low-level residential
formaldehyde exposure including decreased PEFRs (Krzyzanowski et al., 1990).
Worker exposures associated with cross-shift differences in spirometric values are
consistent with formaldehyde-induced sensory irritation. Concordance has also been reported
between subjective irritant response and measured changes in pulmonary function further
supporting the possibility that cross-shift and short-term evidence of bronchial constriction may
be a reflexive response to sensory irritation.
A well-conducted residential epidemiology study by Krzyzanowski et al. (1990) was
considered to be the strongest among the candidate studies on the adverse pulmonary function
effects of formaldehyde for the purposes of deriving an RfC.
6.1.3.4. Asthmatic Responses and Increased Atopic Symptoms
The health effects of respiratory function, asthma and increased atopic response, have
been shown to be clinically related. For example, asthma affects pulmonary function and may be
triggered by an allergic response. These and other data suggest that there may be mechanistic
links between these two health effects. Formaldehyde-induced sensitization (Section 4.2.1.5)
may enhance the asthmatic response or may enhance an individual's response to an allergen
(Section 4.4). In both cases, sensitization results in phenotypic switching - or an individual
exhibiting clinical symptoms of a predisposition to asthma or atopy. Because of the connection
between the two endpoints, they are considered together herein.
Several cross-sectional studies have described a positive association between
formaldehyde concentration and asthma prevalence. A study on risk factors for the initial
physician diagnosis of asthma have shown concentration-dependent associations between
formaldehyde exposure and asthma (Rumchev et al., 2002). In a categorical analysis, Rumchev
et al. (2002) observed statistically significant effects above in-home formaldehyde
concentrations of 60 (J,g/m3, with increased but non-significant effects at 50-59 |ig/m3 that were
consistent with a concentration-response relationship. No effect was apparent at concentrations
in the next lower interval between 30-49 (j,g/m3. Garrett et al. (1999) reported a borderline
statistically significant association between bedroom formaldehyde concentrations and an
increased risk of atopy. The authors computed a respiratory symptom score for each child based
on the frequency of each of eight respiratory symptoms and this score was substantially and
This document is a draft for review purposes only and does not constitute Agency policy.
6-9 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
statistically significantly higher among the asthmatic children compared to non-asthmatic
children. Health effects were reported at formaldehyde concentrations greater than 50 (j,g/m3 but
the lowest formaldehyde concentration interval at which health effects were observed was 20-50
[j,g/m3. The findings of Garrett et al. (1999) are supported by the results of a chamber study
reported by Casset et al. (2006) of 19 sensitized adult asthmatics exposed to formaldehyde at a
concentration of 100 (j,g/m3 for 30 minutes. Casset and colleagues observed an increased
bronchial responsiveness to mite allergen exposure (p = 0.05) and noted the provocative dose
(PD20 for FEV1) for mite allergen was 34.3 ng after formaldehyde exposure and 45.4 ng after
air exposure. However, in study by Ezratty et al. (2007) exposure to 500 |ig/m3 formaldehyde
did not affect an allergen-induced increase in responsiveness to methacholine (p = 0.42) and
there was no formaldehyde-associated effect on the airway inflammatory response.
These observed health effects in humans are similar to the outcome of studies in
laboratory animals that show that formaldehyde can exacerbate existing immunogenic
hypersensitivity to known allergens (Sadakane et al., 2002; Tarkowski and Gorski, 1995; Riedel
et al., 1996). While potentiation varied based on sensitization protocols and formaldehyde
exposure regimens, the results support the finding that formaldehyde exposure can aggravate a
Type-I hypersensitivity response and may do so via a neurogenically initiated response.
Formaldehyde itself does not function as an allergen recognized by the immune system (Lee et
al., 1984) and does not appear to trigger formation of formaldehyde-specific IgE. Although
formaldehyde exposure has been reported to alter cytokine levels and immunoglobulins in some
experimental systems (Fujimaki et al., 2004a; Ohtsuka et al., 2003), these effects do not support
an immunogenically mediated type-I hypersensitivity. In studies in which either egg protein
(ovalbumin, OVA)-sensitized or dust mite (DerF)-sensitized animals were exposed to
formaldehyde, OVA-specific and DerF-specific antibody production was increased over
sensitization alone, suggesting that formaldehyde may potentiate sensitization responses (Riedel
et al., 1996; Sadakane et al., 2002). Formaldehyde-induced sensitivity responses may be
neurogenic in origin based on findings that neurogenic factors such as nerve growth factor
(NGF) and substance P were associated with formaldehyde exposure in sensitization protocols
(Fujimaki et al., 2004).
6.1.3.5. Effects on the Immune System
Formaldehyde-induced systemic immunomodulation in laboratory animals has been
documented in the literature (Leach et al., 1983; Dean et al. 1984; Adams et al., 1987). A
number of studies have evaluated the ability of formaldehyde to induce systemic immunotoxic
effects in humans (Ohtani et al., 2004a, b; Erdei et al., 2003; Thrasher et al., 1990, 1987; Pross et
This document is a draft for review purposes only and does not constitute Agency policy.
6-10 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
al., 1987). Some studies have reported altered innate immune responses associated with
formaldehyde exposure (Erdei et al., 2003), while others have noted adaptive immune response
suppression associated with formaldehyde exposure (Thrasher et al., 1990, 1987) and changes
associated with alterations to a predominant T—lymphocyte helper 2 (Th2) pattern (Ohtani et
al., 2004a, b). In contrast, Pross et al. (1987) did not observe formaldehyde-associated changes
in systemic immune function.
Diverse studies have investigated the possibility that formaldehyde exposure leads to
increased respiratory tract infections (Lyapina et al., 2004; Krzyzanowski et al., 1990; Holness
and Nethercott, 1989). Lyapina et al. (2004) reported increased respiratory tract infections and
decreased neutrophil respiratory burst activity in formaldehyde-exposed workers (at 722 ppb
TWA). Incidences of doctor-diagnosed chronic bronchitis were more prevalent in children
under age 15 living in homes with higher formaldehyde (>60 ppb) readings in the kitchen (p <
0.001) (Krzyzanowski et al., 1990). Holness and Nethercott (1989) also report increased chronic
bronchitis in formaldehyde-exposed funeral workers (380 ppb average exposure).
6.1.3.6. Neurological Effects
Formaldehyde exposure via inhalation has been shown to adversely impact nervous
system function in laboratory animals and humans, although human data for formaldehyde-
induced neurological effects are limited. Studies in formaldehyde-exposed histology technicians
provide evidence of neurological impairment, including lack of concentration, impaired memory,
disturbed sleep, impaired balance, variations in mood and irritability. These effects were
significantly correlated with increasing duration of exposure to formaldehyde, but the findings
are not conclusive due to confounding by concomitant exposures to other neurotoxic solvents
(Kilburn et al., 1985, 1987). In a prospective study, Weisskopf et al. (2009) found a strong
association between duration of formaldehyde exposure and death from amyotrophic lateral
sclerosis (ALS), but information regarding exposure levels was not available. Short-term studies
with controlled exposure to humans (chamber studies) also provide limited support for changes
in cognitive function immediately following a single, controlled formaldehyde exposure (Bach et
al., 1990; Lang et al. 2008).
Available animal data provide substantial evidence of behavioral changes in animals
following single or short-term repeated inhalation exposures to relatively low levels of
formaldehyde. Among the animal studies, none of the available studies examined effects on
nervous system function following chronic formaldehyde inhalation, however.
Reported perturbations in nervous system function following formaldehyde exposure in
animal studies include reductions in motor activity, lack of habituation, impairment in
This document is a draft for review purposes only and does not constitute Agency policy.
6-11 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
acquisition of a new learning task, deficits in retention of a previously learned task, increases in
corticosterone levels, sensitization to cocaine-induced locomotor activity, and enhanced fear
conditioning using an olfactory conditioned stimulus (CS) (see Table 4-57). Behavioral effects
have been seen in multiple laboratories and in studies conducted by different investigators using
a variety of testing paradigms. Many of these effects were observed at acute exposure levels at
or below 1.0 ppm, and some persisted days to weeks after termination of exposure.
More limited data indicate possible effects on the development of the nervous system,
including changes in brain structure and in the behavior of offspring (Table 4-57). Similarly,
there is very little information regarding the mechanism by which effects on the nervous system
might be produced. The data regarding behavioral sensitization provide some support for a
stress-related mechanism for those specific findings, but the applicability of this mechanism to
the behavioral changes seen in the other studies, including the learning deficits and
developmental findings, has not been evaluated. Although there are data supporting stimulation
of the trigeminal nerve by formaldehyde (and documenting the relevance of this interaction to
the sensory irritation caused by formaldehyde), there are no data supporting a causal relationship
between irritant properties of formaldehyde and the behavioral and neurodevelopmental effects
in humans that occur following formaldehyde exposure. In summary, none of the available data
provide sufficient information to allow a determination of the mode of action for effects of
formaldehyde on the adult or developing nervous system.
6.1.3.7. Reproductive and Developmental Effects
Formaldehyde inhalation exposure has been associated with adverse developmental and
reproductive outcomes in both epidemiologic studies and experimental animal studies. Observed
developmental outcomes include fetal loss, structural alterations, growth retardation, and delays
in functional development.
Several occupational studies found an increased risk of spontaneous abortions among
formaldehyde-exposed women (Taskinen et al., 1999, 1994; John et al., 1994; Seitz and Baron,
1990; Axelsson et al., 1984). The Taskinen et al. (1999) study examined several reproductive
outcomes in women employed in the wood-processing industry, with a range of average daily
formaldehyde exposures. The authors found that formaldehyde was associated with a more than
three-fold increased risk of spontaneous abortion, and with a nearly 50% decrease in a measure
of delayed conception indicating reduced fertility, an increased time to pregnancy, and an
increased risk for endometriosis in this study. In experimental animal studies, early fetal death
was noted following maternal formaldehyde exposures (Kitaev et al., 1984; Sheveleva, 1972),
supporting the epidemiologic findings that the spontaneous abortion is likely related to
This document is a draft for review purposes only and does not constitute Agency policy.
6-12 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
formaldehyde exposure. Kitaev et al. (1984) hypothesized that formaldehyde may affect
reproductive function by stimulating the hypothalamus-pituitary-gonadal (HPG) axis, based on
their observations of increased ovary weight, increased number of ovulating cells, and changes
in blood levels of gonadotropins (LH and FSH) in female rats. Additionally, Maronpot et al.
(1986) reported endometrial hypoplasia with a lack of ovarian luteal tissue in formaldehyde-
exposed female rats. This finding may be relevant to the increased risk for endometriosis noted
in the Taskinen et al. (1999) study. However, additional human and animal studies are needed to
better understand the effects of inhalation exposure to formaldehyde on developmental outcomes
after early gestational windows of exposure or on the female reproductive system.
The findings of some occupational studies have suggested formaldehyde-related
associations with congenital malformations and low birth weight. In numerous experimental
animal studies, developmental effects have been noted following inhalation exposures to
formaldehyde (Table 4-68). Exposure of rat dams to formaldehyde during pregnancy has been
shown to result in significantly decreased fetal weight gain (Martin, 1990; Saillenfait et al.,
1989; Kilburn and Moro, 1985). Other studies have noted changes in relative organ weight,
undescended testes, biochemical changes (e.g., ascorbic acid), and blood acidosis (Senichenkova
and Chebotar, 1996; Senichenkova, 1991; Kilburn and Moro, 1985; Gofmekler and
Bonashevskaya, 1969; Gofmekler, 1968; Pushkina et al., 1968).
Studies designed to assess adult male reproductive system toxicity in rats following
repeated inhalation exposures to formaldehyde have found concentration-dependent decreases in
Leydig cell number and quality, degeneration of seminiferous tubules, decreases in testes weight,
alterations in sperm measures, decreased testosterone levels, alterations in trace metals in the
testes, and/or dominant lethal effects (Guseva, 1972; Ozen et al., 2002, 2005; Sarsilmaz et al.,
1999; Xing et al., 2007; Zhou et al., 2006) (Table 4-71).
6.1.3.8. Effects on General Systemic Toxicity
Extrapulmonary effects such as changes in liver function enzymes and focal, chronic
inflammation in the heart and kidney have been observed due to formaldehyde exposure in
experimental animal studies. Most of these changes occurred at exposures of 20 ppm, and those
that occurred at lower formaldehyde exposures (3.7 ppm) were confounded by coexposures. The
underlying modes of action of liver, kidney, and cardiac effects have not been elucidated, and the
human relevance is unknown.
This document is a draft for review purposes only and does not constitute Agency policy.
6-13 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
6.1.3.9. Summary
Formaldehyde-induced eye, nose and throat irritation, decreased pulmonary function,
decreased mucociliary clearance and histopathological lesions have been extensively
documented in human and laboratory animal studies. These health effects are commonly noted
at the portal of entry as expected for exposure to a reactive gas. In addition, effects on immune
system responses and on the nervous and reproductive systems, including developmental effects,
have also been reported. An association between formaldehyde exposure and increased
incidence and severity of response to allergens (i.e., asthma and atopy) has been noted in
humans. This effect, which has also been studied in laboratory animals, might occur via a
neurogenic mode of action. A limited database of information that evaluates neurological effects
in humans following formaldehyde exposure demonstrates a potential for adverse outcomes, and
studies in laboratory animals have reported a variety of formaldehyde-induced neurobehavioral
and neurodevelopmental effects. Formaldehyde has also been associated with adverse
reproductive outcomes. Human studies have reported an association between formaldehyde
exposure and decreased fertility as well as an increased risk of spontaneous abortions. Other
human studies have suggested formaldehyde-related associations with congenital malformations,
low birth weight, and endometriosis. Animal studies have noted a variety of developmental
effects, including fetal death, structural alterations, and growth retardation (e.g., delayed fetal
skeletal ossification and decreased fetal body weight) following inhalation exposure to
formaldehyde, and adverse reproductive effects have been observed in both males and females.
6.1.4 Carcinogenicity in Humans and Laboratory Animals
6.1.4.1 Carcinogenicity in Humans
Based on the total weight of evidence, including the results from a large and well-
followed longitudinal cohort study of 25,619 industrial workers and several case-control studies,
the epidemiologic evidence is sufficient to characterize the association between formaldehyde
nasopharyngeal cancer as causal in humans (Hauptmann et al., 2004; Hildesheim et al., 2001;
Vaughan et al., 2000). As further evaluated below, the evidence supporting a positive association
between formaldehyde exposure and NPC is unlikely due to chance, bias or confounding.
However, it should be noted that other smaller studies of formaldehyde-exposed workers did not
document increased NPC mortality (e.g., Coggon et al., 2003; Pinkerton et al., 2004). These
smaller study sizes yielded effect estimates with wide confidence intervals that were not
statistically inconsistent with the increased risk of mortality from nasopharyngeal cancer
reported in Hauptmann et al. (2004).
This document is a draft for review purposes only and does not constitute Agency policy.
6-14 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Luce et al. (2002) evaluated pooled data from 12 case-control studies conducted in seven
countries using a common job-exposure matrix and demonstrated a statistically significant
increased risk between formaldehyde exposure and sinonasal cancer exhibiting a concentration-
response relationship providing further causal evidence of carcinogenicity. This analysis was
based on a very large dataset of 930 cases and 3136 controls, enabling the investigators to
control for multiple potential sources of bias and confounding and to conduct separate analyses
by histological type. These results are particularly convincing, as the association was
consistently seen for a rare sub-type of sinonasal cancer which normally accounts for only 10%
of the reported cases.
In addition to the evidence of formaldehyde carcinogenicity in the nasopharynx, nose and
sinuses, other upper respiratory tract sites of direct contact with formaldehyde upon inhalation
(i.e., larynx, mouth and salivary gland) also showed evidence of increasing relative risk with
increasing average intensity and peak exposure in a large cohort study with exposure estimates
for the individual workers, although these trends did not reach the level of statistical significance
(Hauptmann et al., 2004). However, Hauptmann and colleagues (2004) concluded that in spite
of the small numbers of deaths from these rare cancers of the upper respiratory tract, the positive
associations of increased cancer risk with increased formaldehyde exposure were consistent with
the carcinogenicity of formaldehyde at these sites of first contact. Case-control studies also
provide evidence of an association between formaldehyde exposure and oral squamous cell
carcinoma (SCC), esophageal, and laryngeal cancers, and hypopharyngeal cancer (Gustavsson et
al., 1998; Laforest et al., 2000.)
The finding that formaldehyde inhalation causes nasal squamous cell carcinoma in
rodents (Section 4.2.1.2) further supports the determination of a causal association of
formaldehyde exposure and increased risk of upper respiratory tract cancer in humans. Both
humans and animals developed tumors within the upper respiratory tract, the site expected to
receive direct exposure to formaldehyde.
Several researchers have argued that the relationship between formaldehyde exposure
and nasopharyngeal cancer based on existing studies has not been determined. Several
limitations, such as the rarity of the cancer and the imprecise estimates of exposure, are often
inherent in epidemiologic methods and exposure assessment. These constraints limit the ability
of epidemiologic studies to statistically detect associations and can lead to false negatives. The
results of the largest cohort study of nasopharyngeal cancer (Hauptmann et al., 2004) showed
statistically significant concentration-response relationships with increased risk of cancer
associated with increased formaldehyde exposure. However, even though this study was based
on 25,619 workers, only 9 cases of nasopharyngeal cancer were observed, compared to an
This document is a draft for review purposes only and does not constitute Agency policy.
6-15 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
expected number of 5 cases, for a relative rate of 2.1 (with a confidence interval of 1.05 - 4.21)
(Hauptmann et al., 2004).
The next largest cohort study of nasopharyngeal cancer was based on 14,014 workers
(Coggon et al., 2003) and reported only 1 case compared to an expected number of 2 cases, for a
relative risk of 0.5 (with an estimated 95% confidence interval of 0.07 - 3.55; see Bosetti et al.,
2007). To put this finding into perspective, it is helpful to note not only the relative risk but also
that this effect estimate is highly unstable due to a lack of statistical power. The large width of
this interval (0.07 - 3.55) indicates that the range of possible true values includes both increased
and decreased NPC mortality and therefore does not contradict the evidence of elevated risk of
nasopharyngeal cancer mortality associated with formaldehyde exposure reported by Hauptmann
et al. (2004). The even smaller study of 11,039 textile workers by Pinkerton et al. (2004)
reported no cases of nasopharyngeal cancer compared to an expected number of one case -
yielding an effective relative risk of zero with a highly unstable 95% confidence interval
estimated at 0 - 3.00 (see Bosetti et al., 2007). While true that Pinkerton et al. (2004) did not
report an increased risk of nasopharyngeal cancer, this study did not have sufficient statistical
power to rule out a true association with less than a 3-fold increase in risk and therefore is
likewise not inconsistent with the finding by Hauptmann et al. (2004). Thus, results from these
cohort studies, with limited power to detect the relatively rare upper respiratory tract cancers
(e.g., NPC), are given less weight in the overall evaluation.
The largest occupational cohort study, conducted by the NCI (Hauptman et al., 2004), did
report statistically significant associations of formaldehyde exposure with carcinogenicity at the
sites of first contact with sufficient statistical power to rule out the null hypothesis of no
association. The NCI investigations controlled for potential selection bias due to the healthy
worker effect and for several potential confounders, including calendar year, age, sex, race, and
pay category. However, other potential sources of bias or confounding have been suggested with
respect to the strength of these data to support a causal conclusion.
Following reports of increased risk of NPC associated with formaldehyde exposure, a
series of analyses of similar data were undertaken by Marsh and coworkers (Marsh et al., 2007a,
b, 2002, 1996; Marsh and Youk, 2005). Briefly, these studies focused on the specific findings
from a single plant in the NCI cohort (Wallingford, Connecticut) that generated the majority of
the NPC cases. Marsh et al. (1996) confirm a significant adverse association of formaldehyde
with nasopharyngeal cancer but note the effects are predominantly among workers at the
Wallingford plant with less than one year employment. Marsh et al. (2002) report a five-fold
excess in risk of nasopharyngeal cancer associated with formaldehyde in both short-term and
long-term workers but note that the increase was concentrated among workers hired during
This document is a draft for review purposes only and does not constitute Agency policy.
6-16 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
1947-1956. Marsh and Youk (2005) re-evaluated the same Wallingford workers and reported a
regional rate-based standardized mortality ratio (SMR) of 10.32 (95% CI = 3.79 - 22.47)
compared to 0.65 (95% CI = 0.08 - 2.33) for workers at the nine other plants combined.
However, Marsh and Youk (2005) also show that rate-based mortality ratios standardized to both
U.S. and local populations were elevated (non-significantly) not only at the Wallingford plant
but individually at each of the four other plants at which a single case of nasopharyngeal cancer
was reported: Plant 2 (SMRus = 5.35), Plant 3 (SMRus = 1.99), Plant 7 (SMRus = 106), and
Plant 10 (SMRus = 1.44). It should be noted that Plant 1 (Wallingford) and Plant 2 had both the
two highest median formaldehyde exposures and the two highest reported excess risks (Marsh
and Youk, 2005).
In another re-analysis of the NCI cohort data on the workers at the Wallingford plant,
Marsh and coworkers (2007a) suggested that an imprecise assessment of formaldehyde exposure
and an inability of the study to separate formaldehyde exposure from other potential chemical or
particulate exposures may have confounded the observed association between formaldehyde and
cancer. However, there was no evidence of any differential measurement error that could have
produced the observation of a spurious association. Any non-differential exposure measurement
error (i.e., random error in the exposure assessment) would likely have led to an attenuated
observed effect of formaldehyde that was less than that which would otherwise have been
observed in the absence of measurement error.
The potential for confounding by particulates was explicitly examined by Hauptmann et
al. (2004) and it was shown that there was an exposure-response relationship with formaldehyde
among individuals with high particulate exposures - alleviating the potential for confounding
and thereby strengthening the causal interpretation of the formaldehyde relationship with an
increased risk of NPC. Marsh and coworkers (Marsh et al., 2007b) later suggested the reported
formaldehyde association was confounded by an association between silversmithing and NPC.
However, careful examination of that analysis (Marsh et al., 2007a) suggests that multiple
comparisons may have led to the reported observation with silversmithing. Additionally, the
reported effect was inconsistently reported between the results and the abstract sections using
different confidence intervals, and both sets of confidence intervals around the reported
association were extremely unstable spanning up to several hundred-fold. No prior studies
identified an associated between silversmithing and NPC. Thus it may be that silversmithing is
an artifactual potential confounder.
The increased NPC mortality observed in the NCI cohort (Hauptmann et al., 2004) has
been thoroughly examined for sources of bias and confounding by both the primary researchers
and Marsh and coworkers (Marsh et al., 2007a, b, 2002, 1996; Marsh and Youk, 2005). Despite
This document is a draft for review purposes only and does not constitute Agency policy.
6-17 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
the extensive scrutiny of these results, no convincing and consistent alternative hypothesis of
causation has been identified. Taken together with the statistically significant association
demonstrating an exposure-response relationship within exposed workers, these data support the
conclusion that the association between formaldehyde exposure and increased risk of NPC is
causal.
Therefore, after a thorough examination of potential confounders, the association
between formaldehyde exposure and NPC mortality in the NCI cohort remains significant and
provides a positive exposure-response relationship. Additionally, case-control studies, which
have greater statistical power than cohort studies for rare diseases, provide strong additional
evidence in support of a causal association between formaldehyde exposure and the incidence of
NPC (Hildesheim et al., 2001; Vaughan et al., 2000). As these studies draw from different
demographic groups, regions of the world, and evaluate various confounding factors, there is
little potential for these consistently reported associations to be artifactual, confounded by
common exposures, or a result of bias or chance.
Numerous epidemiologic studies have also reported an association between
formaldehyde-exposed workers, especially "professional" workers (e.g., pathologists,
embalmers, and funeral directors), and increased risk of lymphohematopoietic cancers (See
Table 4-82). Positive associations between formaldehyde exposure and lymphohematopoietic
cancers have been reported for chemical workers (Wong et al., 1983; Bertazzi et al., 1986),
embalmers (Walrath and Fraumeni, 1983, 1984; Hayes et al., 1990), anatomists and pathologists
(Harrington and Shannon 1975; Hall et al., 1991; Levine at al., 1984; Stroup et al., 1986;
Matanoski et al., 1989). However, clear associations (in terms of overall standardized mortality
ratios (SMRs) or proportional mortality ratios (PMRs) were not reported in analyses for garment
workers, iron-foundry workers, and a large US industrial cohort (Pinkerton et al., 2004;
Andjelkovich et al., 1995; Beane Freeman et al., 2009; Marsh et al., 1996), although associations
were observed in some of these studies when exposure-response relationships were considered.
Several published meta-analyses are available which more formally assess the strength of
association between formaldehyde exposure and mortality from all lymphohematopoietic
cancers. Pooled SMRs indicate stronger associations for professional workers (embalmers,
anatomists and pathologists) than industry workers (Table 4-83). Bosetti et al. (2008) found
similar relationships, with a pooled SMR of 1.31 (95% CI 1.16-1.47) for 'professionals' (i.e.
embalmers, anatomists and pathologists) versus a pooled estimate of 0.85 (95% CI 0.74-0.96) for
industrial workers. A recent meta-analysis by Zhang et al. (2009) reports a summary relative
risk of 1.25 (95% CI 1.09-1.43) for both professional and industry workers for all
lymphohematopoietic cancers (ICD 9 codes 200-209).
This document is a draft for review purposes only and does not constitute Agency policy.
6-18 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Two well-designed cohort studies found significant positive associations between
formaldehyde-exposed professional workers and lymphohematopoietic cancer, particularly
leukemia, using cumulative exposure measures not previously used and using internal
comparison groups. The largest cohort study of industrial workers exposed to formaldehyde
(N=25,619), with the most extensive exposure assessment (Blair et al., 1986; Stewart et al.,
1986) and with the cohort followed for a median duration of 35 years (Hauptmann et al., 2003)
demonstrated that formaldehyde was a risk factor for lymphohematopoietic cancers, independent
of other risk factors, such as benzene exposure and smoking. This finding was re-confirmed
with an additional 10 years of follow-up (Beane Freeman et al., 2009). Another industrial cohort
study reported a significant increase in the risk of leukemia in garment workers 20 years after
their initial exposure and in workers with 10 or more years of exposure to formaldehyde
(Pinkerton et al. 2004). A third large occupational cohort study (Coggon et al., 2003) that did
not evaluate their findings with regard to latency reported somewhat lower mortality from
leukemia and other lymphatic and hematopoietic cancers than expected compared to national
rates.
The associations between myeloid leukemia and formaldehyde exposure are strong and
consistent (Table 4-84). Of the four studies which formally assess myeloid leukemia mortality,
all are positive, including cohorts of both professional and industrial workers (Beane Freeman et
al., 2009; Hayes et al., 1990; Pinkerton at al., 2003; Stroup et al., 1986). Although few cases
exist for further subtype analysis, the available data indicate either no differences in SMRs for
acute myeloid leukemia (AML) versus chronic myeloid leukemia (CML) (Hayes et al., 1990;
Pinkerton et al., 2003) or suggest CML is more prominent (Blair et al., 2000; Stroup et al.,
1986). The association between formaldehyde exposure and myeloid leukemia in embalmers has
recently been confirmed in a large nested case control study by Hauptman et al (2009) which
includes cases identified from the previous studies of Hayes et al. (1990) and Walrath and
Fraumeni (1983 and 1984). Exposure estimates were based on interviews with next-of kin for
duration of job actively embalming and total number of embalmings performed. Strong and
statistically significant exposure-response relationships are demonstrated for duration of
exposure, total number of embalmings performed and estimated cumulative exposure to
formaldehyde with odds ratios of 13.6 (1.6-119.7), 12.7(1.4-112.8) and 13.2(1.5-115.4)
respectively (Hauptmann et al., 2009).
The reported associations between formaldehyde exposure and lymphohematopoietic
cancers in general, and leukemia (especially myeloid leukemia) in particular, were in workers
exposed in very different environments (i.e., mortuary, chemical industry and garment industry).
Since coexposures to other agents are considerably different between these work environments,
This document is a draft for review purposes only and does not constitute Agency policy.
6-19 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
it is unlikely that influence of confounding exposures plays a role in the observed associations.
There is no evidence of bias in the published reports, and the consistency across numerous
studies over time is sufficient to conclude that the results are not due to chance.
It has been argued that it is biologically implausible for a highly reactive agent such as
formaldehyde, whose primary action is expected to be at the portal of entry, to cause acute
lymphoid or myeloid leukemias (ALL and AML, respectively), which are both commonly
believed to arise from transformation of stem cells in the bone marrow. The modes of action
(MOAs) by which formaldehyde may induce these observed cancers are unknown, although it
has been postulated that circulating stem cells (Hauptmann et al., 2003) (e.g., early progenitor
cells in circulating blood or pluripotent cells in nasal/oral passages) may travel to bone marrow
where they become leukemic stem cells (Zhang et al., 2009). In contrast, the mechanism for the
chronic lymphatic leukemia, lymphomas, multiple myelomas (from plasma B-cells) and
unspecified lymphohematopoietic cancers may involve an etiology in peripheral tissues, such as
cells, cell aggregates, germinal centers and lymph nodes. An association of these cancers to a
reactive exogenous agent primarily acting at the point of entry is biologically plausible.
6.1.4.2 Carcinogenicity in Laboratory Animals
The carcinogenic potential of formaldehyde is well documented in numerous animal
bioassays, especially for sites of first contact. Inhalation exposure of formaldehyde induced
primarily squamous cell carcinomas (SCC) in nasal passages of rats (Feron et al., 1988;
Holmstrom et al., 1989a; Woutersen et al., 1989; Tobe et al., 1985; Kamata et al., 1997; Albert et
al., 1982; Sellakumar, 1985; Kerns et al., 1983; Monticello et al., 1996) and mice (Battelle
Columbus Laboratories, 1981; Swenberg et al., 1980; Kerns et al., 1983; CUT, 1982).
Formaldehyde given as 0.5% formalin orally in drinking water to adult rats induced higher
incidences of papillomas in the forestomach, adenomatous hyperplasia in the fundus, and
adenocarcinomas in the pylorus in a 40-week study using an initiation-promotion protocol in rats
(Takahashi et al., 1986). Soffritti et al. (1989) observed a significant increase in rare tumors in
the gastro-intestinal (GI) tract, including both benign (papillomas and acanthomas of the
forestomach and adenomas) and malignant tumors (adenocarcinomas and leiomyosarcomas) in
rats given formaldehyde in drinking water. Formaldehyde is toxic at the portal of entry in
rodents, causing increased cell proliferation, DPX formation, and focal lesions in the GI tract or
upper respiratory tract (depending on the route of exposure). The portal of entry toxicity of
formaldehyde further supports a finding of formaldehyde induced POE cancer in animal
bioassays.
This document is a draft for review purposes only and does not constitute Agency policy.
6-20 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Direct support for lymphohematopoietic cancers in animal bioassays is less convincing.
Although many of the available chronic studies did not examine lymphoma/1 eukemia incidence,
two studies provide positive evidence. Inhalation exposure of formaldehyde increased
lymphoma in female mice and leukemia in female F344 rats, but not male rats (Battelle
Laboratories, 1981). Drinking water exposure to formaldehyde caused a dose-dependent
increase in all hemolymphoreticular neoplasias, especially lymphoblastic leukemias and
lymphomas in both male and female Sprauge-Dawley rats (Soffritti et al., 1989, 2002).
Conversely, no increases were seen in male Wistar rats when exposed to formaldehyde in
drinking water at similar levels (Til et al., 1989) or male rats after chronic inhalation exposures
(Sellakumar et al., 1985).
6.1.4.3 Carcinogenic Mode(s) of Action
Multiple plausible modes of action (MOAs) are presented in the document so as to
explore ways in which a combination of factors may contribute to cancer incidence in a
population exposed to formaldehyde. Multiple MO As for formaldehyde-induced cancer can be
reasonably supported based on various known biological actions of formaldehyde (e.g.,
mutation, cell proliferation, cytotoxicity and regenerative cell proliferation). Additionally,
alternative actions, such as immunosuppression or viral reactivation, are possible, although few
data exist to evaluate their potential relevance. Rather than a single MO A, it is plausible that a
combination of these factors contribute to cancer incidence in an exposed population.
Considering multiple factors may help to better understand the biological and mechanistic basis
for the increases in cancer incidence observed in exposed human populations. Unlike animal
bioassays, results in human epidemiological studies reflect not only the effects of the agent of
concern but also numerous other risk factors (e.g., viral status, diet, smoking, etc.). Additionally,
human studies may be impacted by biological human variability across individuals, cancer
biology (sub-types) and wide variability in exposure regimens in human populations.
A preponderance of the evidence supports a role of mutagenic activity in formaldehyde's
carcinogenic MOA both for respiratory tract cancer and lymphohematopoietic cancers. As
reviewed in Section 4.3, numerous studies provide evidence of formaldehyde's direct mutagenic
activity and supports the relevance these data to formaldehyde's carcinogenicity. It can be
shown that:
1) Formaldehyde directly interacts with DNA, generating DNA-protein cross-links and
DNA adducts (in vitro, in vivo) in multiple species,
This document is a draft for review purposes only and does not constitute Agency policy.
6-21 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
2) DNA-protein cross-links exhibit a dose-response relationship to formaldehyde
exposure in respiratory tract of laboratory animals and are observed at exposure
concentrations of relevance to some people (0.3 ppm, 0.7 ppm),
3) Formaldehyde-induced DNA-protein cross-links have been associated with
formaldehyde-induced micronuclei and chromosomal aberrations (in vitro),
4) Mutations induced by formaldehyde due to small deletions and rearrangements in
DNA in various experimental systems are consistent with formaldehyde's
observed clastogenic effects (micronuclei and chromosomal aberrations) (in vitro,
in vivo),
5) Formaldehyde-induced mutations and clastogenic effects occur at levels below where
significant cytotoxicity is detected (in vitro),
6) Formaldehyde exposure has been correlated to similar increased micronuclei and
chromosomal aberrations in human buccal and oral cells corresponding to sites
where formaldehyde-induced tumors arise, and
7) Chromosomal damage in blood-borne immune cells, relevant to agent-induced
lymphohematopoietic cancers has been documented in formaldehyde exposed
workers including increased micronuclei and chromosomal aberrations, increased
incidence and aneuploidy in hematopoietic stem cells.
In addition, mutations may arise indirectly from formaldehyde-induced DNA damage
during cell proliferation or due to errors in DNA repair mechanisms. Therefore, formaldehyde's
DNA reactivity on a population of proliferating cells strengthens the role of formaldehyde-
induced mutagenicity in its carcinogenic MOA. The nasal and gut mucosa are tissues which are
continually sloughing and regenerating cells (Junqueira et al., 1992). Mucosal cells proliferate
in response to environmental challenges in order to repair cell damage, increase adaptive
response and remodel tissue. Additionally, since the pseudostratified epithelium of the
respiratory tract is only 1-2 cells in depth, cells with proliferative capacity would be directly
impacted by formaldehyde during exposure. Formaldehyde-induced clastogenic effects have
been demonstrated in these tissues (e.g. nasal) in humans, as well as in tissues which possess
stratified epithelium (e.g. buccal). Therefore, formaldehyde would not need to transport beyond
the portal of entry to directly impact and induce DNA mutations in routinely proliferating cells.
In regards to generating the observed clastogenic effects (micronuclei and chromosomal
aberrations in peripheral blood lymphocytes, aneuploidy in circulating hematopoietic stem cells),
it is less clear as to where formaldehyde is making contact with components of the immune
system. Mature lymphocytes present in nasal and gut tissues, and would be vulnerable to the
This document is a draft for review purposes only and does not constitute Agency policy.
6-22 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
direct toxic actions of formaldehyde including genotoxicity. Since mature lymphocytes
routinely traffic through the body and clonally respond in response to an immune challenge, the
observed effects in peripheral blood lymphocytes (micronuclei and chromosomal aberrations)
are consistent with direct action on these cells. Lymphohematopoietic cancers are known to
arise from mature lymphocytes including: Hodgkin lymphoma, multiple myeloma some
leukemia and non-Hodgkin lymphoma (Greaves 2004, Harris et al., 2000).
Formaldehyde may also be directly acting upon circulating stem cells or more mature
progenitor cell in the peripheral blood (Zhang et al., 2010). Any genetic damage sustain by
circulating cells could contribute to a broad spectrum of lymphohematopoietic cancers if those
cells returned to the bone marrow and contributed to hematopoiesis. Evidence of bone marrow
toxicity and stem cell aneuploidy has been reported in formaldehyde exposed workers (Zhang et
al., 2010). Finally, formaldehyde is readily hydrated in aqueous systems, existing in equilibrium
with its hydrated form methylene glycol, which is able to transport through the blood. It has
been hypothesized that this hydration reaction may allow formaldehyde to act systemically and
therefore on the bone marrow directly (Zhang et al., 2010.) Formaldehyde-induced DNA
damage, and resulting mutation in the bone marrow and circulating stem cells could contribute to
any of the lymphohematopoietic cancers including leukemia (both lymphoid and myeloid) as
well as myeloproliferative disorders.
Cell replication allows unrepaired DNA damage to be "fixed" into heritable changes to
the genome. Therefore, increased cell proliferation could serve not only to increase the
mutagenic effects of formaldehyde on a given tissue but also to enhance the mutagenic effects of
other agents in the diet or in the environment. Since epidemiological studies include humans
exposed to a range of agents in the environment, increased cell proliferation could contribute to
increased cancer incidence. The promotion studies in animal bioassays, though limited in
number, support the relevance of formaldehyde's ability to enhance the actions of other agents
(initiators) on tumor formation.
Although the other biologic effects discussed above have not been explicitly tested in
animal systems, the available data are consistent with these actions contributing to the
carcinogenic potential of formaldehyde. For example, localized immunosuppression by
formaldehyde may serve to increase viral reactivation (e.g., EBV, HPV etc.) or decrease tissue
surveillance and immune activity against preneoplastic cells. Both these actions could contribute
to increased cancer risk in a human population, which may not be evident in animal bioassays,
where the animals are not subject to the many risk factors for human cancer. Even the simple
action of the breakdown of the mucociliary apparatus could increase cancer incidence by
This document is a draft for review purposes only and does not constitute Agency policy.
6-23 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
increasing toxic insult to the URT and increasing URT infections. Again, these actions may be
relevant to human populations, but they have not been adequately tested in animal bioassays.
Animal bioassays suggest a role for regenerative proliferation in contributing to
formaldehyde's carcinogenicity. However, these data are not evidence against a role of direct
mutagenic action either in the observed tumorigenicity or in the potential low-dose
carcinogenicity of formaldehyde. As reviewed, a role for mutagenic action is also consistent
with the results of the animal bioassays (Crump et al, 2008; Subramaniam et al., 2007, USEPA
2008). The mutagenic effects of formaldehyde are well-documented to occur below levels of
significant cytotoxicity. This observation is important for the relevance of formaldehyde-
induced mutagenicity to human health risk. Given the above sequence of evidence - from the
nature of formaldehyde's DNA reactivity through clastogenic effects observed in human cells
from the various tumor sites - there is an adequate weight of evidence (WOE) to consider
formaldehyde-induced mutations relevant to human carcinogenic risk. Although occupational
exposures may have resulted in high episodic exposures (especially historically), it is unlikely
that any worker would have endured repeated exposures which resulted in gross focal lesions to
the upper respiratory tract (URT) or oro-digestive tract as seen in the animal bioassays. It is
noteworthy that even without these gross formaldehyde-induced lesions, cancer incidence is
increased from occupational (and perhaps non-occupational) exposures to formaldehyde.
Therefore, we believe formaldehyde carcinogenicity can be attributed, at least in part, to a
mutagenic MOA.
6.1.5 Cancer Hazard Characterization
Formaldehyde is carcinogenic to humans by the inhalation route of exposure. There is
sufficient evidence of causal associations between formaldehyde exposure and nasopharyngeal
cancers as well as sinonasal cancers. There is supporting evidence for cancers of the mouth and
throat in humans as well as strong evidence for nasal tumors in animal bioassays. Taking these
findings together along with mode of action considerations, it is concluded that there is sufficient
evidence of a causal relationship between formaldehyde inhalation exposure and upper
respiratory tract cancers as a group.
Epidemiologic studies also provide evidence of a causal association of inhalation
exposure to formaldehyde and lymphohematopoietic cancers as a group and leukemias as a
group, with the strongest evidence for myeloid leukemia. There is supporting evidence both in
cohort and case-control studies for specific sub-types of lymphohematopoietic cancers, including
myeloid leukemia, multiple myeloma and Hodgkin lymphoma. There is limited supporting
evidence in animal bioassays for both leukemia and lymphoma. It should be noted that although
This document is a draft for review purposes only and does not constitute Agency policy.
6-24 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
several carcinogenic modes of action have been discussed for leukemia, that this remains an area
of significant scientific debate.
6.2 DOSE-RESPONSE CHARACTERIZATION
6.2.1 Noncancer Toxicity: Reference Concentration (RfC)
The portals of entry are major targets for formaldehyde, as can be seen in many studies,
because formaldehyde is highly reactive and water soluble. Human and laboratory animal
studies demonstrate that formaldehyde also causes systemic effects, including neurotoxicity,
reproductive toxicity, developmental toxicity, and immunotoxicity, although the data are less
extensive than those supporting the sensory irritation and respiratory tract effects. Critical data
gaps have been identified and uncertainties associated with data deficiencies are more fully
discussed in Chapter 5 and summarized below.
6.2.1.1 Assessment Approach Employed
RfC values for noncancer effects are derived using EPA's RfC methodologies (U.S.
EPA, 1994, 1993, EPA 2002b). EPA reviewed the existing literature and identified health
effects associated with formaldehyde exposure, defining health effect categories where evidence
was sufficient: sensory irritation, respiratory tract pathology, pulmonary effects, asthma,
increased allergic sensitization, immune function, neurological and behavioral effects and
reproductive and developmental effects. Specific key studies were identified within each health
effects category which provided adequate exposure-response information to support RfC
derivation (Table 5-4). Although not all identified endpoints are represented by these studies, at
least one study was identified for each category. A screening process (described in section
5.1.3.1) was used to identify key studies for a variety of health effects that would best inform the
derivation of the RfC. For each selected key study, a candidate RfC (cRfC) was derived. In
several cases more than one alternative was considered for application of the uncertainty factor
(UF) addressing human variability (Table 5-6).
6.2.1.2 Derivation of Candidate Reference Concentrations
Seven studies were selected as key studies for further consideration in RfC derivation
(Section 5.3.1, Table 5-4). Candidate RfCs from these studies address various health effects
including: sensory irritation, respiratory effects, asthma, increased allergic sensitization, and
decreased fecundity (Table 5-6). From these studies three co-critical studies were selected which
provide similar cRfCs for related health effects (Rumchev et al., 2002; Garrett et al., 1999;
Krzyzanowski et al., 1999). These three studies identify serious health effects in residential
This document is a draft for review purposes only and does not constitute Agency policy.
6-25 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
populations including children: increased asthma incidence, decreased pulmonary function,
increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al., 2002;
Garrett et al., 1999; Krzyzanowski et al., 1999). Asthma, allergic sensitization, altered
pulmonary function, and symptoms of respiratory disease are not only clinically related, but
etiologically related, and it is reasonable that they should be considered together. These health
effects are observed below the exposure levels that result in sensory irritation, and the resulting
cRfCs are correspondingly lower—ranging from 2.8 tol 1 ppb—depending on the study,
endpoint considered, and the application of alternative uncertainty factors for human variability
(Table 6-1). Additionally, these cRfCs are considered protective of the decreased fecundability
density ratio (FDR) reported by Taskinen et al. (1999) which yielded a cRfC of 8.6 ppb. One of
the uncertainties in the cRfC for decreased FDR is the use of a time-weighted exposure metric
which does not address possible contributions of peak exposure levels to the observed health
effect thus; it is possible that a cRfC of 8.6 ppb is lower than is needed for protection against
decreased FDR.
As discussed in section 6.2.1.4, there are uncertainties in establishing an RfC which are
not fully captured in the quantitative process or the standard uncertainty factors, as such it is
acknowledged by EPA that the RfC is not exact, perhaps spanning an order of magnitude. The
range of RfCs from the critical studies (even with various alternative considered for the human
variability uncertainty factor are in close agreement spanning only V2 order of magnitude.)
Therefore EPA is considering a simple mean of these cRfCs as adequately representative of the
three co-critical studies. Alternatives are to take the median as a different way to represent the
three studies together, or the lowest cRfC as most protective. There is little numerical difference
in the result of these decisions.
6.2.1.3 Adequacy of Overall Data Base for RfC Derivation
The database of available laboratory animal studies, clinical and epidemiological studies,
and supporting mechanistic information for formaldehyde is substantial. Many of the health
effects are well studied in animals and humans, especially those endpoints related to sensory
irritation and respiratory effects at the portal of entry, such as impacts on respiratory tract
pathology, asthma and reduced pulmonary function. This is reflected in the number and high
quality of human studies presented in Table 5-4 and supporting data summarized in Chapter 4.
This document is a draft for review purposes only and does not constitute Agency policy.
6-26 DRAFT—DO NOT CITE OR QUOTE
-------
Table 6-1: Summary of candidate reference concentrations (RfC) for co-critical studies
Endpoint
Study
Study
size
Homes
Children
POD (ppb)
Application of
study-specific UF
cRfC 1
(PPb)
ufl
UFS | UFh
Respiratory effects / asthma and sensitization
Reduction of
PEFR in children
(10%)
Krzyzanowski
et al. (1990)
208
Yes
Yes
BMCLio = 17
1
1
3
5.6
Asthma
prevalence
Rumchev et al.
(2002)
192
Yes
Yes
NOAEL = 33
1
3
Alternative A
3
3.3
Alternative B
l
11
Asthma, atopy
and severity of
allergic
sensitization
Garrett et al.
(1999)
148
Yes
Yes
LOAEL = 28
3
1
Alternative A
3
2.8
Alternative B
l
9.3
O
Notes: 1: The final RfC will be rounded to one significant digit per EPA policy. Since the Candidate RfC is an interim calculation two-significant digits are
retained as is common practice in mathematics {i.e. one significant diget more that the final result, to avoid rounding errors compounding across multiple
mathematical manipulations.
O
O
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
The data also indicate effects in other health effect categories, specifically neurotoxic
effects, reproductive toxicity, and developmental toxicity (Section 5.1.2). These non-portal of
entry effects are areas where additional research may be warranted to reduce uncertainty and
better characterize the potential for health effects and the formaldehyde concentrations at which
they might occur in humans.
EPA guidance indicates that an uncertainty factor for database deficiencies should be
applied where there is an indication that the existing studies may not completely characterize the
hazard of a specific agent. This may be the result of lacking studies to assess toxicity to key
functional areas or organ systems, or where "... a review of existing data may also suggest that a
lower reference value might result if additional data were available." (EPA 2002b)
Application of an uncertainty factor of 3 was considered by EPA based on the lack of a
satisfactory two-generation study to fully evaluate the effects of formaldehyde exposure on
reproductive and developmental endpoints and limitations of the available studies evaluating
neurotoxic effects. An uncertainty factor of 3 rather than 10 was considered given the relative
completeness of the database across all major health effect categories such that it is believed all
major health effects have been identified at least qualitatively. The observed adverse health
effect levels (LOAELs) for those endpoints where the database is not adequate for alternative
RfC derivation are above the range of candidate RfCs; however, it is unclear if the candidate
RfCs would be protective of these other health effects (neurotoxic, reproductive and
developmental) since NOAELs were not identified for several observed health effects.
Therefore EPA is considering several options to address database deficiencies in the final
RfC.
This document is a draft for review purposes only and does not constitute Agency policy.
6-28 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Approaches to the application of a database uncertainty factor:
Options EPA is considering include:
(1) Provide an RfC derived from studies of respiratory and allergenic responses and protective
of sensory irritation effects with a database uncertainty factor of one given significant data on
formaldehyde, but noting that further research reproductive, developmental and neurotoxic
effects would be valuable.
(2) Provide an RfC with a database uncertainty factor of one, with this RfC explicitly
identified as being protective of the well-studied effects.
(3) Apply a database UF of 3 to the RfC derived from studies of respiratory and allergenic
responses to reflect the potential that reproductive, developmental, or neurotoxic effects might
occur at lower doses:
(3) Provide both an RfC identified as protective of the better-studied effects and an RfC with
a database uncertainty factor of 3 incorporated to account for limits to the data on
reproductive, developmental and neurotoxic effects.
It is unclear what uncertainty factors are appropriate to account for human variability and
deficiencies in the overall database. For this reason, several alternatives have been presented.
6.2.1.4 Uncertainties in the Reference Concentration (RfC)
A number of uncertainties that underlie the RfC for formaldehyde are discussed in this
section. A fundamental uncertainty in an RfC is that the critical study(ies) and endpoint(s)
selected reflect an actual hazard, i.e., a chemically related effect. As summarized in Section
6.1.3, there is strong and consistent evidence, from both human and laboratory animal studies,
for the critical effects that form the basis of the RfC for formaldehyde. This section pertains to
uncertainties in the quantitative derivation of the RfC.
Point of Departure (POD)
Most of the studies considered for RfC derivation did not provide enough data to support
benchmark dose modeling. Rather, the PODs for most studies were LOAELs or NOAELs,
This document is a draft for review purposes only and does not constitute Agency policy.
6-29 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
which have a number of shortcomings relative to a POD obtained from benchmark dose-
response modeling (i.e., a benchmark concentration or dose):
¦ LOAELs and NOAELs are a reflection of the particular exposure/dose levels used in a
study, contributing some inaccuracy to the POD determination.
¦ LOAELs and NOAELs are often determined based on statistical significance and, thus,
reflect the number of study subjects or test animals. Studies are typically dissimilar in
detection ability and statistical power, with smaller studies tending to identify higher
exposure levels as NOAELs relative to larger, but otherwise similarly designed, studies.
¦ Different LOAELs and NOAELs represent different response rates, so direct qualitative
and quantitative comparisons are not possible.
PODs identified from benchmark dose models overcome some of the deficiencies
associated with LOAELs and NOAELs. Benchmark models were used for two inhalation data
sets—Hanrahan et al. (1984) and Krzyzanowski et al. (1990).
It should also be noted, however, that even for benchmark concentrations/doses there is
often uncertainty, in particular for continuous responses, about what response level to select as
the benchmark response, i.e., where to define the cut-point between a level of change that is not
adverse and one that is adverse. In addition, benchmark dose models currently in use are purely
mathematical models and are not intended to accurately reflect the biology of the effect being
modeled.
Another source of uncertainty in the POD is the adjustment for continuous exposure.
RfCs are meant to apply to continuous (24 hour/day) exposures. Exposure patterns in human
and laboratory animal inhalation studies are typically not for continuous exposures, and
assumptions must be made in converting reported exposure levels to equivalent continuous
exposures. Similarly, there are uncertainties about potential dose rate effects, in particular the
effect of peak exposures in occupational studies.
Extrapolation from Laboratory Animal Data to Humans
Because the inhalation database for formaldehyde contains many human studies for a
variety of health effects, it was not necessary to rely on animal data for the endpoints from which
the RfC was derived. Thus, unlike for most RfCs, this is not a source of uncertainty in the RfC
for formaldehyde.
This document is a draft for review purposes only and does not constitute Agency policy.
6-30 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Human Variation
Heterogeneity among humans is another uncertainty associated with extending results
observed in a limited human study population or laboratory animal experiment to a larger, more
diverse human population.
For three of the studies used to derive the RfC, a value of 3 was used for the human
variability UF (rather than the default value of 10) because the studies had an apparent over-
representation of populations expected to have increased susceptibility (section 5.5.3.1):
¦ The residential study by Ritchie and Lehnen (1987) evaluated eye, nose, and throat
irritation in a large number of subjects, including children and the elderly. As a result of
the study's participation criteria, individuals with greater sensitivity were potentially
over-represented.
¦ Thirty percent of the subjects in the residential study by Krzyzanowski et al. (1990) are
children, who are more sensitive to formaldehyde-associated decreases in peak expiratory
flow rates (PEFR) than adults. The candidate RfC determination for this study focused
on the results in the children, among which asthmatics were over-represented (roughly 3-
times) compared to the national average.
¦ Garrett et al. (1999) conducted a cross-sectional survey of allergy and asthma-like
symptoms in children with or without a doctor's diagnosis of asthma. The study was
designed to include a high proportion of asthmatic children, a sensitive population for the
effects being studied.
EPA notes, however, that, while a human variability UF of 3 rather than 10 was used to
attempt to account for certain special attributes of these studies/effects, there is still uncertainty
about how much of the overall population heterogeneity is actually reflected even in these
relatively diverse residential studies.
Subchronic-to-Chronic Extrapolation
RfCs are intended to apply to chronic lifetime exposures. If a study is subchronic
(typically less than 10% of lifetime), an UF for subchronic-to-chronic extrapolation is generally
applied to the candidate RfC for that study. For the human residential and occupational studies
comprising the key studies for the RfC in this assessment, the average durations of exposure in
the households or workplaces under study is unknown. In this assessment, these studies were
considered chronic in nature and no subchronic-to-chronic UF was applied. However, there is
uncertainty about whether or not the responses observed fully reflected the potential effects of
This document is a draft for review purposes only and does not constitute Agency policy.
6-31 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
chronic exposure, especially for effects in children, where effects on the developing respiratory
and immune systems, for example, could be predisposing the children to further health effects
later in life.
6.2.1.5 Conclusions
Seven different non-cancer health effects were identified from formaldehyde inhalation
exposure studies, including: 1) sensory irritation of the eyes, nose, and throat, 2) upper
respiratory tract pathology, 3) pulmonary function, 4) asthma and atopy, 5) neurologic and
behavioral toxicity, 6) reproductive and developmental toxicity, and 7) immunological toxicity.
Of note, epidemiological evidence is available for most of these noncancer effects. EPA has
derived candidate RfCs for critical effects based on seven key studies. Three co-critical studies
were selected which provide similar cRfCs for related adverse health effects observed in
residential populations including children i.e., increased asthma incidence, decreased pulmonary
function, increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al.,
2002; Garrett et al., 1999; Krzyzanowski et al., 1999). The resulting cRfCs fall in a range
between 2.8 and 11 ppb, depending on the study, or endpoints considered, and the application of
alternative uncertainty factors for human variability (Table 6-1). The RfC is taken as the
average of the cRfCs from the three co-critical studies (See Section 6.2.1.2).
EPA has assessed the adequacy of the overall database for RfC derivation, and although
the database is quite large, and provides significant information on well studied POE effects.
There are remaining uncertainties in the database. Most notably, there is a need for additional
exposure-response information for observed neurotoxic effects, reproductive and developmental
effects as well as a two-generation study to evaluate the effects of formaldehyde exposure on
reproductive and developmental endpoints. EPA is considering 4 options to address database
uncertainties in the final RfC (Section 6.2.1.3). It is unclear what uncertainty factors are
appropriate to account for human variability and deficiencies in the overall database. For this
reason, several alternatives have been presented. EPA is seeking advice from the NAS and the
public on this matter.
6.2.2. Cancer Risk Estimates
6.2.2.1. Choice of Data
As explained above, the human epidemiologic data and the animal bioassay data indicate
multiple sites of concern, remote as well as at the portal of entry. The quantitative cancer risk
derivations in this document consider the risks of lymphohematopoietic cancers and solid
cancers of the respiratory tract. When adequate human data are available, as is the case with
This document is a draft for review purposes only and does not constitute Agency policy.
6-32 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
formaldehyde, it is generally preferable to base cancer risk estimates on the human data rather
than on data from experimental animals because of the inherent uncertainties associated with
interspecies extrapolation. Sufficient exposure-response data from a large, high-quality
epidemiologic study for the quantitative estimation of risk were available for some
lymphohematopoietic cancers and for nasopharyngeal cancer.15 Risk estimates based on nasal
tumors in rats were also derived for comparison with the estimates based on human data. The
data used for the quantitative risk assessment are as follows:
1. Nasopharyngeal cancer (NPC): The dose-response modeling of NPCs is based on results
from a large NCI cohort study of over 25,000 workers in 10 U.S. plants producing or
using formaldehyde (Hauptmann et al., 2004).
2. Lymphohematopoietic cancers: The dose-response modeling of select
lymphohematopoietic cancers is based on results from a more recent follow-up study (of
lymphohematopoietic malignancies only) of the same NCI cohort (Beane Freeman et al.,
2009).
3. Squamous cell carcinoma (SCC) in the upper and lower respiratory tract: An increased
incidence of nasal SCC was seen in two large long-term bioassays using F344 rats (Kerns
et al., 1983; Monticello et al., 1996). Although other studies in laboratory animals exist,
4. these two studies, when combined, provided the most robust data for analyses. The nasal
tumor incidence data from these rat bioassays is used for extrapolating the risk of SCC to
the entire human respiratory tract.16
6.2.2.2. Analysis of Epidemiologic Data
The NCI cohort consisted of 25,619 workers employed in 10 plants prior to 1966. A
follow-up through 1994 presented exposure-response analyses for 9 NPC deaths, as well as
15 Only one other epidemiological study was available with quantitative exposure estimates for the individual
workers. It was a much smaller study (it focused on one of the ten plants covered in the selected study), and it
evaluated only pharyngeal cancers.
16 That is, we do not assume site concordance between rat and human. This is reasonable because the respiratory
and transitional cell types considered to be at risk of SCC in the upper respiratory tract are also prevalent in the
lower human respiratory tract. Greater fractional penetration of formaldehyde is thought to occur posteriorly in the
human respiratory tract compared to the rat (Kimbell et al. 2001, Overton et al. 2001). Furthermore, some
epidemiological studies reported an increase in lung cancer with formaldehyde exposure (Gardner et al. 1993, Blair
et al. 1990, 1986), and lesions were seen in the lower respiratory tract of rhesus monkeys exposed to formaldehyde
(Monticello et al. 1989).
This document is a draft for review purposes only and does not constitute Agency policy.
6-33 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
analyses of deaths from other solid cancers (Hauptmann et al., 2004). The most recent follow-up
(through 2004; lymphohematopoietic cancers only) analyzed 319 deaths attributed to
lymphohematopoietic malignancy from a total of 13,951 deaths (Beane Freeman et al., 2009). A
detailed exposure assessment was conducted for each worker, based on exposure estimates for
different jobs held and tasks performed (Stewart et al., 1986). Exposure estimates were made
using several different metrics—peak exposure, average intensity, cumulative exposure, and
duration of exposure. Respirator use and exposures to formaldehyde-containing particulates and
other chemicals were also considered. Relative Risks (RRs) were estimated using log-linear
Poisson regression models stratified by calendar year, age, sex, and race and adjusted for pay
category (salary/wage/unknown). The NCI investigators used the low-exposure category as the
reference category to "minimize the impact of any unmeasured confounding variables since
nonexposed workers may differ from exposed workers with respect to socioeconomic
characteristics."
Although other upper respiratory tract cancers were also identified as being causally
associated with formaldehyde exposure in the weight-of-evidence analysis in section 4.5, NPC
was the only upper respiratory tract cancer with exposure-response data adequate for the
derivation of unit risk estimates in the Hauptmann et al. (2004) follow-up study of solid tumors.
Similarly, the weight-of evidence analysis in section 4.5 concluded that there were causal
relationships between formaldehyde exposure and all lymphohematopoietic cancers as a group as
well as leukemias as a group (with the strongest evidence for myeloid leukemia). However,
from the Beane Freeman et al. (2009) follow-up study of lymphohematopoietic malignancies,
only all leukemias combined and Hodgkin lymphoma were judged to have exposure-response
data adequate for the derivation of unit risk estimates.
For the NPCs, significant trends were observed for the cumulative and peak exposure
metrics. The cumulative exposure metric provides a good fit to the data (p trend = 0.029 for all
person-years). Since this is generally the preferred metric for quantitative risk assessment for
environmental exposure to carcinogens, cumulative exposure is chosen as the exposure metric
for the risk estimate calculations for NPC in this assessment. For the latency of solid cancers,
including nasopharyngeal tumors, a 15-year lag interval was used by Hauptmann et al. (2004).
For the lymphohematopoietic cancers, using the peak exposure metric, statistically
significant log-linear trends were observed for all lymphohematopoietic cancers, Hodgkin
lymphoma, and leukemia (the latter only when the unexposed person-years were included)
(Beane Freeman et al., 2009). Using the average exposure metric, there was a significant trend
for Hodgkin lymphoma. Similar results were seen with the cumulative exposure metric,
although the trends were only of borderline significance (Hodgkin lymphoma p trends = 0.06
This document is a draft for review purposes only and does not constitute Agency policy.
6-34 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
and 0.08 with and without the unexposed person-years, respectively; leukemiap trends = 0.08
and 0.12 with and without the unexposed person-years, respectively). For the latency of
lymphohematopoietic cancers, a 2-year lag interval was used by Beane Freeman et al. (2009).
Although the peak exposure metric provides the most statistically robust dose-response
relationship, it is not clear how to extrapolate RR estimates based on the peak exposure estimates
to meaningful estimates of lifetime extra risk of cancer from environmental exposures. The
average exposure metric is also problematic because it suggests that duration of exposure is not
important, i.e., exposure to a given exposure level for one year conveys the same amount of risk
as exposure to the same level for 70 years.
Cumulative exposure is generally the preferred metric for quantitative risk assessment for
environmental exposure to carcinogens. Given the consistency of increased mortality from
Hodgkin lymphoma and leukemia overall (exposed versus unexposed) and for each exposure
metric (Table 5-12), indicating risk from these cancers is more than chance, a determination was
made that the cumulative exposure results for these two cancer types constituted the best data
sets from which to calculate unit risk estimates for lymphohematopoietic cancers from the NCI
cohort.
Regression coefficients from the NCI log-linear trend test models for the NPCs
(Hauptmann et al., 2004) and the various lymphohematopoietic cancers (Beane Freeman et al.,
2009) were provided by Drs. Hauptmann and Beane Freeman, respectively. These trend tests
were of the form RR = eP*exposure xhe coefficients (i.e., P) were used in lifetable analyses to
calculate lifetime extra cancer risks from formaldehyde exposure (Section 5.2). Extra risk
estimates for cancer incidence for the three cancer types were approximated by assuming that
cancer incidence and cancer mortality have the same dose-response relationships and then using
background cause-specific incidence rates instead of mortality rates in the lifetable analysis.
Points of departure (PODs) based on the dose-response modeling of these cancers were
calculated as the exposure concentration at which the 95% upper confidence bound on extra risk
was 0.0005 (i.e., 0.05%) forNPC and for Hodgkin lymphoma and 0.005 (i.e., 0.5%) for
leukemia (Sections 5.2.2 and 5.2.3). These values approximate the lower confidence bounds on
dose at these extra risk levels. The values for these extra risk levels, 0.0005 and 0.005, were
chosen because they are near the lower end of the observable range of the data. Having such low
response levels associated with the points of departure is warranted because of the low
background lifetime risks for these cancer types (e.g., 0.00022 for NPC mortality). Higher extra
risk levels would entail extrapolation above the range of the bulk of the observable data to obtain
PODs. The resulting effective concentration values for the selected extra risk values for cancer
incidence are presented in Table 6-2.
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 5 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Linear low-dose extrapolation from the PODs was used to derive unit risk estimates for
NPC, Hodgkin lymphoma, and leukemia, as discussed in Section 6.2.2.4. To obtain an
approximate (upper bound) unit risk estimate of the total cancer risk from formaldehyde
exposure, risk estimates for these three cancer types (NPC, Hodgkin lymphoma, and leukemia)
were combined assuming a normal distribution (Section 5.2.4). This was considered the most
reasonable approach for estimating total cancer risk from the available data; however, it should
be noted that this estimate may not reflect all of the cancer types associated with formaldehyde
exposure.
Table 6-2: Effective concentrations (lifetime continuous exposure levels)
predicted for specified extra cancer risk levels for selected formaldehyde-
related cancers1
Cancer Type
Extra Risk Level
EC2 (ppm)
LEC3 (ppm)
NPC
0.0005
0.074
0.046
Hodgkin
lymphoma
0.0005
0.052
0.030
Leukemias
0.005
0.16
0.088
1. calculated including all person-years (see section 5.2)
2. effective concentration.
3. 95% lower confidence bound on the EC; this value is the POD.
6.2.2.3. Analysis of Laboratory Animal Data
Various bioassays have been conducted studying the effects of formaldehyde on rats,
mice, and rhesus monkeys and have been discussed at length earlier in this document. Of these,
two inhalation bioassays of rats, when combined, allow for the most robust characterization of
the long-term dose-response relationship in a laboratory species. These long-term bioassays
found an increased incidence of nasal SCCs in rats exposed to formaldehyde by the inhalation
route (Monticello et al., 1996; Kerns et al., 1983). In the combined data, rats were exposed to 0,
0.7, 2.0, 6.0, 9.93, and 14.96 ppm (0, 0.86, 2.5, 7.4, 12.2, and 18.4 mg/m3) exposure
concentrations of formaldehyde (Monticello et al. 1996; Kerns et al. 1983). SCCs were observed
only at 6 ppm and higher exposure concentrations.
A large amount of mechanistic information relevant to the dose-response relationship of
formaldehyde in the respiratory tract has been generated either following or in conjunction with
these two bioassays, as reviewed in Chapter 3, 4 and 5. This information includes the following:
This document is a draft for review purposes only and does not constitute Agency policy.
6-36 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
1. Measurements of DNA-protein cross-links (DPXs) formed by formaldehyde in F344 rats
and rhesus monkeys (Casanova et al., 1989, 1994). Several PBPK models have been
developed in the literature based on these data. Some of these efforts integrated the data
in both species (Casanova et al., 1991; Conolly et al., 2000; Klein et al., 2009).
2. Measurements of cell proliferation in F344 rats and rhesus monkeys (Monticello et al.,
1989, 1990, 1991, 1996).
3. Simulations of airflow in anatomically realistic representations of the upper respiratory
tract of the F344 rat, rhesus monkey and human, and in an idealized representation of the
human lower respiratory tract, using computer and physical models (Kimbell et al., 1993;
Kepler et al., 1998; Subramaniam et al., 1998). These simulations were used to predict
regional formaldehyde dosimetry in the corresponding sections of the respiratory tract of
these three species (Kimbell et al., 2001a, b; Overton et al., 2001).
The combined nasal tumor incidence data were analyzed using a multistage-weibull time-
to-tumor approach as well as biologically based dose-response (BBDR) models derived from
Conolly et al. (2003) [see Crump et al. (2005), Subramaniam et al. (2007), and Appendix E for
details]. The BBDR approach enabled integration of the mechanistic information and the time-
to-tumor incidence data within a single conceptual framework.
6.2.2.4. Extrapolation Approaches
An EPA inhalation unit risk is developed to estimate cancer risk from environmental
exposures or in order to determine exposure levels corresponding with cancer risks as low as 1
excess cancer in 10,000 or 1 excess cancer in 1 million. As neither data from animal studies, nor
human epidemiological studies, provide direct observation of these low level risks, the observed
exposure response relationship is extrapolated to estimate low dose risk. The model used to
extrapolate below the range of exposures clearly associated with increased risk of health effects
has a great influence on the inhalation unit risk, as there may be several orders of magnitude
difference between the observed risk and the target risk range. In the absence of empirical data
or a biologically-informed model, the EPA applies a simple straight line extrapolation from the
point of departure to zero exposure (U.S. EPA, 2005a). The Mode of Action evaluation reviews
available data and determines if an MOA can be sufficiently established and whether it informs
the shape of the exposure-response relationship.
This document is a draft for review purposes only and does not constitute Agency policy.
6-37 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Low-dose extrapolation for Lymphohematopoietic cancers:
Formaldehyde is a mutagen, and known to act directly on cells at the site of first contact.
Clastogenic effects have been documented in formaldehyde-exposed workers including
peripheral blood lymphocytes and circulating stem cells (Zhang et al., 2010). Thus a mutagenic
MOA has been hypothesized for lymphohematopoeitic cancers, and supports a linear low-dose
extrapolation of human cancer risk. Additionally, formaldehyde may also induce some form of
bone marrow toxicity, as suggested by observed pancytopenia in exposed workers (Tang et al.,
2008, Zhang et al., 2010). However, as the mechanism of transport to the bone marrow, and
biological activity leading to the observed toxicity are unknown, this information does not
inform the low-dose extrapolation. Although the mechanisms underlying formaldehyde-induced
leukemia and lymphoma are still largely speculative, there is little doubt of an association
between formaldehyde exposures and lymphohematopoeitic cancer mortality, especially for
myeloid leukemia. Therefore, without a known MOA which would justify an alternative
approach, and with a hypothesized mutagenic MOA under consideration which supports a simple
straight line extrapolation from the point of departure to zero risk at zero exposure, this is
applied when estimating human cancer risk from both leukemia and Hodgkins lymphoma from
the NCI cohort.
Low-dose extrapolation for cancer of the upper respiratory tract:
There are multiple plausible MO As for formaldehyde carcinogenesis regarding upper
respiratory tract cancers (Section 4.5.3), however they are not all applicable to the lower end of
the exposure response curve. For example, although regenerative cell proliferation associated
with focal and gross tissue lesions due to cell death may contribute to the high incidence of rat
nasal tumor in F344 rats, these mechanisms may not be operative in the low exposure region
expected for human environmental exposure (e.g. less than lppm) and therefore may not inform
low-dose extrapolation. There are MO As which are more appropriate to the low-dose region.
Specifically, formaldehyde is a known mutagen, may inhibit DNA repair activity and may have
additional activity as a tumor promoter. Finally, other affects such as formaldehyde-induced cell
proliferation, imunosuppression and disruption of the mucociliary apparatus may influence both
the level of tissue damage and ultimately cancer incidence.
EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a) recommend using
biologically based dose-response (BBDR) models for extrapolation when data permit. Conolly
et al. (2003, 2004) developed BBDR models to predict squamous cell carcinoma risk in the rat
This document is a draft for review purposes only and does not constitute Agency policy.
6-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
and human respiratory tract at exposures well below the range of the observed animal data.17
The primary conclusion from their modeling effort was that human exposure standards
protective of effects of formaldehyde-induced cytotoxicity should be sufficient to protect from
the potential carcinogenic effects of formaldehyde. The authors assessed that such a conclusion
was
conservative in the face of model uncertainties.18 This assessment's evaluation of the BBDR
models and alternative implementations (detailed in Section 5.3) finds that these models may not
provide conservative estimates of human cancer risk below the range of observed data.
The current assessment evaluated the uncertainties in the above BBDR model extensively
and examined alternative parameterizations of the modeling in Conolly et al. (2003, 2004).
These alternatives and the original model were equally consistent with the experimental data but
resulted in maximum likelihood estimates of added human risk that ranged from negative to
large positive values at environmental exposure concentrations. Model uncertainty far exceeded
statistical uncertainty (Table E-4 in Appendix E). Each of these models, including the modeling
in Conolly et al.,
1. was judged to be just as biologically plausible,
2. described the rat tumor incidence data equally well,
3. was based on different characterizations of the same empirical cell kinetic data, and
4. was based on the same empirical data on DPX measurements.
This assessment's evaluation19 of the above models (detailed in Section 5.3) concluded
that these models, including alternative implementations of those in Conolly et al. (2003, 2004),
were too uncertain to be useful for low-dose extrapolation of risk. It may be noted that the
sensitivity analyses on the basis of which these conclusions were reached have been criticized as
resulting in implausible risk estimates (given the epidemiologic data) as a consequence of
implementing model variations that are not biologically reasonable (Conolly et al. 2009). This
17 In that sense, the authors used the modeling as if it were a BBDR model even though they termed it as
"biologically-motivated".
18 Based on their modeling, Conolly et al. (2003, 2004) concluded that the directly mutagenic action of formaldehyde
does not play a significant role in formaldehyde carcinogenicity. Respiratory cancer risks associated with inhaled
formaldehyde were predicted to be de minimis (10 6 or less) at relevant human exposure levels when an upper bound
on the model estimate for the directly mutagenic action of formaldehyde was used.
19 The scope of this evaluation was informed by views provided by several experts convened by EPA in October
2004. The participants were Drs. Rory Conolly, Kenny Crump, Linda Hanna, Dale Hattis, Julia Kimbell, George
Lucier, Christopher Portier and Fred Miller (guest participant). The meeting agenda and summary are provided in
Appendix H.
This document is a draft for review purposes only and does not constitute Agency policy.
6-39 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
criticism was rebutted by Crump et al. (2009) on biological and epidemiological grounds. These
debates have been discussed fully in Appendix F. In particular, the assessment concludes that:
• When used for the purpose of extrapolating risk, the BBDR models did not appear to
reasonably constrain either
> risk estimates extrapolated from the F344 rat to the human, regardless of whether the
extrapolation was carried out at low or comparable exposures, or
> risk estimates for the F344 rat when extrapolated outside the range of observable
data.
• Furthermore, human risk calculated from these BBDR models was numerically unstable
when certain parameter conditions were realized (Section 5.3.3 and Appendix F).
• Therefore, clonal growth modeling was not found to be a useful approach for human
extrapolation of rodent risk estimates. The current assessment concludes that the result in
Conolly et al. cannot be considered to be "conservative in the face of model
uncertainties."
However, using the BBDR model to characterize the dose-response in the range of the
available data was judged to have the advantage of utilizing the available biological and
dosimetry data on formaldehyde in an integrated manner as well as providing statistically sound
descriptions of the empirical tumor incidence data. Therefore, this assessment uses the BBDR
modeling of the rat data to derive multiple PODs (for SCC in the respiratory tract) in the range of
the observed data and uses model-derived internal dose estimates. For the reasons detailed
above, the BBDR modeling is not used to extrapolate far below the observed data.
The lowest observed incidence of SCC in the bioassays used in the dose-response
assessment was equal to 0.0087 (at 6 ppm exposure). In addition, the BBDR modeling used data
on cell proliferation and formation of DPXs that informed the modeling of the tumor data at the
lower exposure concentrations of 0.7 and 2.0 ppm. Thus, the available data supported estimation
of response levels below the 10% response level commonly used in BMD analyses of tumor
data. Therefore, points of departure corresponding to 95% statistical upper bound levels of extra
risk of 0.005, 0.01 and 0.1 were estimated when the BBDR modeling was used.
Summary:
As discussed earlier in the hazard characterization, formaldehyde is a direct-acting
mutagen, and its genotoxic effects have been observed following human occupational
This document is a draft for review purposes only and does not constitute Agency policy.
6-40 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
exposures.20 Furthermore, a low-dose nonlinear MO A for formaldehyde-induced
lymphohematopoietic cancers, NPCs, or cancers in other regions of the respiratory tract has not
been established. In particular, the formation of DPXs by formaldehyde, considered a dose
surrogate for the molecular dose associated with formaldehyde's mutagenic action, has been
observed at doses well below those considered cytotoxic. Therefore, linear low-dose
extrapolation from the suitably chosen PODs was considered most appropriate for all the cancers
(whether the PODs were based on epidemiological data or rodent bioassay data), which is also in
accordance with EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a).
6.2.2.5. Inhalation Unit Risk Estimates for Cancer
The epidemiological and rodent inhalation data indicate multiple sites of concern. Unit
risk estimates calculated separately from these data are summarized in Table 6-3.
As can be seen in the Table 6-3, the unit risk estimate based on human data for NPC is in
the range of the estimates calculated for respiratory tract cancer from the rodent nasal cancer
data. Experimental animal data were inadequate for estimating risk of lymphohematopoietic
cancers. The unit risk estimate for Hodgkin lymphoma is also in the same range, while the unit
risk estimate for leukemia and the total cancer unit risk estimate are up to 4-fold higher.
As documented in EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a),
when high-quality human data are available, they are generally preferred over laboratory animal
data for quantitative risk assessment. Thus, the preferred (plausible upper bound) unit risk
estimate in this assessment is the value of 8.1 x 10"2 per ppm (6.6 x 10~5 per jig/m3) based on
(adult) human data for NPC, Hodgkin lymphoma, and leukemia. Note that, as discussed in
Section 6.2.2.6 below, if there is early-life exposure, the age-dependent adjustment factors
(ADAFs) should be applied, in accordance with EPA's Supplemental Guidance for Assessing
Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b).
6.2.2.6. Early-Life Susceptibility
There are no chemical-specific data for quantitatively addressing the susceptibility of
different life stages to carcinogenicity from inhalation exposure to formaldehyde. As
documented in section 4.5, formaldehyde is a mutagenic carcinogen and the weight of evidence
suggests that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
MOA. Therefore, increased early-life susceptibility should be assumed and, if there is early-life
20 While formaldehyde may also contribute to mutations indirectly, such an effect is likely to be relevant only at the
higher doses.
This document is a draft for review purposes only and does not constitute Agency policy.
6-41 DRAFT—DO NOT CITE OR QUOTE
-------
1 exposure, the ADAFs should be applied, in accordance with EPA's Supplemental Guidance for
2 Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA, 2005b). See
3 Section 5.4.4 for details on the application of the ADAFs.
4
5 Table 6-3 Inhalation unit risk estimates based on epidemiological and
6 experimental animal data
7
Cancer typea
Dose metric
Unit Risk Estimateb
(ppm1)
Based on Epidemiological Data
Nasopharyngeal
Cumulative exposure
0.011
Hodgkin lymphoma
Cumulative exposure
0.017
Leukemia
Cumulative exposure
0.057
All three cancer sites combined:
0.081°
Based on Experimental Animal Data
Squamous cell carcinoma
of the respiratory tract
Local dose (flux) of
formaldehyde in
pmol/mm2/hour
0.011 - 0.022d
8
9 a the unit risk estimates are all for cancer incidence.
10 bthese unit risk estimates do not include ADAFs (see Section 6.2.2.6 below).
11 0 this total cancer unit risk estimate is an estimate of the upper bound on the sum of risk estimates calculated
12 for the 3 individual cancer types (nasopharyngeal cancer, Hodgkin lymphoma, and leukemia); it is not the
13 sum of the individual (upper bound) unit risk estimates (see Section 5.2.4).
14 d values are similar to estimates from Schlosser et al. (2003). These authors determined their PODs based
15 on tumor and cell proliferation as endpoints, and extrapolated benchmark exposure concentrations to
16 humans using formaldehyde flux to the tissue and DPX concentrations as internal dose metrics.
17
18
19 Accordingly, for full lifetime exposures, the overall (plausible upper bound) unit risk
20 estimate is 0.13 per ppm (1.1 x 10~4 per jig/m3) for the three cancer types (NPC, Hodgkin
21 lymphoma, and leukemia) combined (see Table 5-26 for calculations).
22
23 6.2.2.7. Uncertainties in the Quantitative Risk Estimates
24 Uncertainties in the risk estimates based on the human data are discussed in detail in
25 Sections 5.2.2.4 and 5.2.3.4. Major uncertainties inherent in the NPC, Hodgkin lymphoma, and
26 leukemia risk estimates are
27
This document is a draft for review purposes only and does not constitute Agency policy.
6-42 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
¦ the retrospective exposure estimation,
¦ the appropriateness of the dose-response model and exposure metric, and
¦ the extrapolation from occupational exposures to lower environmental exposures.
In addition, the NPC and Hodgkin lymphoma estimates are limited by the sparse data for these
cancers in the NCI cohort study (estimates are based on the exposure-response modeling of only
9 NPC deaths and 27 Hodgkin lymphoma deaths).
Of note, Marsh et al. (2002, 1996) independently studied one of the 10 plants that was in
the NCI study, and there were large differences in the exposure estimates for that plant from the
two different studies. If the exposure estimates of Marsh et al. (2002) are closer to the true
exposures, then the potency of formaldehyde could be greater than reflected in the risk estimates
derived from the NCI data.
The linear low-dose extrapolation (see Section 6.2.2.4) from the 95% lower bound on the
exposure level associated with the benchmark response is generally considered to provide a
plausible upper bound on the risk at lower exposure levels. The strong association with peak
exposures for all 3 cancer types in the NCI study suggests that dose-rate effects may be operative
(i.e., the risk from peak occupational exposures may be greater than the [linearly] proportional
risks from lower exposures and, similarly, the risk from an occupational cumulative exposure
may be greater than the proportional risk from a lower environmental cumulative exposure).21
Any such dose-rate effects would not be reflected in the linear low-dose extrapolation approach
used in this assessment. Actual low-dose risks may be lower to an unknown extent.
Other significant uncertainties may also remain. For example, risk estimates could not be
derived from the NCI cohort study for rare upper respiratory tract cancers other than NPC. In
addition, although unit risk estimates were derived for Hodgkin lymphoma and leukemia because
they exhibited the strongest trend results of the lymphohematopoietic cancers using the
cumulative exposure metric, it is uncertain which specific lymphohematopoietic cancer subtypes
are associated with formaldehyde exposure. Furthermore, the potential role of particulates in the
NPC risk is unclear. Moreover, as for all occupational epidemiology studies, there is uncertainty
in extrapolating risk from an adult worker population (in this case predominantly white males) to
the more diverse general population.
21 Dose-rate effects are also suggested by the very steep, nonlinear exposure-response relationships observed in the
rodent cancer bioassays, although, in the rodents, this steep increase in tumor incidence at high exposures is thought
to be due to the contribution of cytotoxicity and regenerative proliferation, which is not apparent with the human
exposures (Section 4.5).
This document is a draft for review purposes only and does not constitute Agency policy.
6-43 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
Despite inevitable uncertainties, it is important not to lose sight of the strengths of the
estimates, which are based on human data from a high-quality NCI study. In addition to the use
of internal analyses and the extensive exposure assessment and consideration of potential
confounding or modifying variables, the NCI study has a large cohort that has been followed for
a long time. With the additional follow-up through 2004, reflected in the lymphohematopoietic
cancer results of Beane Freeman et al. (2009), the median duration of follow-up was 42 years,
and the 25,619 cohort members had accrued 998,106 person-years of follow-up.
Significant uncertainties also exist in the risk estimates derived from the rodent bioassay
data. In general, the difficulties in extrapolating from experimental animal bioassays are
considerable, and the use of human data is preferred, while recognizing the different
uncertainties that are present in risk estimates based on epidemiological data.
In the case of formaldehyde, this general uncertainty associated with extrapolation from
rodent data is increased due to the highly curvilinear nature of the dose-response relationships
associated with DPX formation, labeling index data, and tumor responses. The mechanistic
interpretation of these observed data has provided grounds for arguments in the literature that
formaldehyde tumorigenicity (at exposures > 6 ppm) should be uncoupled from its potential
carcinogenicity in the low-dose region.
Quantitative models have been used in the literature to further argue that the observed
risk in animal experiments is entirely due to cell proliferation induced by regenerative
hyperplasia in response to cell injury at cytotoxic doses, i.e., without a relevant role for the direct
mutagenic action of formaldehyde. In the context of using these data for quantitative risk
assessment, this document notes that such an inference of the data has been found to be
extremely uncertain. An analysis of the uncertainties in interpreting the available data has
shown that the directly mutagenic component could be important in explaining the high-dose
effect (Subramaniam et al., 2007).
While acknowledging these substantial difficulties, the quantitative dose-response
modeling of the rat data does allow inference about upper bound risks for respiratory cancer,
consistent with the observed experimental tumorigenicity. These upper bound risk estimates are
consistent with those estimated from the epidemiological data; however, such a consistency may
be entirely artifactual. As noted earlier, the BBDR modeling helped characterize some of the
uncertainty associated with extrapolating from the rodent data to the environmental risk in
people. The actual risk may be substantially lower or higher than the reasonable upper bound
risk estimated from the animal data.
This document is a draft for review purposes only and does not constitute Agency policy.
6-44 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
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
6.2.2.8 Conclusions
Cancer unit risk estimates for formaldehyde inhalation exposure were derived from both
human and laboratory animal data. As documented in EPA's Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 2005a), when high-quality human data are available, they are generally
preferred over laboratory animal data for quantitative risk assessment. Thus, the preferred unit
risk estimate in this assessment is based on human data for NPC, Hodgkin lymphoma, and
leukemia from a high-quality NCI occupational cohort study (Hauptmann et al., 2004; Beane
Freeman et al., 2009). (The qualitative hazard assessment suggests causal associations between
formaldehyde exposure and other cancer types as well [e.g., other upper respiratory tract cancers
and possibly other lymphohematopoietic cancers; see Section 4.5], but quantitative data from the
NCI cohort study were not amenable for deriving quantitative risk estimates for those cancer
types. Because there were not clear exposure-response data for these cancer types in that cohort
study [based on cumulative exposure], any contributions to the total cancer risk from
environmental formaldehyde exposure for these cancers are not expected to be large; however,
this is a source of uncertainty.)
The unit risk estimate for the total cancer incidence extra risk for these three cancer types
combined based on the (adult) human data is 8.1 x 10"2 per ppm (6.6 x 10"5 per ^ig/m3). As
documented in Section 4.5, formaldehyde is a mutagenic carcinogen and the weight of evidence
suggests that formaldehyde carcinogenicity can be attributed, at least in part, to a mutagenic
MOA. Therefore, as there are no chemical-specific inhalation data on cancer susceptibility at
different life-stages, increased early-life susceptibility is assumed and ADAFs should be applied
in accordance with EPA's Supplemental Guidance for Assessing Susceptibility from Early-Life
Exposure to Carcinogens (U.S. EPA, 2005b). Applying the ADAFs, the overall (upper bound)
unit risk estimate for full lifetime exposure is 0.13 per ppm (1.1 x 10 4 per |ig/m3) for the three
cancer types (NPC, Hodgkin lymphoma, and leukemia) combined. Using this lifetime unit risk
estimate, the upper bound estimate of the cancer risk at the RfC of 1 ppb is 1 x 10 4,
6.3. SUMMARY AND CONCLUSIONS
Seven different non-cancer health effects were identified from formaldehyde inhalation
exposure studies, including: 1) sensory irritation of the eyes, nose, and throat, 2) upper
respiratory tract pathology, 3) pulmonary function, 4) asthma and atopy, 5) neurologic and
behavioral toxicity, 6) reproductive and developmental toxicity, and 7) immunological toxicity.
Of note, epidemiological evidence is available for most of these noncancer effects. EPA has
derived candidate RfCs for critical effects based on seven key studies. Three co-critical studies
were selected which provide similar cRfCs for related adverse health effects observed in
This document is a draft for review purposes only and does not constitute Agency policy.
6-45 DRAFT—DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
residential populations including children i.e., increased asthma incidence, decreased pulmonary
function, increase in respiratory symptoms, and increased allergic sensitization (Rumchev et al.,
2002; Garrett et al., 1999; Krzyzanowski et al., 1999). The resulting cRfCs fall in a range
between 2.8 and 11 ppb, depending on the study, or endpoints considered, and the application of
alternative uncertainty factors for human variability (Table 6-1). The representative RfC for the
co-critical studies is taken as the average of the cRfCs (Section 6.2.1.2).
EPA has assessed the adequacy of the overall database for RfC derivation, and although
the database is quite large, and provides significant information on well studied POE effects.
There are remaining uncertainties in the database. Most notably, there is a need for additional
exposure-response information for observed neurotoxic effects, reproductive and developmental
effects as well as a two-generation study to evaluate the effects of formaldehyde exposure on
reproductive and developmental endpoints. EPA is considering 4 options to address database
uncertainties in the final RfC (Section 6.2.1.3). It is unclear what uncertainty factors are
appropriate to account for human variability and deficiencies in the overall database. For this
reason, several alternatives have been presented. EPA is seeking advice from the NAS and the
public on this matter.
Formaldehyde is carcinogenic to humans by the inhalation route of exposure. There is
sufficient evidence of a causal association between formaldehyde exposure and cancers of the
upper respiratory tracts, with the strongest evidence for nasopharyngeal and sino-nasal cancers.
There is also sufficient evidence of a causal association between formaldehyde exposure and
lymphohematopoietic cancers, with the strongest evidence for Hodgkin lymphoma and leukemia,
particularly myeloid leukemia. The (upper bound) unit risk estimate for the total cancer
incidence based on (adult) human data is 8.1 x 10"2 per ppm (6.6 x 10"5 per (J,g/m3). Applying
the age-dependent adjustment factors for increased early-life susceptibility, the overall combined
cancer unit risk estimate for full lifetime exposure is 0.13 per ppm (1.1 x 10 4 per |ig/m3).
This document is a draft for review purposes only and does not constitute Agency policy.
6-46 DRAFT—DO NOT CITE OR QUOTE
-------
REFERENCES
Abbas, AK; Fausto, N; Kumar, V. (2004) Robbins and Cotran pathologic basis of disease. W.B.
Saunders Co. Silver Spring, MD.
ACGIH (American Conference of Governmental Industrial Hygienists). (2002) Documentation
of the TLVs® and BEIs® with other worldwide occupational exposure values CD-ROM—2002.
American Conference of Governmental Industrial Hygienists, Cincinnati, OH.
Acheson, ED; Barnes, HR; Gardener, MJ; et al. (1984) Formaldehyde in the British chemical
industry: an occupational cohort study. Lancet 1:611-616.
Achkor, H; Diaz, M; Fernandez, MR; et al. (2003) Enhanced formaldehyde detoxification by
overexpression of glutathione-dependent formaldehyde dehydrogenase from Arabidopsis. Plant
Physiol 132:2248-2255.
Adams, DO; Hamilton, TA; Lauer, LD; et al. (1987) The effect of formaldehyde exposure upon
the mononuclear phagocyte system of mice. Toxicol Appl Pharmacol 88:165-174.
Ahmad, I; Whitson, TC. (1973) Formaldehyde: how much of a hazard? Ind Med Surg 42:26-27.
Ahmed, S; Tsukahara, S; Tin-Tin-Tin-Shwe; et al. (2007) Effects of low-level formaldehyde
exposure on synaptic plasticity-related gene expression in the hippocampus of immunized mice.
J. Neuroimmunol 186:104-111.
Akbar-Khanzadeh, F; Mlynek, JS. (1997) Changes in respiratory function after one and three
hours of exposure to formaldehyde in non-smoking subjects. Occup Environ Med 54:296-300.
Akbar-Khanzadeh, F; Vaquerano, MU; Akbar-Khanzadeh, M; et al. (1994) Formaldehyde
exposure, acute pulmonary response, and exposure control options in a gross anatomy
laboratory. Am J Ind Med 26:61-75.
Alarie, Y. (1981) Toxicological evaluation of airborne chemical irritants and allergens using
respiratory reflex reactions. In: Leong, B, ed. Proceedings of the inhalation toxicology and
technology symposium. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.; pp. 207-231.
Albert, RE; Pasternack, BS; Shore, RE; et al. (1975) Mortality patterns among workers exposed
to chloromethyl ethers-a preliminary report. Environ Health Perspect 11:209-214.
Albert, RE; Sellakumar, AR; Laskin, S; et al. (1982) Gaseous formaldehyde and hydrogen
chloride induction of nasal cancer in the rat. J Natl Cancer Inst 68:597-603.
Alderson, T. (1985) Formaldehyde-induced mutagenesis: a novel mechanism for its action.
MutatRes 154:101-10.
Alexandersson, R; Hedenstierna, G. (1988) Respiratory hazards associated with exposure to
formaldehyde and solvents in acid-curing paints. Arch Environ Health 43:222-227.
Alexandersson, R; Hedenstierna, G. (1989) Pulmonary function in wood workers exposed to
formaldehyde: A prospective study. Arch Environ Health 44:5-11.
Alexandersson, R; Hedenstierna, G; Kolmodin-Hedman, B. (1982) Exposure to formaldehyde
effects on pulmonary function. Arch Environ Health 37:279-284.
This document is a draft for review purposes only and does not constitute Agency policy.
R. i DRAFT—DO NOT CITE OR QUOTE
-------
Almaguer, D; Blade, LM. (1990) Health hazard evaluation report: Buckeye Hills Career Center,
Rio Grande, Ohio. National Institute for Occupational Safety and Health, U.S. Department of
Health and Human Services, Cincinnati, OH; HETA 88-153-2072
Almaguer, D; Klein, M. (1991) Health hazard evaluation report: Northwest Vocational School,
Cincinnati, Ohio. National Institute for Occupational Safety and Health, U.S. Department of
Health and Human Services, Cincinnati, OH; HETA 89-170-2100
Amdur, MO. (1960) The response of guinea pigs to inhalation of formaldehyde and formic acid
alone and with a sodium chloride aerosol. Int J Air Pollut 3:201-220.
Ames, BN; Shigenaga, MK; Gold, LS. (1993) DNA lesions, inducible DNA repair, and cell
division: three key factors in mutagenesis and carcinogenesis. Environ Health Perspect
101(Suppl 5):35-44.
Andersen, I; Lundqvist, GR. (1974) Design and performance of an environmental chamber. Int J
Biometeor 14:402-405.
Andersen, I; M0lhave, L. (1983) Controlled human studies with formaldehyde. In: Gibson, JE;
ed. Formaldehyde toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 154-165.
Andersen, ME; Jarabek, AM. (2001) Nasal tissue dosimetry. Issues and approaches for
"Category 1" gases: a report on a meeting held in Research Triangle Park, NC, February 11-12,
1998. Inhal Toxicol 13:415-435.
Andersen, ME; Clewell III, HJ, Bermudez, E; et al. (2008) Genomic signatures and dose
dependent transitions in nasal epithelial responses to inhaled formaldehyde in the rat. Toxicol
Sci 105:368-383.
Andersen, I (1979) Indoor climate: Effects on human comfort, performance, and health in
residential, commercial, and light-industry buildings. In: Fanger, P.O. and Valbj0rn; ed.
Proceedings of the first international indoor climate symposium in Copenhagen Aguust 30-
September 1, 1978. Danish Building Research Institute; Copenhagen; pp. 65-87.
Anderson, H; Dally, KA; Hanrahan, L; et al. (1983) The epidemiology of mobile home
formaldehyde vapor concentration and residents' health status. Chicago, IL: Region V.
Available from NTIS, PB84-123504. EPA/905/1-83-001.
Andersson, M; Agurell, E; Vaghef, H, et al. (2003) Extended-term cultures of human T-
lymphocytes and the comet assay: a useful combination when testing for genotoxicity in vitro?
MutatRes 540:43-55.
Anderson, S; Jensen, O; Oliva, D. (1982) Formaldehyde exposure and lung cancer in Danish
doctors. UgeskrLaeg 144:1571-1573.
Andjelkovich, DA; Janszen, DB; Brown, MH; et al. (1995) Mortality of iron foundry workers:
IV. Analysis of a subcohort exposed to formaldehyde. J Occup Environ Med 37:826-837.
Ang, HL; Deltour, L; Zgombic-Knight, M; et al. (1996a) Expression patterns of class I and class
IV alcohol dehydrogenase genes in developing epithelia suggest a role for alcohol
dehydrogenase in local retinoic acid synthesis. Alcohol Clin Exp Res 20:1050-1064.
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT—DO NOT CITE OR QUOTE
-------
Ang, HL; Deltour, L; Hayamizu, TF; et al. (1996b) Retinoic acid synthesis in mouse embryos
during gastrulation and craniofacial development inked to class IV alcohol dehydrogenase gene
expression. J Biol Chem 271:9526-9534.
Apfelbach, R; Reibenspies, M. (1991) Olfactory learning is inhibited after low-level
formaldehyde gas exposure in the ferret. Chem Senses 16:498
Apfelbach, R; Weiler, E. (1991) Sensitivity to odors in Wistar rats is reduced after low-level
formaldehyde-gas exposure. Naturwissenschaften 78:221-223.
Apfelbach, R; Reibenspies, M; Schmidt, R. (1992) The effect of long-lasting formaldehyde gas
exposure on the olfactory epithelium and on the olfactory discrimination ability in the ferret.
Chem Senses 17:590-591.
Appelman, LM; Woutersen, RA; Zwart, A; et al. (1988) One-year inhalation toxicity study of
formaldehyde in male rats with a damaged or undamaged nasal mucosa. J Appl Toxicol 8:85-
90.
Armstrong, RW; Imrey, PB; Lye, MS; et al. (2000) Nasopharyngeal carcinoma in Malaysian
Chinese: occupational exposures to particles, formaldehyde and heat. Int J Epidemiol 29:991-8.
Arts, JH; Droge, SC; Spanhaak, S; et al. (1997) Local lymph node activation and IgE responses
in brown Norway and Wistar rats after dermal application of sensitizing and non-sensitizing
chemicals. Toxicology 117:229-234.
Ashby, J; Callander, RD; and Rose, FL. (1985) Weak mutagenicity to Salmonella of the
formaldehyde-releasing anti-tumour agent hexamethylmelamine. Mutat Res 142:121-5.
Asian, H; Songur, A; Tunc, AT; et al. (2006) Effects of formaldehyde exposure on granule cell
number and volume of dentate gyrus: A histopathological and stereological study. Brain Res
1122:191-200.
ASTM (American Society for Testing and Materials). (2000) Standard test method for estimating
sensory irritancy of airborne chemicals. Annual book of ASTM standards. Designation E981
(reapproved 2000). Philadelphia, PA: American Society for Testing of Materials.
ATC (American Thoracic Society). (2000) What constitutes an adverse health effect of air
pollution? Am J Respir Crit Care Med 161:665-673.
ATSDR (Agency for Toxic Substances and Disease Registry). (1999) Toxicological profile for
formaldehyde. Public Health Service, U.S. Department of Health and Human Services, Atlanta,
GA. Available online at http://www.atsdr.cdc.gov/toxprofiles/tpl 11 .html.
ATSDR, 2007. An update and revision of ATSDR's February 2007 health consultation:
Formaldehyde sampling of FEMA temporary-housing trailers; Baton Rouge, Louisiana
September-October, 2006. Public Health Service, U.S. Department of Health and Human
Services, Atlanta, GA. Available online at
www.atsdr.cdc.gov/substances/formaldehyde/public_assessment.html.
Auerbach, C; Moutschen-Dahmen, M; Moutschen, J. (1977) Genetic and cytogenetical effects of
formaldehyde and related compounds. Mutat Res 39:317-361.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT—DO NOT CITE OR QUOTE
-------
Axelsson, G; Lutz, C; Rylander, R. (1984) Exposure to solvents and outcome of pregnancy in
university laboratory employees. Br J Ind Med 41:305-312.
Baba, K; Yagi, T; Niwa, S; et al. (2000) Measurement of formaldehyde-specific IgE antibodies
in adult asthmatics. Arerugi 49(5):404-411.
Babiuk, C; Steinhagen, WH; Barrow, CS. (1985) Sensory irritation response to inhaled
aldehydes after formaldehyde pretreatment. Toxicol Appl Pharmacol 79:143-149.
Bach, B; Pedersen, OF; M0lhave, L. (1990) Human performance during experimental
formaldehyde exposure. Environ Int 16:105-114.
Baird, DD; Wilcox, AJ. (1985) Cigarette smoking associated with delayed conception. J Am
Med Assoc 253:2979-2983.
Baker, H; Kawano, T; Margolis, FL; et al. (1983) Transneuronal regulation of tyrosine
hydroxylase expression in olfactory bulb of mouse and rat. J Neurosci 3:69-78.
Ballarin, C; Sarto, F; Giacomelli, L; et al. (1992) Micronucleated cells in nasal mucosa of
formaldehyde-exposed workers. Mutat Res 280:1-7.
Barker, S; Weinfeld, M; Murray, D. (2005) DNA-protein crosslinks: Their induction, repair, and
biological consequences. Mutat Res 589:111-135.
Barrow, CS; Steinhagen, WH; Chang, JCF. (1983) Formaldehyde sensory irritation. In: Gibson,
JE; ed. Formaldehyde toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 16-
25.
Bartnik, FG; Gloxhuber, C; Zimmermann, V. (1985) Percutaneous absorption of formaldehyde
in rats. Toxicol Lett 25:167-172.
Basecke, J; Griesinger, F; Trumper, L; et al. (2002) Leukemia- and lymphoma-associated genetic
aberrations in healthy individuals. Ann Hematol 81:64-75.
Basler, A; v d Hude, W; and Scheutwinkel-Reich, M. (1985) Formaldehyde-induced sister
chromatid exchanges in vitro and the influence of the exogenous metabolizing systems S9 mix
and primary rat hepatocytes. Arch Toxicol 58:10-3.
Battelle Columbus Laboratories. (1981) Final report on a chronic inhalation toxicology study in
rats and mice exposed to formaldehyde. Prepared by Battelle Columbus Laboratories,
Columbus, OH for the Chemical Industry Institute of Toxicology (CUT), Research Triangle
Park, NC; CUT Docket # 10922.
Bauchinger, M; Schmid, E. (1985) Cytogenetic effects in lymphocytes of formaldehyde workers
of a paper factory. Mutat Res 158:195-199.
Beane-Freeman, LE; Blair, A; Lubin, JH; et al. (2009) Mortality from lymphohematopoietic
malignancies among workers in formaldehyde industries: the National Cancer Institute cohort. J
Natl Cancer Inst 101:751-761.
Beland, FA; Fullerton, NF; Heflich, RH. (1984) Rapid isolation, hydrolysis and chromatography
of formaldehyde-modified DNA. J Chromatogr 308:121-131.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT—DO NOT CITE OR QUOTE
-------
Bender, JR; Mullin, LS; Graepel, GJ; et al. (1983) Eye irritation response of humans to
formaldehyde. Am Ind Hyg Assoc J 44:463-465.
Benedetti, MS; Whomsley, R; Canning, M. (2007) Drug metabolism in the paediatric population
and in the elderly. Drug Discov Today 12:599-610.
Berke, JH. (1987) Cytologic examination of the nasal mucosa in formaldehyde-exposed workers.
J Occup Med 29:681-684.
Bermudez, E; Allen, PF. (1984) The assessment of DNA damage and repair in rat nasal
epithelial cells. Carcinogenesis 5:1453-1458.
Bermudez, E; Chen, Z; Gross, EA, et al. (1994) Characterization of cell lines derived from
formaldehyde-induced nasal tumors in rats. Mol Carcinog 9:193-199.
Bermudez, E; Delehanty, LL. (1986) The effects of in vitro formaldehyde treatment on the cells
of the rat nasal epithelium. Environ Mutagen 8 (Suppl.6): 11.
Bertazzi, PA; Pesatori, AC; Radice, L; et al. (1986) Exposure to formaldehyde and cancer
mortality in a cohort of workers producing resins. Scand J Work Environ Health 12:461-468.
Bertazzi, PA; Pesatori, A; Guercilena, S; et al. (1989) Carcinogenic risk for resin producers
exposed to formaldehyde: extension of follow-up. Med Lav 80:111-122.
Bhalla, DK; Mahavni, V; Nguyen, T; et al. (1991) Effects of acute exposure to formaldehyde on
surface morphology of nasal epithelia in rats. J Toxicol Environ Health 33:171-188.
Biagini, RE; Moorman, WJ; Knecht, EA; et al. (1989) Acute airway narrowing in monkeys from
challenge with 2.5 ppm formaldehyde generated from formalin. Arch Environ Health 44:12-17.
Bickis, M; Krewski, D. (1989) Statistical issues in the analysis of the long-term carcinogenicity
bioassay in small rodents: An empirical evaluation of statistical decision rules. Fundam Appl
Toxicol 12:202-221.
Binzak, BA; Vockley, JG; Jenkins, RB; et al. (2000) Structure and analysis of the human
dimethylglycine dehydrogenase gene. Mol Genet Metab 69:181-187.
Black, KA; Eells, JT; Noker, PE; et al. (1985) Role of hepatic tetrahydrofolate in the species
difference in methanol toxicity. Proc Natl Acad Sci USA 82:3854-3858.
Blackburn, GR; Dooley, JFI; Schreiner, CA; et al. (1991) Specific identification of
formaldehyde-mediated mutagenicity using the mouse lymphoma L5178Y TK +/- assay
supplemented with formaldehyde dehydrogenase. In Vitro Toxicol 4:121-132.
Blair, A; Stewart, P; O'Berg, M; et al. (1986) Mortality among industrial workers exposed to
formaldehyde. JNatl Cancer Inst 76:1071-1084.
Blair, A; Stewart, PA; Hoover, RN; et al. (1987) Cancers of the nasopharynx and oropharynx
and formaldehyde exposure. J Natl Cancer Inst 78:191-193.
Blair, A; Stewart, PA. (1990) Correlation between different measures of occupational exposure
to formaldehyde, Am JEpidemiol 131:510-516.
Blair, A; Stewart, PA; Hoover, RN. (1990) Mortality from lung cancer among workers employed
in formaldehyde industries. Am J Ind Med 17:683-699.
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT—DO NOT CITE OR QUOTE
-------
Blair, A; Linos, A; Stewart, A; et al. (1993) Evaluation of risks for Non-Hodgkin's Lymphoma
by occupation and industry exposures from a case-control study. Am J Ind Med 23:301-312.
Blasiak, J; Trzeciak, A; Malecka-Panas, E, et al. (2000) In vitro genotoxicity of ethanol and
acetaldehyde in human lymphocytes and the gastrointestinal tract mucosa cells. Toxicol In Vitro
14:287-295.
Blejer, HP; Miller, BH. (1966) Occupational health report of formaldehyde concentrations and
effects on workers at the Bayly Manufacturing Company, Visalia, California. Bureau of
Occupational Health, California Department of Public Health, Los Angeles, CA; Study No. S-
1806.
Bloom, B; Cohen, RA; Freeman, G. (2009) Summary health statistics for U.S. children: National
Health Interview Survey, 2008. National Center for Health Statistics. Vital Health Stat 10(244).
Boeniger, MF. (1987) Formate in urine as a biological indicator of formaldehyde exposure: A
review. Am Ind Hyg Assoc J 48:900-908.
Boffetta, P; Stellman, SD; Garfinkel, L. (1989) A case-control study of multiple myeloma nested
in the American Cancer Society prospective study. Int J Cancer 43:554-559.
Bogdanffy, MS; Morgan, PH; Starr, TB; et al. (1987) Binding of formaldehyde to human and rat
nasal mucus and bovine serum albumin. Toxicol Lett 38:145-154.
Bogdanffy, MS; Sarangapani, R; Plowchalk, DR; et al. (1999) A biologically based risk
assessment for vinyl acetate-induced cancer and noncancer inhalation toxicity. Toxicol Sci
51:19-35.
Bogdanffy, MS; Plowchalk, DR; Sarangapani, R, et al. (2001) Mode of action-based dosimeters
for interspecies extrapolation of vinyl acetate inhalation risk. Inhal Toxicol 13:377-396.
Boja, JW; Nielsen, JA; Foldvary, E; et al. (1985) Acute low-level formaldehyde behavioural and
neurochemical toxicity in the rat. Prog Neuro-Psychopharm 9:671-674.
Bolumar, F; Olsen, J; Boldsen, J. (1996) Smoking reduces fecundity: A European multicenter
study on fertility and subfecundity. The European study group on infertility and subfecundity.
Am JEpidemiol 143:578-587.
Bolumar, F; Olsen, J; Boldsen, J. (1997) Smoking reduces fecundity: A European multicenter
study on infertility and subfecundity. The European study group on infertility and subfecundity.
Am J Epidemiol 145:81-82.
Bond, G; Flores, G; Shellenberger, R. (1986) Nested case-control study of lung cancer among
chemical workers. Am J Epidemiol 124:53-66.
Boreiko, CJ; and Ragan, DL (1983). Formaldehyde effects in the C3H/1OTV2 cell transformation
assay. In: Gibson, JE,ed, Formaldehyde toxicity, Washington, DC: Hemisphere Publishing
Corporation, 63-71.
Bos, PMJ; Zwart, A; Reuzel, PGJ; et al. (1992) Evaluation of the sensory irritation test for the
assessment of occupational health risk. Crit Rev Toxicol 21:423-450.
Bosetti, C; McLaughlin, JK; Tarone, RE; et al. (2008) Formaldehyde and cancer risk: a
quantitative review of cohort studies through 2006. Annal Oncol 19:29-43.
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT—DO NOT CITE OR QUOTE
-------
Bosworth, D; Crofton-Sleigh, C; Venitt, S. (1987) A forward mutation assay using ampicillin-
resistance in Escherichia coli designed for investigating the mutagenicity of biological samples.
Mutagenesis 2:455-467.
Boy sen, M; Zadig, E; Digernes, V; et al. (1990) Nasal mucosa in workers exposed to
formaldehyde: a pilot study. Br J Ind Med 47:116-121.
Brenneman, KA; James, RA; Gross, EA; et al. (2000) Olfactory neuron loss in adult male CD
rats following subchronic inhalation exposure to hydrogen sulfide. Toxicol Pathol 28:326-333.
Brinton, LA; Blot, W; Becker, J; et al. (1984) A case-control study of cancers of the nasal cavity
and paranasal sinuses. Am J Epidemiol 119:896-906.
Brooks, SM. (1985) Occupational asthma. In Bronchial Asthma; Mechanism and Therapeutics.
Second edition. Eds Weiss, EB; Segal, MS; Stein, M. Little, Brown and Co, Boston/Toronto.
Brutlag, D; Schlehuber, C; Bonner, J. (1969) Properties of formaldehyde-treated nucleohistone.
Biochemistry 8:3214-3218.
Buckley, LA; Jiang, XZ; James, RA; et al. (1984) Respiratory tract lesions induced by sensory
irritants at the RD50 concentration. Toxicol Appl Pharmacol 74:417-429.
Bull, RJ. (2000) Mode of action of liver tumor induction by trichloroethylene and its
metabolites, trichloroacetate and dichloroacetate. Environ Health Perspect 108(Suppl 2):241-
259.
Burgaz, S; £akmak, G; Erdem, O; et al. (2001) Micronuclei frequencies in exfoliated nasal
mucosa cells from pathology and anatomy laboratory workers exposed to formaldehyde.
Neoplasma 48:144-147.
Burgaz, S; Erdem, O; £akmak, G; et al. (2002) Cytogenetic analysis of buccal cells from shoe-
workers and pathology and anatomy laboratory workers exposed to n-hexane, toluene, methyl
ethyl ketone and formaldehyde. Biomarkers 7:151-161.
Burge, PS; Harries, MG; Lam, WK; et al. (1985) Occupational asthma due to formaldehyde.
Thorax 40:255-260.
Burt, RD; Vaughan, TL; McKnight, B. (1992) Descriptive epidemiology and survival analysis of
nasopharyngeal carcinoma in the United States. Int J Cancer 52:549-556.
Buss, J; Kuschinsky, K; Kewitz, H; et al. (1964) Enteric resorption of formaldehyde. Naunyn
Schmiedebergs Arch Pharmacol 247:380-381.
Cain, H. (2001) Bronchoprovocation testing. Clin Chest Med 22:651-659.
Cain, W; See, L; Tosun, T. (1986) Irritation and odor from formaldehyde: chamber studies. In:
IAQ '86: Managing indoor air for health and energy conservation: proceedings of the American
Society of Heating, Air Conditioning Engineers Conference, Atlanta, GA; pp. 126-137.
CalEPA (2008) California OEHHA. Hot Spots Guidelines. Air Toxicology and Epidemiology.
Acute, 8-hour and Chronic Reference Exposure Levels. Appendix D.
http://oehha.ca.gov/air/hot_spots/2008/AppendixDl_final.pdf
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT—DO NOT CITE OR QUOTE
-------
Callas, PW; Pastides, H; Hosmer, DW, Jr. (1996) Lung cancer mortality among workers in
formaldehyde industries. J Occup Environ Med 38:747-748.
Cap, P; Dryahina, K; Pehal, F; et al. (2008) Selected ion flow tube mass spectrometry of exhaled
breath condensate headspace. Rapid Commun Mass Spectrom. 22(18):2844-2850.
Carraro, E; Gasparini, S; Petrini, T; et al. (1997) Immune response prevalence to formaldehyde-
human serum albumin molecular adduct in a healthy population. J Environ Pathol Toxicol
Oncol 16(2-3 ): 215—218.
Carson, S; Goldhamer, R; Weinberg, MS. (1966) Characterization of physical, chemical and
biological properties of mucus in the intact animal. Ann NY Acad Sci 130:935-943.
Casanova, M; Heck, HD. (1987) Further studies of the metabolic incorporation and covalent
binding of inhaled [3H]- and [14C]formaldehyde in Fischer-344 rats: effects of glutathione
depletion. Toxicol Appl Pharmacol 89:105-121.
Casanova, M; Heck, HD; Everitt, JI; et al. (1988) Formaldehyde concentrations in the blood of
rhesus monkeys after inhalation exposure. Food Chem Toxicol 26:715-716.
Casanova, M; Devo, DF; Heck, HD. (1989) Covalent binding of inhaled formaldehyde to DNA
in the nasal mucus of Fischer 344 rats: analysis of formaldehyde and DNA by high performance
liquid chromatography and provisional pharmacokinetic interpretation. Fundam Appl Toxicol
12:397-417.
Casanova, M; Morgan, KT; Steinhagen, WH; et al. (1991) Covalent binding of inhaled
formaldehyde to DNA in the respiratory tract of rhesus monkeys: pharmacokinetics, rat-to-
monkey interspecies scaling, and extrapolation to man. Fundam Appl Toxicol 17:409-428.
Casanova, M; Morgan, KT; Gross, EA; et al. (1994) DNA-protein cross-links and cell
replication at specific sites in the nose of F344 rats exposed subchronically to formaldehyde.
Fundam Appl Toxicol 23:525-536.
Casanova, M; Bell, DA; and Heck, HD. (1997) Dichloromethane metabolism to formaldehyde
and reaction of formaldehyde with nucleic acids in hepatocytes of rodents and humans with and
without glutathione S-transferase T1 and Ml genes. Fundam Appl Toxicol 37:168-180.
Casanova, M; and Heck, HA. (1997) Lack of evidence for the involvement of formaldehyde in
the hepatocarcinogenicity of methyl tertiary-butyl ether in CD-I mice. Chem Biol Interact
105:131-143.
Casanova, M; Cole, P; Collins, JJ; et al. (2004) Re: mortality from lymphohematopoietic
malignancies among workers in formaldehyde industries. J Natl Cancer Inst 96:966-967.
Casanova-Schmitz, M; Heck, HD. (1983) Effects of formaldehyde exposure on the extractability
of DNA from proteins in the rat nasal mucosa. Toxicol Appl Pharmacol 70:121-132.
Casanova-Schmitz, M; David, RM; Heck, HD. (1984a) Oxidation of formaldehyde and
acetaldehyde by NAD+-dependent dehydrogenases in rat nasal mucosal homogenates. Biochem
Pharmacol 33:1137-1142.
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT—DO NOT CITE OR QUOTE
-------
Casanova-Schmitz, M; Starr, TB; Heck, HD. (1984b) Differentiation between metabolic
incorporation and covalent binding in the labeling of macromolecules in the rat nasal mucosa
and bone marrow by inhaled [14C]- and [3H]-formaldehyde. Toxicol Appl Pharmacol 76:26-44.
Cascieri, T; Clary, JJ. (1992) Formaldehyde—oral toxicity assessment. Comments Toxicol
4:295-304.
Cassee, FR; Arts, JH; Groten, JP; et al. (1996) Sensory irritation to mixtures of formaldehyde,
acrolein, and acetaldehyde in rats. Arch Toxicol 70:329-337.
Cassee, FR; Feron, VJ. (1994) Biochemical and histopathological changes in nasal epithelium of
rats after 3-day intermittent exposure to formaldehyde and ozone alone or in combination.
Toxicol Lett 72:257-268.
Cassee, FR; Groten, JP; Feron, VJ. (1996) Changes in the nasal epithelium of rats exposed by
inhalation to mixtures of formaldehyde, acetaldehyde, and acrolein. Fundam Appl Toxicol
29:208-218.
Casset, A; Purohit, A; Marchand, C; et al. (2006) The bronchial response to inhaled
formaldehyde. Rev Mai Respir 23(1 Suppl):3S25-3S34.
Cassidy, SL; Dix, KM; Jenkins, T. (1983) Evaluation of a testicular sperm head counting
technique using rats exposed to dimethoxyethyl phthalate (DMEP), glycerol alpha-
monochlorohydrin (GMCH), epichlorohydrin (ECH), formaldehyde (FA), or methyl
methanesulphonate (MMS). Arch Toxicol 53:71-78.
Castonguay, A; Pepin, P; Briere, N. (1991) Modulation of 4-(methylnitrosamino)-l-(3-pyridyl)-l
butanone demethylation and denitrosation by rat liver microsomes. Cancer Lett 59:67-74.
CDC (U.S. Centers for Disease Control and Prevention). (2000) Measuring childhood asthma
prevalence before and after the 1997 redesign of the National Health Interview Survey — United
States. MMWR 49:908-911.
CDC. (2008) Final report on formaldehyde levels in FEMA-supplied travel trailers, park
models, and mobile homes. Available from
http://www.cdc.gOv/nceh/ehhe/trailerstudy/residents.htm#final.
Ceder, R; Merne, M; Staab, C; et al. (2007) The application of normal, SV40 T-antigen-
immortalized and tumour-derived oral keratinocytes, under serum-free conditions, to the study of
the probability of cancer progression as a result of environmental exposure to chemicals.
Alternatives to Laboratory Animals 35:621-639.
Cederbaum, Al; Qureshi, A. (1982) Role of catalase and hydroxyl radicals in the oxidation of
methanol by rat liver microsomes. Biochem Pharmacol 31:329-335.
Chang, JC; Barrow, CS. (1984) Sensory irritation tolerance and cross-tolerance in F-344 rats
exposed to chlorine or formaldehyde gas. Toxicol Appl Pharmacol 76:319-327.
Chang, JC; Steinhagen, WH; Barrow, CS. (1981) Effect of single or repeated formaldehyde
exposure on minute volume of B6C3F1 mice and F-344 rats. Toxicol Appl Pharmacol 61:451-
459.
This document is a draft for review purposes only and does not constitute Agency policy.
R-9 DRAFT—DO NOT CITE OR QUOTE
-------
Chang, JC; Gross, EA; Swenberg, JA; et al. (1983) Nasal cavity deposition, histopathology, and
cell proliferation after single or repeated formaldehyde exposures in B6C3F1 mice and F-344
rats. Toxicol Appl Pharmacol 68:161-176.
Chebotarev, AN; Titenko, NV; Selezneva, TG, et al. (1986) [Comparison of chromosome
aberrations, sister chromatid exchanges and unscheduled DNA synthesis in the evaluation of the
mutagenicity of environmental factors], Tsitol Genet 20:109-115.
Cheng, YS; Hansen, GK; Su, FY; et al. (1990) Deposition of ultrafine aerosols in rat nasal
molds. Toxicol Appl Pharmacol 106:222-233.
Cheng, G; Shi, Y; Sturla, SJ; et al. (2003) Reactions of formaldehyde plus acetaldehyde with
deoxyguanosine and DNA: formation of cyclic deoxyguanosine adducts and formaldehyde
cross-links. Chem Res Toxicol 16:145-152.
Cheng, XM; Zhao, J; Feng, B; et al. (2008) Relationship between susceptibility of formaldehyde
metabolism and genetic polymorphisms of ALD2 and cytochrome P4502E1. Zhonghua Yu Fang
YiXueZaZhi 42:582-587.
Cheng, G; Wang, M; Upadhyaya, P, et al. (2008) Formation of formaldehyde adducts in the
reactions of DNA and deoxyribonucleosides with alpha-acetates of 4-(methylnitrosamino)-l-(3-
pyridyl)-l-butanone (NNK), 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanol (NNAL), andN-
nitrosodimethylamine (NDMA). Chem Res Toxicol 21:746-751.
Chernoff, N; Kavlock, RJ. (1982) An in vivo teratology screen utilizing pregnant mice. J
Toxicol Environ Health 10:541-550.
Cho, JY; Min, N; Franzen, L; et al. (1996) Rapid down-regulation of tyrosine hydroxylase
expression in the olfactory bulb of naris-occluded adult rats. J Comp Neurol 369:264-276.
Chowdhury, AR; Gautam, AK; Patel, KG; et al. (1992) Steroidogenic inhibition in testicular
tissue of formaldehyde exposed rats. Ind J Physiol Pharmacol 36:162-168.
CUT (Chemical Industry Institute of Toxicology). (1982) Summary of final report on
formaldehyde study. CUT Activities 2:1, 9-10.
CUT. (1999) Formaldehyde: Hazard characterization and dose-response assessment for
carcinogenicity by the route of inhalation (Revised Edition). Research Triangle Park, NC:
Chemical Industry Institute of Toxicology.
Claeys, S; Cuvelier, C; Quatacker, J; et al. (1996a) Ultrastructural investigation of M-cells and
lymphoepithelial contacts in naso-pharangeal associated lymphoid tissue (NALT). Acta
Otolaryngeal Supple 523:40-42.
Claeys, S; Cuvelier, C; Van Cauwenberge, P. (1996b) Immunohistochemical analyses of the
lymphoepithelium in human nasopharyngeal associated lymphoid tissue (NALT) Acta
Otolaryngol Suppl 523:38-39.
Coggon, D; Pannett, B; Acheson, ED. (1984) Use of job-exposure matrix in an occupational
analysis of lung and bladder cancers on the basis of death certificates. J Natl Cancer Inst
72(l):61-65.
This document is a draft for review purposes only and does not constitute Agency policy.
R. i o DRAFT—DO NOT CITE OR QUOTE
-------
Coggon, D; Harris, EC; Poole, J; et al. (2003) Extended follow-up of a cohort of British
chemical workers exposed to formaldehyde. J Natl Cancer Inst 95:1608-1615.
Cohen-Hubal, EA; Schlosser, PM; Conolly, RB; et al. (1997) Comparison of inhaled
formaldehyde dosimetry predictions with DNA-protein cross-link measurements in the rat nasal
passages. Toxicol Appl Pharmacol 143:47-55.
Cole, P; Axten, C. (2004) Formaldehyde and leukemia: an improbable causal relationship.
Regul Toxicol Pharmacol 40:107-112.
Collins, JJ. (2004) Formaldehyde exposure and leukemia. Occup Environ Med 61:875-876.
Collins, JJ; Lineker, GA. (2004) A review and meta-analysis of formaldehyde exposure and
leukemia. Regul Toxicol Pharmacol 40:81-91.
Collins, JJ; Acquavella, JF; Esmen, NA. (1997) An updated analysis of formaldehyde exposure
and upper respiratory tract cancers. J Occup Environ Med 39: 639-651.
Collins, JJ; Esmen, NA; Hall, TA. (2001a) A review and meta-analysis of formaldehyde
exposure and pancreatic cancer. Am J Ind Med 39:336-345.
Collins, JJ; Ness, R; Tyl, RW; et al. (2001b) A review of adverse pregnancy outcomes and
formaldehyde exposure in human and animal studies. Regul Toxicol Pharmacol 34:17-34.
Cometto-Muniz, JE; Cain, WS. (1992) Sensory irritation. Relation to indoor air pollution. Ann
NY Acad Sci 641:137-151.
Conaway, CC; Whysner, J; Verna, LK, et al. (1996) Formaldehyde mechanistic data and risk
assessment: endogenous protection from DNA adduct formation. Pharmacol Ther 71:29-55.
Connor, TH; Barrie, MD; Theiss, JC; et al. (1983) Mutagenicity of formalin in the Ames assay.
MutatRes 119:145-149.
Connor, TH; Theiss, JC; Hanna, HA; et al. (1985) Genotoxicity of organic chemicals frequently
found in the air of mobile homes. Toxicol Lett 25:33-40.
Conolly, RB, (1995) Cancer and non-cancer risk assessment: Not so different if you consider
mechanisms. Toxicology 102:179-188.
Conolly, RB. (2002) The use of biologically based modeling in risk assessment. Toxicology
181-182:275-279.
Conolly, RB. (2009) Commentary on "Toxicity testing in the 21st century: Implications for
human health risk assessment" by Krewski et al. Risk Anal 29:480-481.
Conolly, RB; Lutz, WK. (2004) Nonmonotonic dose-response relationships: Mechanistic basis,
kinetic modeling, and implications for risk assessment. Toxicol Sci 77:151-157.
Conolly, RB; Lilly, PD; Kimbell, JS. (2000) Simulation modeling of the tissue disposition of
formaldehyde to predict nasal DNA-protein cross-links in Fischer 344 rats, rhesus monkeys, and
humans. Environ Health Perspect 108 Suppl 5:919-924.
Conolly, RB; Kimbell, JS; Janszen, D; et al. (2003) Biologically motivated computational
modeling of formaldehyde carcinogenicity in the F344 rat. Toxicol Sci 75:432-447.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 11 DRAFT—DO NOT CITE OR QUOTE
-------
Conolly, RB; Kimbell, JS; Janszen, D; et al. (2004) Human respiratory tract cancer risks of
inhaled formaldehyde: dose-response predictions derived from biologically-motivated
computational modeling of a combined rodent and human dataset. Toxicol Sci 82:279-296.
Conolly, RB; Miller, FJ; Kimbell, JS; et al. (2009) Formaldehyde risk assessment. Ann Hyg
53:181-184.
Cook, RJ; Champion, KM; Giometti, CS. (2001) Methanol toxicity and formate oxidation in
NEUT2 mice. Arch Biochem Biophys 393:192-198.
Coon, RA; Jones, RA; Jenkins, LJ, Jr; et al. (1970) Animal inhalation studies on ammonia,
ethylene glycol, formaldehyde, dimethylamine, and ethanol. Toxicol Appl Pharmacol 16:646-
655.
Correa, A; Gray, RH; Cohen, R; et al. (1996) Ethylene glycol ethers and risks of spontaneous
abortion and subfertility. Am J Epidemiol 143:707-717.
Cosma, GN; Marchok, AC. (1988) Benzo[a]pyrene- and formaldehyde-induced DNA damage
and repair in rat tracheal epithelial cells. Toxicology 51:309-320.
Cosma, GN; Jamasbi, R; Marchok, AC. (1988) Growth inhibition and DNA damage induced by
benzo[a]pyrene and formaldehyde in primary cultures of rat tracheal epithelial cells. Mutat Res
201:161-168.
Cosma, GN; Wilhite, AS; and Marchok, AC. (1988) The detection of DNA-protein cross-links
in rat tracheal implants exposed in vivo to benzo[a]pyrene and formaldehyde. Cancer Lett 42:13-
21.
Costa, M; Zhitkovich, A; Harris, M; et al. (1997) DNA-protein cross-links produced by various
chemicals in cultured human lymphoma cells. J Toxicol Environ Health 50:433-449.
Costa, S; Coelho, P; Costa, C, et al. (2008) Genotoxic damage in pathology anatomy laboratory
workers exposed to formaldehyde. Toxicology 252:40-48.
Coste, A; Rateau, JG; Roudot-Thoraval, F; et al. (1996) Increased epithelial cell proliferation in
nasal polyps. Arch Otol Head Neck Surg 122:432-436.
Couch, DB; Allen, PF; and Eales, HC. (1982) The mutagenicity of formaldehyde to Salmonella
typhimurium [Abstract], Environ Mutagen 4:336-337.
Cox, DR. (1972) Regression models and life tables (with discussion). J Roy Stat Soc (Series B)
34:187-220. (as cited in Battelle Columbus Laboratories [1981]).
Cox, D; Hinkley, D. (1974) Theoretical Statistics. Chapman and Hall, London, UK.
Cox, D; Oakes, D. (1984) Analysis of Survival Data. Chapman and Hall, London, UK.
Craft, TR; Bermudez, E; Skopek, TR. (1987) Formaldehyde mutagenesis and formation of DNA-
protein crosslinks in human lymphoblasts in vitro. Mutat Res 176:147-155.
Cralley, LV. (1942) The effect of irritant gases upon the rate of ciliary activity. J Ind Hyg
Toxicol 24:193-198.
Crosby, RM; Richardson, KK; Craft, TR; et al. (1988) Molecular analysis of formaldehyde-
induced mutations in human lymphoblasts and E. coli. Environ Mol Mutagen 12:155-166.
This document is a draft for review purposes only and does not constitute Agency policy.
R-12 DRAFT—DO NOT CITE OR QUOTE
-------
Crump, KS. (1994a) Limitations of biological models of carcinogenesis for low-dose
extrapolation. Risk Anal 14:883-886.
Crump, KS. (1994b) Use of mechanistic models to estimate low-dose cancer risks. Risk Anal
14:1033-1038.
Crump, K. (2002) Benchmark analysis. In: Encyclopedia of Environmetrics. Eds: El-Shaarawi,
A; Piegorsch, WW. Wiley, West Sussex, UK.
Crump, KS; Subramaniam, RP; Van Landingham, C. (2005) A numerical solution to the non-
homogenous two-stage MVK model of cancer. Risk Anal 25:921-926.
Crump, K; Subramanian, RP; Van Landingham, C. (2006) Acknowledgment of prior solution to
MVK cancer model. Risk Anal 26:3-4.
Crump, KS; Chen, C; Fox, JF; et al. (2008) Sensitivity analysis of biologically motivated model
for formaldehyde-induced respiratory cancer in humans. Ann Occup Hyg 52:481-495.
Crump, KS; Chen, C; Fox, JF; et al. (2009) Reply to Conolly et al. (2009) Ann Hyg 53:184—
189.
Dahl, AR; Hadley, WM. (1983) Formaldehyde production promoted by rat nasal cytochrome P-
450-dependent monooxygenases with nasal decongestants, essences, solvents, air pollutants,
nicotine, and cocaine as substrates. Toxicol Appl Pharmacol 67:200-205.
Dalbey, WE. (1982) Formaldehyde and tumors in hamster respiratory tract. Toxicology 24:9-
14.
Dalhamn, T. (1956) Mucus flow and ciliary activity in the trachea of healthy rats and rats
exposed to respiratory irritant gases (SO2, H3N, HCHO): a functional and morphologic (light
microscopy and electron microscopy) study, with special reference to technique. Acta Physiol
Scand 36(Suppl): 1-161.
Dallas, CE; Badeaux, P; Theiss, JC; et al. (1989) The influence of inhaled formaldehyde on rat
lung cytochrome P450. Environ Res 49:50-59.
Dallas, CE; Scott, MJ; Ward, JB, Jr.; et al. (1992) Cytogenetic analysis of pulmonary lavage and
bone marrow cells of rats after repeated formaldehyde inhalation. J Appl Toxicol 12:199-203.
Dally, K; Hanrahan, L; Woodbury, M; et al. (1981) Formaldehyde exposure in nonoccupational
environments. Arch Environ Health 36(6):277-284.
Dalton, P. (2003) Upper airway irritation, odor perception and health risk due to airborne
chemicals. Toxicol Lett 140-141:239-248.
Dasgupta PK, J Li, G Zhang, WT Luke, WA McClenny, J Stutz and A Fried. Summertime
ambient formaldyde in five U.S. metropolitan areas: Nashville, Atlanta, Houston, Philadelphia
and Tampa. Environ Sci Technol 39:4767-4783.
Daston, GP. (2008) Gene expression, dose-response, and phenotypic anchoring: Applications for
toxicogenomics in risk assessment. Toxicol Sci 105:233-234.
Day, JH; Lees, RE; Clark, RH; et al. (1984) Respiratory response to formaldehyde and off-gas of
urea formaldehyde foam insulation. Can Med Assoc J 131:1061-1065.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 13 DRAFT—DO NOT CITE OR QUOTE
-------
De Ceaurriz, JC; Micillino, JC; Bonnet, P; et al. (1981) Sensory irritation caused by various
industrial airborne chemicals. Toxicol Lett 9:137-143.
De Flora, S. (1981) Study of 106 organic and inorganic compounds in the Salmonella/microsome
test. Carcinogenesis 2:283-98.
de Flora, S; Camoirano, A; Zanacchi, P, et al. (1984) Mutagenicity testing with TA97 and
TA102 of 30 DNA-damaging compounds, negative with other Salmonella strains. Mutat Res
134:159-165.
De Stefani, E; Boffetta, P; Brennan, P; et al. (2005) Occupational exposures and risk of
adenocarcinoma of the lung in Uruguay. Cancer Causes and Control 16:851-856.
Dean, JH; Lauer, LD; House, RV; et al. (1984) Studies of immune function and host resistance
in B6C3F1 mice exposed to formaldehyde. Toxicol Appl Pharmacol 72:519-529.
Dearman, RJ; Humphreys, N; Skinner, RA; et al. (2005) Allergen-induced cytokine phenotypes
in mice: role of CD4 and CD8 T cell populations. Clin Exp Allergy 35:498-505.
Degitz, SJ; Zucker, RM; Kawanishi, CY; et al. (2004a) Pathogenesis of methanol-induced
cranofacial defects in C57BL/6J mice. Birth Defects A Clin Mol Teratol 70:172-178.
Degitz, SJ; Rogers, JM; Zucker, RM; et al. (2004b) Developmental toxicity of methanol:
Pathogenesis in CD-I and C57BL/6J mice exposed in whole embryo culture. Birth Defects Res
A Clin Mol Teratol. 70(4): 179-184.
Delfino, RJ. (2002) Epidemiologic evidence for asthma and exposure to air toxics: linkages
between occupational, indoor, and community air pollution research. Environ Health Perspect
110 Suppl 4:573-589.
Delfino, RJ; Gong, H Jr; Linn, WS; et al. (2003) Asthma symptoms in Hispanic children and
daily ambient exposures to toxic and criteria air pollutants. Environ Health Perspect 111 :647-
656.
Dell, L; Teta, MJ. (1995) Mortality among workers at a plastics manufacturing and research and
development facility: 1946-1988. Am J Ind Med 28:373-384.
Deltour, L; Foglio, MH; Duester, G. (1999) Metabolic deficiencies in alcohol dehydrogenase
Adhl, Adh3, and Adh4 null mutant mice. Overlapping roles of Adhl and Adh4 in ethanol
clearance and metabolism of retinol to retinoic acid. J Biol Chem 274:16796-16801.
Demkowicz-Dobrzanski, K; and Castonguay, A. (1992) Modulation by glutathione of DNA
strand breaks induced by 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone and its aldehyde
metabolites in rat hepatocytes. Carcinogenesis 13:1447-1454.
Dicker, E; Cederbaum, Al. (1986) Inhibition of the low-Km mitochondrial aldehyde
dehydrogenase by diethyl maleate and phorone in vivo and in vitro. Implications for
formaldehyde metabolism. Biochem J 240:821-827.
Dikalova, AE; Kadiiska, MB; Mason, RP. (2001) An in vivo ESR spin-trapping study: free
radical generation in rats from formate intoxication—role of the Fenton reaction. Proc Natl Acad
Sci USA 98:13549-13553.
This document is a draft for review purposes only and does not constitute Agency policy.
R-14 DRAFT—DO NOT CITE OR QUOTE
-------
Dillon, D; Combes, R; Zeiger, E. (1998) The effectiveness of Salmonella strains TA100, TA102
and TA104 for detecting mutagenicity of some aldehydes and peroxides. Mutagenesis 13:19-26.
Dinsdale, D; Riley, RA; Verschoyle, RD. (1993) Pulmonary cytochrome P450 in rats exposed to
formaldehyde vapor. Environ Res 62:19-27.
Doenecke, D. (1978) Digestion of chromosomal proteins in formaldehyde treated chromatin.
Hoppe Seylers Z Physiol Chem 359:1343-1352.
Doi, S; Suzuki, S; Morishita, M; et al. (2003) The prevalence of IgE sensitization to
formaldehyde in asthmatic children. Allergy 58(7):668-671.
Doll, R; Peto, R. (1978) Cigarette smoking and bronchial carcinoma: Dose and time
relationships among regular smoker and lifetime nonsmokers. J Epidem Commun Health
32:303-313.
Donovan, SM; Krahn, DF; Stewart, JA; et al. (1983) Mutagenic activities of formaldehyde
(HCHO) and hexamethylphosphoramide (HMPA) in reverse and forward Salmonella
typhimurium mutation assays. Environ Mutagen 5:476.
Doolittle, DJ; Furlong, JW; and Butterworth, BE. (1985) Assessment of chemically induced
DNA repair in primary cultures of human bronchial epithelial cells. Toxicol Appl Pharmacol
79:28-38.
Doty, RL; Cometto-Muniz, JE; Jalowayski, AA; et al. (2004) Assessment of upper respiratory
tract and ocular irritative effects of volatile chemicals in humans. Crit Rev Toxicol 34:85-142.
Dragan, YP; Hully, J; Baker, K; et al. (1995) Comparison of experimental and theoretical
parameters of the Moolgavkar-Venzon-Knudson incidence function for the stages of initiation
and promotion in rat hepatocarcinogenesis. Toxicology 102:161-175.
Dresp, J; and Bauchinger, M. (1988) Direct analysis of the clastogenic effect of formaldehyde in
unstimulated human lymphocytes by means of the premature chromosome condensation
technique. MutatRes 204:349-352.
Dubreuil, A; Bouley, G; Godin, J; et al. (1976) Continuous inhalation of small amounts of
formaldehyde: experimental study in the rat. Eur J Toxicol Environ Hyg 9:245-250.
Dulskiene, V; Grazuleviciene, R. (2005) Environmental risk factors and outdoor formaldehyde
and risk of congenital heart malformations. Medicina (Kaunas) 41:787-795.
Dykewicz, M; Patterson, R; Cugell, DW; et al. (1991) Serum IgE and IgGto formaldehyde-
human serum albumin lack of relation to gaseous formaldehyde exposure and symptoms. J
Allergy Clin Immunol 87:48-57.
Edling, C; Jarvholm, B; Andersson, L; et al. (1987) Mortality and cancer incidence among
workers in an abrasive manufacturing industry. Br J Ind Med 44:57-59.
Edling, C; Hellquist, H; Odkvist, L. (1988) Occupational exposure to formaldehyde and
histopathological changes in the nasal mucosa. Br J Ind Med 45:761-765.
Egle, JL, Jr. (1972) Retention of inhaled formaldehyde, propionaldehyde, and acrolein in the
dog. Arch Environ Health 25:119-124.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 15 DRAFT—DO NOT CITE OR QUOTE
-------
Ehlers, CL. (2007) Variations in ADH and ALDH in Southwest California Indians. Alcohol Res
Health 30:14-17.
Emri, G; Schaefer, D; Held, B; et al. (2004) Low concentrations of formaldehyde induce DNA
damage and delay DNA repair after UV irradiation in human skin cells. Exp Dermatol 13:305—
315.
Environment California. (2008) New Report: Formaldehyde from baby nursery furniture
contaminates indoor air Californis's green chemistry initative can protect consumers. Available
at http://www.environmentcalifornia.org/newsroom/environmental-health/environmental-
health/new-report-formaldehyde-from-baby-nursery-furniture-contaminates-indoor-air.
Epstein, S.S.; Shafner, H. (1968) Chemical mutagens in human environment. Nature 219: 385.
Epstein, S.S. et al (1972) Detection of chemical mutagens by the dominant lethal assay in the
mosue. Toxicol. Appl. Pharmacol. 23:288.
Erdei, E; Bobvos, J; Brozik, M; et al. (2003) Indoor air pollutants and immune biomarkers
among Hungarian asthmatic children. Arch Environ Health 58:337-347.
Estonius, M; Svensson, S; Hoog, JO. (1996) Alcohol dehydrogenase in human tissues:
localization of transcript coding for five classes of the enzyme. FEBS Lett 397:338-342.
Ezratty, V; Bonay, M; Neukirch, C; et al. (2007) Effect of formaldehyde on asthmatic response
to inhaled allergen challenge. Environ Health Perspect 115(2):210—214.
Fanucchi, MV; Buckpitt, AR; Murphy, ME; et al. (1997) Naphthalene cytotoxicity of
differentiating Clara cells in neonatal mice. Toxicol Appl Pharmacol 144:96-104.
Fanucchi, MV; Buckpitt, AR; Murphy, ME; et al. (2000) Development of phase II xenobiotic
metabolizing enzymes in differentiating murine Clara cells. Toxicol Appl Pharmacol 168:253-
267.
Fanucchi, MV; Day, KC; Clay, CC; et al. (2004) Increased vulnerability of neonatal rats and
mice to 1-nitronaphthalene-induced pulmonary injury. Toxicol Appl Pharmacol 201:53-65.
Farooqui, MY; Upreti, RK; Ahmed, AE; et al. (1986) Influence of intraperitoneally administered
formaldehyde on bile production and tissue glutathione levels in rats. Res Commun Chem
Pathol Pharmacol 53:233-236.
Fayerweather, WE; Pell, S; Bender, J. (1983) Case-control study of cancer deaths in DuPont
workers with potential exposure to formaldehyde. In: Clary, J; Gibson, J; Waritz, RS; eds.
Formaldehyde: toxicology, epidemiology, and mechanisms. New York: Marcel Dekker; pp. 47-
121.
Fennell, TR. (1994) Development of methods for measuring biological markers of formaldehyde
exposure. Res Rep Health Eff Inst: 1-20; discussion 21-26.
Fernandez, MR; Biosca, JA; Pares, X. (2003) S-nitrosoglutathione reductase activity of human
and yeast glutathione-dependent formaldehyde dehydrogenase and its nuclear and cytoplasmic
localization. Cell Mol Life Sci 60:1013-1018.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 16 DRAFT—DO NOT CITE OR QUOTE
-------
Feron, VJ; Woutersen, RA; Appelman, LM. (1984) Epithelial damage and tumours of the nose
after exposure to four different aldehydes by inhalation. In: Grosdanoff, P; Bass, R;
Hackenberg, U; et al.; eds. Problems of inhalation toxicity studies. Munich: MMV Medizin
Verlag; pp. 587-610.
Feron, VJ; Bruyntjes, JP; Woutersen, RA; et al. (1988) Nasal tumours in rats after short-term
exposure to a cytotoxic concentration of formaldehyde. Cancer Lett 39:101-111.
Feron, VJ; Wilmer, JW; Woutersen, RA; et al. (1989) Inhalation toxicity and carcinogenicity of
formaldehyde in animals: significance for assessment of human health risk. In: Bates, D;
Dungworth, DL; Lee, P; et al.; eds. Assessment of inhalation hazards: integration and
extrapolation using diverse data. Berlin: Springer-Verlag; pp. 131-138.
Feron, VJ; Til, HP; de Vrijer, F; et al. (1991) Aldehydes: occurrence, carcinogenic potential,
mechanism of action and risk assessment. Mutat Res 259:363-385.
Feron, VJ; Arts, JH; Kuper, CF; et al. (2001) Health risks associated with inhaled nasal
toxicants. Crit Rev Toxicol 31:313-347.
Fiddler, W; Miller, AJ; Pensabene, JW; et al. (1984) Investigation on the mutagenicity of N-
nitrosothiazolidine using the Ames Salmonella test. IARC Sci Publ:95-100.
Figuero, WG; Raszkowski, R; Weiss, W. (1973) Lung cancer in chloromethyl ether workers. N
Engl J Med 288:1096-1097.
Fireman, P. (2003) Understanding asthma pathophysiology. Allergy Asthma Proc 24:79-83.
Fishbein, L. (1992) Exposure from occupational versus other sources. Scand J Work Environ
Health 18(Suppl 1):5-16.
Fleig, I; Petri, N; Stocker, WG; et al. (1982) Cytogenetic analyses of blood lymphocytes of
workers exposed to formaldehyde in formaldehyde manufacturing and processing. J Occup Med
24:1009-1012.
Fleisher, J. (1987) Medical students' exposure to formaldehyde in gross anatomy laboratories. N
Y State J Med 87(7):385-388.
Flo-Neyret, C; Lorenzl-Filho, G; Macchione, M; et al. (2001) Effects of formaldehyde on the
frog's mucociliary epithelium as a surrogate to evaluate air pollution effects on the respiratory
epithelium. J Med Bil Res 34:639-643.
Fontignie-Houbrechts, N. (1981) Genetic effects of formaldehyde in the mouse. Mutat Res
88:109-114.
Fontignie-Houbrechts, N; Moutschen-Dahmen, M; Degraeve, N; et al. (1982) Genetic effects in
the mouse of formaldehyde in combination with adenosine and hydrogen peroxide. Mutat Res
104:371-376.
Fornace, AJ, Jr. (1982) Detection of DNA single-strand breaks produced during the repair of
damage by DNA-protein cross-linking agents. Cancer Res 42:145-149.
Franklin, P; Dingle, P; Stick, S. (2000) Raised exhaled nitric oxide in healthy children is
associated with domestic formaldehyde levels. Am J Respir Crit Care Med 161:1757-1759.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 17 DRAFT—DO NOT CITE OR QUOTE
-------
Franks, SJ. (2005) A mathematical model for the absorption and metabolism of formaldehyde
vapour in humans. Toxicol Appl Pharmacol 206:309-320.
Fransman, W; McLean, D; Douwes, J; et al. (2003) Respiratory symptoms and occupational
exposures in New Zealand plywood mill workers. Ann Occup Hyg 47(4):287-295.
Frazelle, JH; Abernethy, DJ; and Boreiko, CJ. (1983) Weak promotion of C3H/10T1/2 cell
transformation by repeated treatments with formaldehyde. Cancer Res 43:3236-3239.
Frei, E; Pool, BL; Plesch, W; et al. (1984) Biochemical and biological properties of prospective
N-nitrodialkylamine metabolites and their derivatives. IARC Sci Publ:491-497.
Friedman, GD; Ury, HK. (1983) Screening for possible drug carcinogenicity: second report of
findings. J Natl Cancer Inst 71:1165-1175.
Frigas, E; Filley, WV; Reed, CE. (1984) Bronchial challenge with formaldehyde gas: lack of
bronchoconstriction in 13 patients suspected of having formaldehyde-induced asthma. Mayo
Clin Proc 59:295-299.
Fujimaki, H; Kurokawa, Y; Kunugita, N; et al. (2004a) Differential immunogenic and
neurogenic inflammatory responses in an allergic mouse model exposed to low levels of
formaldehyde. Toxicology 197:1-13.
Fujimaki, H; Kurokawa, Y; Kakeyama, M; et al. (2004b) Inhalation of low-level formaldehyde
enhances nerve growth factor production in the hippocampus of mice. Neuroimmunomodulation
11:373-375.
Fujimura, Y. (2000) Evidence of M cells as portals of entry for antigens in the nasopharyngeal
lymphoid tissue of humans. Virchows Arch 436:560-566.
Galloway, SM; Bloom, AD; Resnick, M, et al. (1985) Development of a standard protocol for in
vitro cytogenetic testing with Chinese hamster ovary cells: comparison of results for 22
compounds in two laboratories. Environ Mutagen 7:1-51.
Gaiter, D; Carmine, A; Buervenich, S; et al. (2003) Distribution of class I, III and IV alcohol
dehydrogenase mRNAs in the adult rat, mouse and human brain. Eur J Biochem 270:1316-
1326.
Gandhi, M; Aweeka, F; Greenblatt, RM; et al. (2004) Sex differences in pharmacokinetics and
pharmaco-dynamics Ann Rev Pharmacol Toxicol 44:499-523.
Garcia, CL; Mechilli, M; De Santis, LP, et al. (2009) Relationship between DNA lesions, DNA
repair and chromosomal damage induced by acetaldehyde. Mutat Res 662:3-9.
Garcia, GJ; Schroeter, JD; Segal, RA; et al. (2009) Dosimetry of nasal uptake of water-soluble
and reactive gases: a first study of interhuman variability. Inhal Toxicol. 21(7):607-18.
Gardner, MJ; Pannett, B; Winter, PD; et al. (1993) A cohort study of workers exposed to
formaldehyde in the British chemical industry: an update. Br J Ind Med 50:827-834.
Gardner, RJ; Burgess, BA; Kennedy, GL, Jr. (1985) Sensory irritation potential of selected nasal
tumorigens in the rat. Food Chem Toxicol 23:87-92.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 18 DRAFT—DO NOT CITE OR QUOTE
-------
Garrett, MH; Hooper, MA; Hooper, BM; et al. (1999) Increased risk of allergy in children due to
formaldehyde exposure in homes. Allergy 34:330-337.
Garry, VF; Kreiger, RA; and Wiencke, JK. (1981) The mutagenic and cytotoxic effects of
formaldehyde in cultured human lymphocytes [Abstract], Environ Mutagen 3:341.
Gart, JJ; Krewski, D; Lee, PN; et al. (1986) Statistical methods in cancer research. Volume III—
The design and analysis of long-term animal experiments. IARC Sci Publ (79): 1-219.
Gaylor, DW; Zheng, O. (1996) Risk assessment of nongenotoxic carcinogens based upon cell
proliferation/death rates in rodents. Risk Anal 16:221-225.
Gaylor, DW; Lutz, WK; Conolly, RB. (2004) Statistical analysis of nonmonotonic dose-response
relationships: Research design and analysis of nasal cell proliferation in rats exposed to
formaldehyde. Toxicol Sci 77:158-164.
Gee, IL; Watson, AFR; Tavernier, G; et al. (2005) Indoor air quality, environmental tobacco
smoke and asthma: A case control study of asthma in a community population. Indoor Built
Environ 14(3-4):215-219.
Georgieva, AV; Kimbell, JS; Schlosser, PM. (2003) A distributed-parameter model for
formaldehyde uptake and disposition in the rat nasal lining. Inhal Toxicol 15:1435-1463.
Gerard, C. (2005) Biomedicine. Asthmatics breathe easier when its SNO-ing. Science
308:1560-1561.
Gerberich, HR; Seaman, GC. (1994) Formaldehyde. In: Kroshwitz, JI; Howe-Grant, M; eds.
Kirk-Othmer encyclopedia of chemical technology. Vol. 11. 4th edition. New York, NY: John
Wiley and Sons; pp. 929-951.
Gerde, P; Cheng, YS; Medinsky, MA. (1991) In vitro deposition of ultrafine aerosols in the nasal
airways of the rat. Fundam Appl Toxicol 16:330-336.
Gerin, M; Siemiatycki, J; Nadon, L; et al. (1989) Cancer risks due to occupational exposure to
formaldehyde: Results of a multi-site case-control study in Montreal Canada. Int J Cancer
44:53-58.
Ginsberg, G; Smolenski, S; Hattis, D; et al. (2002) Population distribution of aldehyde
dehydrogenase-2 genetic polymorphism: implications for risk assessment. Regul Toxicol
Pharmacol 36:297-309.
Ginsberg, G; Hattis, D; Sonawane, B. (2004) Incorporating pharmacokinetic differences between
children and adults in assessing children's risks to environmental toxicants. Toxicol Appl
Pharmacol 198:164-183.
Ginsberg, G; Foos, B; Firestone, M. (2005) Review and Analysis of Inhalation Dosimetry
Methods for Application to Children's Risk Assessment. J. Tox. & Env. Health 68:573-615.
Ginsberg,G; Asgharian, B; Kimbell, J; et al. (2008) Modeling approaches for estimating the
dosimetry of inhaled toxicants in children. J. Tox. & Env. Health 71:166-195.
Global Insight, 2006. Economic primer on formaldehyde. Prepared for Formaldehyde Council,
Inc. http://www.formaldehyde.org/pdfs/GIEconomicPrimer-04-20-06.pdf.
This document is a draft for review purposes only and does not constitute Agency policy.
R. 19 DRAFT—DO NOT CITE OR QUOTE
-------
Gochfeld, M. (2007) Framework for gender differences in human and animal toxicology.
Environ Res 104:4-21.
Gofmekler, V. (1968) Embryotropic action of benzene and formaldehyde inhalation. Hyg Sanit
33:327-332.
Gofmekler, V; Bonashevskaya, T. (1969) Experimental studies of teratogenic properties of
formaldehyde, based on pathological investigations. Hyg Sanit 34:266-268.
Golaliour MJ; Azarhoush R; Ghafari S; Gharravi AM; Fazeli SA; Davarian A. 2007.
Formaldehyde exposure induces histopathological and morphometric changes in the rat testis.
Folia Morphol. 66(3): 167-171.
Golden, R; Pyatt, D; Shields, PG. (2006) Formaldehyde as a potential human leukemogen; An
assessment of biological plausibility. Crit Rev Toxicol 36:135-153.
Goldmacher, VS; and Thilly, WG. (1983) Formaldehyde is mutagenic for cultured human cells.
MutatRes 116:417-422.
Goldstein, BD. (1973) Textiles and the chemical industry: A marriage. J Am Assoc Textile
Chemists and Colorists 10:209-214.
Gordon, MY; Marley, SB; Apperley, JF; et al. (2003) Clinical heterogeneity in chronic myeloid
leukemia reflecting biological diversity in normal persons. Brit J Haematol 122:424-429.
Gordon, CJ; Spencer, PJ; Hotchkiss, J; et al. (2008) Thermoregulation and its influence on risk
assessment. Toxicology 244:87-97.
Gorski, P; Tarkowski, M; Krakowiak, A; et al. (1992) Neutrophil chemiluminescence following
exposure to formaldehyde in healthy subjects and in patients with contact dermatitis.
Immunologia Clinica 11:53-58.
Gottschling, LM; Beaulieu, HJ; Melvin, WW. (1984) Monitoring of formic acid in urine of
humans exposed to low levels of formaldehyde. Am Ind Hyg Assoc J 45:19-23.
Grafstrom, RC; Fornace, AJ, Jr.; Autrup, H; et al. (1983) Formaldehyde damage to DNA and
inhibition of DNA repair in human bronchial cells. Science 220:216-218.
Grafstrom, RC; Fornace, A, Jr.; Harris, CC. (1984) Repair of DNA damage caused by
formaldehyde in human cells. Cancer Res 44:4323-4327.
Grafstrom, RC; Curren, RD; Yang, LL; et al. (1985) Genotoxicity of formaldehyde in cultured
human bronchial fibroblasts. Science 228:89-91.
Grafstrom, RC. (1990) In vitro studies of aldehyde effects related to human respiratory
carcinogenesis. MutatRes 238:175-84.
Grafstrom, RC; Hsu, IC; and Harris, CC. (1993) Mutagenicity of formaldehyde in Chinese
hamster lung fibroblasts: synergy with ionizing radiation and N-nitroso-N-methylurea. Chem
Biol Interact 86:41-49.
Grammer, LC; Harris, KE; Shaughnessy, M; et al. (1990) Clinical and immunologic evaluation
of 37 workers exposed to gaseous formaldehyde. J Allergy Clin Immunol 86:177-181.
This document is a draft for review purposes only and does not constitute Agency policy.
R-20 DRAFT—DO NOT CITE OR QUOTE
-------
Grammer, LC; Harris, KE; Cugell, DW; et al. (1993) Evaluation of a worker with possible
formaldehyde-induced asthma. J Allergy Clin Immunol 92:29-33.
Grasl-Kraupp, B; Luebeck, G; Wagner, A; et al. (2000) Quantitative analysis of tumor initiation
in rat liver: Role of cell replication and cell death (apoptosis). Carcinogenesis 21:1411-1421.
Graves, RJ; Callander, RD; and Green, T. (1994) The role of formaldehyde and S-
chloromethylglutathione in the bacterial mutagenicity of methylene chloride. Mutat Res
320:235-243.
Graves, RJ; Trueman, P; Jones, S; et al. (1996) DNA sequence analysis of methylene chloride-
induced HPRT mutations in Chinese hamster ovary cells: comparison with the mutation
spectrum obtained for 1,2-dibromoethane and formaldehyde. Mutagenesis 11:229-233.
Gray, RH; Wu, LY. (2000) Subfertility and risk of spontaneous abortion. Am J Public Health
90:1452-1454.
Grazuleviciene, R; Dulskiene, V; Vencloviene, J. (1998) Formaldehyde exposure and low birth
weight incidence. J Occup Health 40:61-67.
Greaves, MA. (2004) Biological models for leukemia and lymphoma. IARC Sci Publ
(157):351—372.
Green, DJ; Sauder, LR; Kulle, TJ; et al. (1987) Acute response to 3.0 ppm formaldehyde in
exercising healthy nonsmokers and asthmatics. Am Rev Respir Dis 135:1261-1266.
Green, DJ; Bascom, R; Healey, EM; et al. (1989) Acute pulmonary response in healthy,
nonsmoking adults to inhalation of formaldehyde and carbon. Toxicol Environ Health 28:261-
275.
Griesemer, RA; Boreiko, CJ; de Serres, FJ; et al. (1985) Report on the consensus workshop on
formaldehyde: Carcinogenicity/Histopathology/Genotoxicity Panel report. Environ Health
Perspect 58:340-343.
Guess, H; Crump, K; Peto, R. (1977) Uncertainty estimates for low dose-rate extrapolations of
animal carcinogenicity data. Cancer Res 37:3475-3483.
Guseva, VA. (1972) Study of the gonadotropic effect in male rats under the effect of
formaldehyde administered simultaneously with air and water [translation]. Gig Sanit 37:101-
102.
Gustavsson, P; Jakobsson, R; Johannson, H; et al. (1998) Occupational exposures and squamous
cell carcinoma of the oral cavity, pharynx, larynx, and oesophagus: a case-control study in
Sweden. Occup Environ Med 55:393-400.
Hagiwara, M; Watanabe, E; Barrett, JC, et al. (2006) Assessment of genotoxicity of 14 chemical
agents used in dental practice: ability to induce chromosome aberrations in Syrian hamster
embryo cells. Mutat Res 603:111-120.
Hakim, RB; Gray, RH; Zacur, H. (1995) Infertility and early pregnancy loss. Am J Obstet
Gynecol 172:1510-1517.
Hall, A; Harrington, J; Aw, TC. (1991) Mortality study of British pathologists. Am J Ind Med
20:83-89.
This document is a draft for review purposes only and does not constitute Agency policy.
R-21 DRAFT—DO NOT CITE OR QUOTE
-------
Hamaguchi, F; and Tsutsui, T. (2000) Assessment of genotoxicity of dental antiseptics: ability
of phenol, guaiacol, p-phenolsulfonic acid, sodium hypochlorite, p-chlorophenol, m-cresol or
formaldehyde to induce unscheduled DNA synthesis in cultured Syrian hamster embryo cells.
Jpn J Pharmacol 83:273-276.
Hanahan, D; Weinberg, RA. (2000) The hallmarks of cancer. Cell 100:57-70.
Hanrahan, LP; Dally, KA; Anderson, HA; et al. (1984) Formaldehyde vapor in mobile homes: a
cross sectional survey of concentrations and irritant effects. Am J Public Health 74(9): 1026-
1027.
Hansen, J; Olsen, JH. (1995) Formaldehyde and cancer morbidity among male employees in
Denmark. Cancer Causes Control 6:354-360.
Hanzlik, RP; Fowler, SC; Eells, JT. (2005) Absorption and elimination of formate following oral
administration of calcium formate in female human subjects. Drug Metab Dispos 33:282-286.
Harrington, JM; Shannon, H. (1975) Mortality study of pathologists and medical laboratory
technicians. Br Med J 4:329-332.
Harrington, JM; Oakes, D. (1984) Mortality study of British pathologists 1974-80. Br J Ind Med
41:188-191.
Harris, NL; Jaffe, ES; Diebold, J; et al. (2000a) The World Health Organization classification of
neoplasms of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee
meeting—Airlie House, Virginia, November, 1997. Hematol J 1:53-66.
Harris, NL; Jaffe, ES; Diebold, J; et al. (2000b) The World Health Organization classification of
hematological malignancies report of the Clinical Advisory Committee meeting, Airlie House,
Virginia, November 1997. Mod Pathol 13:193-207.
Harris, C; Wand, S-W; Lauchu, JJ; et al. (2003) Methanol metabolism and embryotoxicity in rat
and mouse conceptuses: comparison of alcohol dehydrogenase (ADH1), formaldehyde
dehydrogenase (ADH3), and catalase. Reprod Toxicol 17:349-357.
Harving, H; Korsgaard, J; Pedersen, OF; et al. (1990) Pulmonary function and bronchial
reactivity in asthmatics during low-level formaldehyde exposure. Lung 168:15-21.
Haseman, JK. (1995) Data analysis: Statistical analysis and use of historical control data. Regul
Toxicol Pharmacol 21:52-59.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2003) Mortality from lymphohematopoietic
malignancies among workers in formaldehyde industries. J Natl Cancer Inst 95:1615-1623.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2004a) Mortality from solid cancers among
workers in formaldehyde industries. Am J Epidemiol 159:1117-1130.
Hauptmann, M; Lubin, JH; Stewart, PA; et al. (2004b) Response. J Natl Cancer Inst 96:967-
968.
Haworth, S; Lawlor, T; Mortelmans, K; et al. (1983) Salmonella mutagenicity test results for 250
chemicals. Environ Mutagen 5 Suppl 1:1-142.
This document is a draft for review purposes only and does not constitute Agency policy.
R-22 DRAFT—DO NOT CITE OR QUOTE
-------
Hayashi, H; Kunugita, N; Arashidani, K; et al. (2004) Long-term exposure to low levels of
formaldehyde increases the number of tyrosine hydroxylase-immunopositive periglomerular
cells in mouse main olfactory bulb. Brain Res 1007:192-197.
Hayes, RB; Raatgever, JW; de Bruyn, A; et al. (1986) Cancer of the nasal cavity and paranasal
sinuses, and formaldehyde exposure. Int J Cancer 37:487-492.
Hayes, RB; Blair, A; Stewart, PA; et al. (1990) Mortality of U.S. embalmers and funeral
directors. Am J Ind Med 18:641-652.
Hayes, RB; Klein, S; Suruda, A, et al. (1997) 06-alkylguanine DNA alkyltransferase activity in
student embalmers. Am J Ind Med 31:361-365.
He, JL; Jin, LF; Jin, HY. (1998) Detection of cytogenetic effects in peripheral lymphocytes of
students exposed to formaldehyde with cytokinesis-blocked micronucleus assay. Biomed
Environ Sci 11:87-92.
Health Canada and Environment Canada. (2001) Priority substances list assessment report:
formaldehyde. Hull, Quebec, Canada: Environment Canada and Health Canada.
Heck, HD, Casanova-Schmitz, M. (1984) The relevance of disposition studies to the toxicology
of formaldehyde. CIIT Activities 4:1-5.
Heck HD; Casanova, M. (1987) Isotope effects and their implications for the covalent binding of
inhaled [3H]- and [14C]formaldehyde in the rat nasal mucosa. Toxicol Appl Pharmacol 89:122-
134.
Heck, HD; Casanova, M. (1999) Pharmacodynamics of formaldehyde: applications of a model
for the arrest of DNA replication by DNA-protein cross-links. Toxicol Appl Pharmacol 160:86-
100.
Heck, HD, Casanova, M. (2004) The implausibility of leukemia induction by formaldehyde: A
critical review of the biological evidence on distant-site toxicity. Regul Toxicol Pharmacol
40:92-106.
Heck, HD; White, EL; Casanova-Schmitz, M. (1982) Determination of formaldehyde in
biological tissues by gas chromatography/mass spectrometry. Biomed Mass Spectrom 9:347-
353.
Heck, HD; Chin, TY; Schmitz, MC. (1983) Distribution of [14C] formaldehyde in rats after
inhalation exposure. In: Gibson, J E , eds. Formaldehyde Toxicity. Washington, DC:
Hemisphere Publishing Corporation; pp. 26-37.
HeckHd, A; Casanova, M; L., C-W, et al. (1986). The formation of DNA-protein crosslinks by
aldehydes present in tobacco smoke. In: Hoffman, D and CC Harris,ed, Mechanisms in Tobacco
Carcinogenesis: Cold Spring Harbor, New York: CSH Press, 215-230.
Heck, HD; Casanova-Schmitz, M; Dodd, PB; et al. (1985) Formaldehyde (CH20) concentrations
in the blood of humans and Fischer-344 rats exposed to CH20 under controlled conditions. Am
IndHyg Assoc J 46:1-3.
Heck, HD; Casanova, M; Starr, TB. (1990) Formaldehyde toxicity—new understanding. Crit Rev
Toxicol 20:397-426.
This document is a draft for review purposes only and does not constitute Agency policy.
R-23 DRAFT—DO NOT CITE OR QUOTE
-------
Hedberg, JJ; H00g, JO; Nilsson, JA; et al. (2000) Expression of alcohol dehydrogenase 3 in
tissue and cultured cells from human oral mucosa. Am J Pathol 157:1745-1755.
Hedberg, JJ; Grafstrom, RC; Vondracek, M; et al. (2001a) Micro-array chip analysis of
carbonyl-metabolizing enzymes in normal, immortalized and malignant human oral
keratinocytes. Cell Mol Life Sci 58:1719-1726.
Hedberg, JJ; Backlund, M; Stromberg, P; et al. (2001b) Functional polymorphism in the alcohol
dehydrogenases (ADH3) promoter. Pharmacogenetics 11:815-824.
Hedberg, JJ; Griffiths, WJ; Nilsson, SJ; et al. (2003) Reduction of S-nitrosoglutathione by
human alcohol dehydrogenase 3 is an irreversible reaction as analyzed by electrospray mass
spectrometry. Eur J Biochem 270:1249-1256.
Heikkila, P; Priha, E; Savola, A. (1991) Exposure at work. 14. Formaldehyde. Institute of
Occupational Health. Finnish Work Environmental Fund.
Heineman, EF; Olsen, JH; Pottern, LM; et al. (1992) Occupational risk factors for multiple
myeloma among Danish men. Cancer Causes Control 3:555-568.
Hemminki, K; Mutanen, P; Salonienmi, I; et al. (1982) Spontaneous abortions in hospital staff
engaged in sterilizing instruments with chemical agents. Br Med J 285:1461-1463.
Hemminki, K; Kyyronen, P; Lindbohm, ML. (1985) Spontaneous abortions and malformations
in the offspring of nurses exposed to anaesthetic gases, cytostatic drugs, and other potential
hazards in hospitals, based on registered information of outcome. J Epidemiol Community
Health 39:141-147.
Hendrick, DJ; Lane, DJ. (1975) Formalin asthma in hospital staff. Br Med J l(5958):607-608.
Hendrick, DJ; Lane, DJ. (1977) Occupational formalin asthma. Br J Ind Med 34:11-18.
Hendrick, DJ; Rando, RJ; Lane, DJ; et al. (1982) Formaldehyde asthma: challenge exposure
levels and fate after five years. J Occup Med 24:893-897.
Herbert, FA; Hessel, PA; Melenka, LS; et al. (1994) Respiratory consequences of exposure to
wood dust and formaldehyde of workers manufacturing oriented strand board. Arch Environ
Health 49(6):465-470.
Hernandez, O; Rhomberg, L; Hogan, K; et al. (1994) Risk assessment of formaldehyde. J
Hazard Mater 39:161-172.
Hess, DT; Matsumoto, A; Kim, SO; et al. (2005) Protein S-nitrosylation: purview and
parameters. Nat Rev Mol Cell Biol 6:150-166.
Hester, SD; Benavides, GB; Yoon, L; et al. (2003) Formaldehyde-induced gene expression in
F344 rat nasal respiratory epithelium. Toxicology 187:13-24.
Hester, SD; Barry, WT; Zou, F; et al. (2005) Transcriptomic analysis of F344 rat nasal
epithelium suggests that the lack of carcinogenic response to glutaraldehyde is due to its greater
toxicity compared to formaldehyde. Toxicol Pathol 33:415-424.
This document is a draft for review purposes only and does not constitute Agency policy.
R-24 DRAFT—DO NOT CITE OR QUOTE
-------
Hikiba, H; Watanabe, E; Barrett, JC; et al. (2005) Ability of fourteen chemical agents used in
dental practice to induce chromosome aberrations in Syrian hamster embryo cells. J Pharmacol
Sci 97:146-152.
Hildesheim, A; Dosemeci, M; Chan, CC; et al. (2001) Occupational exposure to wood,
formaldehyde, and solvents and risk of nasopharyngeal carcinoma. Cancer Epidemiol
Biomarkers Prev 10:1145-1153.
Hileman, B. (1984) Formaldehyde: assessing the risk. Environ Sci Technol 18:216A-221A.
Hilton, J; Dearman, RJ; Harvey, P; et al. (1998) Estimation of relative skin sensitizing potency
using the local lymph node assay: a comparison of formaldehyde with glutaraldehyde. Am J
Contact Dermat 9:29-33.
Hines, RN; McCarver, DG. (2002) The ontogeny of human drug-metabolizing enzymes: phase I
oxidative enzymes. J Pharmacol Exp Ther 300:361-366.
Holly, EA; Aston, DA; Ahn, DK; et al. (1996) Intraocular melanoma linked to occupations and
chemical exposures. Epidemiology 7:55-61.
Holmstrom, M; Wilhelmsson, B. (1988) Respiratory symptoms and pathophysiological effects of
occupational exposure to formaldehyde and wood dust. Scand J Work Environ Health 14:306-
311.
Holmstrom, M; Wilhelmsson, B; Hellquist, H. (1989a) Histological changes in the nasal mucosa
in rats after long-term exposure to formaldehyde and wood dust. Acta Otolaryngol 108:274-
283.
Holmstrom, M; Rynnel-Dagoo, B; Wilhelmsson, B. (1989b) Antibody production in rats after
long-term exposure to formaldehyde. Toxicol Appl Pharmacol 100:328-333.
Holmstrom, M; Wilhelmsson, B; Hellquist, H; et al. (1989c) Histological changes in the nasal
mucosa in persons occupationally exposed to formaldehyde alone and in combination with wood
dust. Acta Otolaryngol 107:120-129.
Holness, DL; Nethercott, JR. (1989) Health status of funeral service workers exposed to
formaldehyde. Arch Environ Health 44:222-228.
Hoog, JO; Hedberg, JJ; Stromberg, P; et al. (2001) Mammalian alcohol dehydrogenase -
functional and structural implications. J Biomed Sci 8:71-76.
Hoog, JO; Stromberg, P; Hedberg, JJ; et al. (2003) The mammalian alcohol dehydrogenases
interact in several metabolic pathways. Chem Biol Interact 143-144:175-181.
Hoogenveen, R; Clewell, H; Andersen, M; et al. (1999) An alternative exact solution of the two-
stage clonal growth model of cancer. Risk Anal 19:9-14.
Horton, A; Tye, R; Stemmer, K. (1963) Experimental carcinogenesis of the lung. Inhalation of
gaseous formaldehyde or an aerosol of coal tar by C3H mice. J Natl Cancer Inst 30:31-43.
Horvath, E, Jr; Anderson, J; Pierce, WE; et al. (1988) Effects of formaldehyde on the mucous
membranes and lungs: A study of an industrial population. JAMA 259:701-707.
This document is a draft for review purposes only and does not constitute Agency policy.
R-25 DRAFT—DO NOT CITE OR QUOTE
-------
Hou, Y; Hu, WG; Hirano, T; et al. (2002) A new intra-NALT route elicits mucosal and systemic
immunity against Moraxella catarrhalis in a mouse challenge model. Vaccine 20:2375-2381.
HSDB (Hazardous Substances Data Bank). (2001) Formaldehyde. National Library of
Medicine, Bethesda, MD. Available from .
Hurni, H; Ohder, H. (1973) Reproduction study with formaldehyde and hexamethylenetetramine
in beagle dogs. Food Cosmet Toxicol 11:459-462.
IARC (International Agency for Research on Cancer). (1995) IARC monographs on the
evaluation of carcinogenic risks to humans. Vol. 62. Wood dust and formaldehyde. Lyon,
France: International Agency for Research on Cancer. Available from
and
.
IARC (2006) IARC Monographs on the evaluation of carcinogenic risks to humans.
Formaldehyde, 2-Butoxyethanol and l-tert-Butoxy-2-propanol, vol. 88. IARC, Lyon, France.
http://monographs.iarc.fr/ENG/Monographs/vol88/mono88-6A.pdf
Iarmarcovai, G; Botta, A; Orsiere, T. (2006a) Number of centromeric signals in micronuclei and
mechanisms of aneuploidy. Toxicol Lett 166:1-10.
Iarmarcovai, G; Orsiere, T; Sari-Minodier, I; et al. (2006b) Different pathways of centromeric
micronuclei formation in peripheral blood lymphocytes of cancer patients and workers
occupationally exposed to mutagens/carcinogens? Advances in Molecular and Clinical Aspects
of Cancer Prevention,4th International Conference on Cancer Prevention 2006, 4:54-55,
Switzerland.
Iarmarcovai, G; Bonassi, S; Sari-Minodier, I; et al. (2007) Exposure to genotoxic agents, host
factors, and lifestyle influence the number of centromeric signals in micronuclei: a pooled re-
analysis. Mutat Res 615:18-27.
Ibanes, JD; Leininger, JR; Jarabek, AM; et al. (1996) Reexamination of respiratory tract
responses in rats, mice, and rhesus monkeys chronically exposed to inhaled chlorine. Inhal Tox
8:859-876.
ICRP (International Commission on Radiological Protection). (1994) Human respiratory tract
model for radiological protection. Publication 66 Annals of the ICRP Vol 24; Pergamon Press,
Oxford, UK.
Ihrig, A; Hoffmann, J; Triebig, G. (2006) Examination of the influence of personal traits and
habituation on the reporting of complaints at experimental exposure to ammonia. Int Arch
Occup Environ Health 79:332-338.
ILSI (International Life Sciences Institute). (2003) Final report of a workshop to develop a
framework for assessing risks to children from exposures to environmental agents. ILSI Risk
Science Institute, Washington, DC.
Im, H; Oh, E; Mun, J; et al. (2006) Evaluation of toxicological monitoring markers using
proteomic analysis in rats exposed to formaldehyde. J Proteome Res 5:1354-1366.
This document is a draft for review purposes only and does not constitute Agency policy.
R-26 DRAFT—DO NOT CITE OR QUOTE
-------
Ionescu, J; Marinescu, D; Tapu, V; et al. (1978) Experimental chronic obstructive lung disease.
I. Bronchopulmonary changes induced in rabbits by prolonged exposure to formaldehyde.
Morphol Embryol 24:233-242.
IPCS (International Programme on Chemical Safety). (1989) Environmental Health Criteria.
Vol. 89. Formaldehyde. Geneva, Switzerland: World Health Organization. Available from
.
IPCS (International Programme on Chemical Safety). (1997) Environmental Health Criteria 196
— Methanol 180.
IPCS. (2002) Formaldehyde. Concise international chemical assessment document. Vol. 40.
World Health Organization, Geneva, Switzerland. Available online at
http://www.inchem.org/documents/cicads/cicads/cicad40.htm.
Iravani, J. (1974) Clearance function of the respiratory ciliated epithelium in normal and
bronchitic rats. Inhaled Part 1:143-148.
Iravani, J; Melville, GN. (1974) Long-term effect of cigarette smoke on mucociliary function in
animals. Respiration 31:358-366.
Ishidate, JM; Sofuni, T; and Yoshikawa, K. (1981) Chromosomal aberration tests in vitro as a
primary screening tool for environmental mutagens and/or carcinogens. GANN Monogr Cancer
Res 27:95-108.
Ito, K; Sakamoto, T; Hayashi, Y; et al. (1996) Role of tachykinin and bradykinin receptors and
mast cells in gaseous formaldehyde-induced airway microvascular leakage in rats. Eur J
Pharmacol 307:291-298.
Jaeger, RJ; Gearhart, JM. (1982) Respiratory and metabolic response of rats and mice to
formalin vapor. Toxicology 25:299-309.
Jaiswal, S; Traver, D; Miyamoto, T; et al. (2003) Expression of BCR/ABL and BCL-2 in
myeloid progenitors leads to myeloid leukemias. Proc Natl Acad Sci USA 100:10002-10007.
Jakab, GJ. (1992) Relationship between carbon black particulate-bound formaldehyde,
pulmonary antibacterial defenses, and alveolar macrophage phagocytosis. Inhal Toxicol 4:325-
342.
James, JT; Limero, TF; Beck, SW; et al. (2002) Environmental Monitoring Air Quality. Report
of the Lunar-Mars Support Test Project (LMLSTP4.1). In: Isolation: NASA Experiments in
Closed Environment Living. Eds: Lane, HW; Sauer, RL; Feeback, DL. Univeit, Inc, San Diego,
CA. 104:177-191. Data also available at National Aeronautics and Space Administration Life
Sciences Data Archive, Houston, TX. Available at
http://lsda.jsc.nasa. gov/docs/research/research_detail.cfm?experiment_type_code=24&researcht
ype=.
Jarabek, AM; Fisher, JW; Rubenstein, R; et al. (1994) Mechanistic insights aid the search for
CFC substitutes: risk assessment of HCHC-123 as an example. Risk Anal 14:231-250.
Javdan, M; Taher, TE. (2000) The short and long term effects of formaldehyde vapour (2 ppm
and 5 ppm) on rat nasal epithelium. Pathol Int 50:A77.
This document is a draft for review purposes only and does not constitute Agency policy.
R-27 DRAFT—DO NOT CITE OR QUOTE
-------
Jeffcoat, AR; Chasalow, F; Feldman, DB; et al. (1983) Disposition of [14C] formaldehyde after
topical exposure to rats, guinea pigs, and monkeys. In: Gibson, J E , eds. Formaldehyde
Toxicity. Washington, DC: Hemisphere; pp. 38-50.
Jensen, DE; Belka, GK; Du Bois, GC. (1998) S-Nitrosoglutathione is a substrate for rat alcohol
dehydrogenase class III isoenzyme. Biochem J 331 ( Pt 2):659-668.
Johannsen, FR; Levinskas, GJ; Tegeris, AS. (1986) Effects of formaldehyde in the rat and dog
following oral exposure. Toxicol Lett 30:1-6.
John, EM; Savitz, DA; Shy, CM. (1994) Spontaneous abortions among cosmetologists.
Epidemiology 5:147-155.
Joos, GF. (2003) Bronchial hyperresponsiveness: too complex to be useful? Curr Opin
Pharmacol 3:233-238.
Jornvall, H; Hoog, JO; Persson, B; et al. (2000) Pharmacogenetics of the alcohol dehydrogenase
system. Pharmacology 61:184-191.
Junqueira, LC; Carneiro, J; Kelley, RO. (1992) Basic Histology. Appleton and Lange, Norwalk,
CN.
Jung, R; Engelhart, G; Herbolt, B; et al. (1992) Collaborative study of mutagenicity with
Salmonella typhimurium TA102. Mutat Res 278:265-270.
Jurvelin, J; Vartiainen, M; Jantunen, M; et al. (2001) Personal exposure levels and
microenvironmental concentrations of formaldehyde and acetaldehyde in the Helsinki
metropolitan area. J Air Waste Manag Assoc 51:17-24.
K S Crump Group. (2001) TOX RISK dose-response software, version 5.2 (2000-2001). KS
Crump Group.
Kakeyama, M; Kurokawa, Y; Kunugita, N; et al. (2004) Low dose exposure to formaldehyde
alters glutamate NMDA receptor mRNA expressions in the rat neocortex. Neurotoxicology
25:717.
Kalasz, H. (2003) Biological role of formaldehyde, and cycles related to methylation,
demethylation, and formaldehyde production. Mini Rev Med Chem 3:175-192.
Kallen, RG; Jencks, WP. (1966a) The dissociation constants of tetrahydrofolic acid. J Biol Chem
241:5845-5850.
Kallen, RG; Jencks, WP. (1966b) The mechanism of the condensation of formaldehyde with
tetrahydrofolic acid. J Biol Chem 241:5851-5863.
Kamata, E; Nakadate, M; Ogawa, Y; et al. (1996a) Acute inhalation toxicity study of
formaldehyde in rats: effect of vapor on the pulmonary surfactant. Oyo Yakuri 51:33-37.
Kamata, E; Nakadate, M; Uchida, O; et al. (1996b) Effects of formaldehyde vapor on the nasal
cavity and lungs of F-344 rats. J Environ Pathol Toxicol Oncol 15:1-8.
Kamata, E; Nakadate, M; Uchida, O; et al. (1997) Results of a 28-month chronic inhalation
toxicity study of formaldehyde in male Fisher-344 rats. J Toxicol Sci 22:239-254.
This document is a draft for review purposes only and does not constitute Agency policy.
R-28 DRAFT—DO NOT CITE OR QUOTE
-------
Kane, LE; Alarie, Y. (1977) Sensory irritation to formaldehyde and acrolein during single and
repeated exposures in mice. Am Ind Hyg Assoc J 38:509-522.
Kane, LE; Alarie, Y. (1978) Evaluation of sensory irritation from acrolein-formaldehyde
mixtures. Am Ind Hyg Assoc J 39:270-274.
Kane, LE; Barrow, CS; Alarie, Y. (1979) A short-term test to predict acceptable levels of
exposure to airborne sensory irritants. Am Ind Hyg Assoc J 40:207-229.
Kedderis, GL; Hollenberg, PF. (1983) Steady state kinetics of chloroperoxidase-catalyzed N-
demethylation reactions. J Biol Chem 258:12413-12419.
Kelly, JT; Prasad, AK; Wexler, AS. (2000) Detailed flow patterns in the nasal cavity. J Appl
Physiol 89:323-337.
Kennedy, G; Slaich, PK; Golding, BT; et al. (1996) Structure and mechanism of formation of a
new adduct from formaldehyde and guanosine. Chem Biol Interact 102:93-100.
Kepler, GM; Richardson, RB; Morgan, KT; et al. (1998) Computer simulation of inspiratory
nasal airflow and inhaled gas uptake in a rhesus monkey. Toxicol Appl Pharmacol 150:1-11.
Kernan, GJ; Ji, BT; Dosemeci, M; et al. (1999) Occupational risk factors for pancreatic cancer:
A case-control study based on death certificates from 24 U.S. states. Am J Ind Med 36:260-270.
Kerns, WD; Pavkov, KL; Donofrio, DJ; et al. (1983) Carcinogenicity of formaldehyde in rats
and mice after long-term inhalation exposure. Cancer Res 43:4382-4392.
Khamgaonkar, MB; Fulare, MB. (1991) Pulmonary effects of formaldehyde exposure—an
environmental-epidemiological study. Ind J Chest Dis Allied Sci 33:9-13.
Kiernan, JA. (2000) Formaldehyde, formalin, paraformaldehyde and glutaraldehyde: What they
are and what they do. Microscopy Today 00-1 pp. 8-12.
Kilburn, KH. (1994) Neurobehavioral impairment and seizures from formaldehyde. Arch
Environ Health 49:37-44.
Kilburn, KH; Moro, A. (1985) Reproductive and maternal effects of formaldehyde (HCH0) in
rats. Fed Proc 44:535.
Kilburn, KH; Warshaw, R. (1992) Neurobehavioral effects of formaldehyde and solvents on
histology technicians: repeated testing across time. Environ Res 58:134-146.
Kilburn, KH; Seidman, BC; Warshaw, R. (1985) Neurobehavioral and respiratory symptoms of
formaldehyde and xylene exposure in histology technicians. Arch Environ Health 40:229-233.
Kilburn, KH; Warshaw, R; Thornton, JC. (1987) Formaldehyde impairs memory equilibrium and
dexterity in histology technicians effects which persist for days after exposure. Arch Environ
Health 42:117-120.
Kim, CW; Song, JS; Ahn, YS; et al. (2001) Occupational asthma due to formaldehyde. Yonsei
Med J 42:440-445.
Kim, KC; Hisatsune, A; Kim, DJ; et al. (2003) Pharmacology of airway goblet cell mucin
release. J. Pharmacol Sci 92:301-307.
This document is a draft for review purposes only and does not constitute Agency policy.
R-29 DRAFT—DO NOT CITE OR QUOTE
-------
Kimbell, JS; Gross, EA; Joyner, DR; et al. (1993) Application of computational fluid dynamics
to regional dosimetry of inhaled chemicals in the upper respiratory tract of the rat. Toxicol Appl
Pharmacol 121:253-263.
Kimbell, JS; Godo, MN; Gross, EA; et al. (1997a) Computer simulation of inspiratory airflow in
all regions of the F344 rat nasal passages. Toxicol Appl Pharmacol 145:388-398.
Kimbell, JS; Gross, EA; Richardson RB; et al. (1997b) Correlation of regional formaldehyde
flux predictions with the distribution of formaldehyde-induced squamous metaplasia in F344 rat
nasal passages. MutatRes 380:143-154.
Kimbell, JS; Subramaniam, RP; Gross, EA; et al. (2001a) Dosimetry modeling of inhaled
formaldehyde: comparisons of local flux predictions in the rat, monkey, and human nasal
passages. Toxicol Sci 64:100-110.
Kimbell, JS; Overton, JH; Subramaniam, RP; et al. (2001b) Dosimetry modeling of inhaled
formaldehyde: binning nasal flux predictions for quantitative risk assessment. Toxicol Sci
64:111-121.
Kimbell, JS. (2006) Nasal dosimetry of inhaled gases and particles: Where do inhaled agents go
in the nose? Toxicol Pathol 34:270-273.
Kinzler, KW; Vogelstein, B. (1997) Cancer-susceptibility genes. Gatekeepers and caretakers.
Nature 386:761-763.
Kitaev, EM; Savchenko, O; Lovchikov, VA; et al. (1984) Embryonic development and various
indicators of reproductive function in rats after inhalation of formaldehyde in the preimplantation
phase [translation], Akush Ginekol 10:49-52.
Kitaeva, LV; Kitaev, E; Pimenova, MN. (1990) Cytopathic and cytogenetic effects of chronic
inhalation of formaldehyde on the female rat's germ marrow cells [translation], Tsitologiia
32:1212-1216.
Kitaeva, LV; Mikheeva, EA; Shelomova, LF; et al. (1996) Genotoxic effect of formaldehyde in
somatic human cells in vivo [translation], Genetika 32(9): 1287-1290. Translated from Russian
by Ted Crump, Translator. NIH Library Translation. National Institutes of Health, Public Health
Service, U.S. Department of Health and Human Services; NIH-03-640.
Klaassen, CD. (2001) Casarett and Doull's Toxicology: The Basic Science of Poisons. 6th
Edition. McGraw-Hill, New York, NY.
Klein, M; Sinha, B; Subramaniam, RP. (2009) Statistical inferences from formaldehyde DNA-
protein cross-link data: Improving methods for characterization of uncertainty. J.
Biopharmaceutical Statistics. [Submitted]
Kligerman, AD; Phelps, MC; Erexson, GL. (1984) Cytogenetic analysis of lymphocytes from
rats following formaldehyde inhalation. Toxicol Lett 21:241-246.
Klonisch, T; Steger, K; Kehlen, A; et al. (2004) INSL-3 ligand-receptor system in the equine
testis. Biol Reprod 68:1975-1981.
This document is a draft for review purposes only and does not constitute Agency policy.
R-30 DRAFT—DO NOT CITE OR QUOTE
-------
Koppel, C; Tenczer, J; Peixoto-Menezes, KM. (1991) Formation of formaldehyde adducts from
various drugs by use of methanol in a toxicological screening procedure with gas
chromatography-mass spectrometry. J Chromatogr 563:73-81.
Korczynski, RE. (1994) Formaldehyde exposure in the funeral industry. Appl Occup Environ
Hyg 9:575-579.
Krakowiak, A; Gorski, P; Pazdrak, K; et al. (1998) Airway response to formaldehyde inhalation
in asthmatic subjects with suspected respiratory formaldehyde sensitization. Am J Ind Med
33:274-281.
Kramps, JA; Peltenburg, LT; Kerklaan, PR; et al. (1989) Measurement of specific IgE antibodies
in individuals exposed to formaldehyde. Clin Exp Allergy 19:509-514.
Krewski, D; Crump, KS; Farmer, J; et al. (1983) A comparison of statistical methods for low
dose extrapolation utilizing time-to-tumor data. Fundam Appl Toxicol 3:140-160.
Kriebel, D; Sama, SR; Cocanour, B. (1993) Reversible pulmonary responses to formaldehyde. A
study of clinical anatomy students. Am Rev Respir Dis 148:1509-1515.
Kriebel, D; Myers, D; Cheng, M; et al. (2001) Short-term effects of formaldehyde on peak
expiratory flow and irritant symptoms. Arch Environ Health 56(1): 11-18.
Kreiger, RA; and Garry, VF. (1983) Formaldehyde-induced cytotoxicity and sister-chromatid
exchanges in human lymphocyte cultures. MutatRes 120:51-55.
Krimsky, MM. (1986) Letter to the editor regarding Sallakumar et al., 1985. Toxicol Appl
Pharmacol 85:295.
Krzyzanowski, M; Quackenboss, JJ; Lebowitz, MD. (1990) Chronic respiratory effects of indoor
formaldehyde exposure. Environ Res 52:117-125.
Kuby, J. (1991) Immunology. 2nd edition. New York, NY: W.H. Freeman and Company.
Kulle, TJ. (1993) Acute odor and irritation response in healthy nonsmokers with formaldehyde
exposure. Inhal Toxicol 5:323-332.
Kulle, TJ; Cooper, GP. (1975) Effects of formaldehyde and ozone on the trigeminal nasal
sensory system. Arch Environ Health 30:237-243.
Kulle, TJ; Sauder, LR; Hebel, R; et al. (1987) Formaldehyde dose-response in healthy
nonsmokers. JAPCA 37:919-924.
Kum, C; Kiral, F; Sekkin, S; et al. (2007) Effects of xylene and formaldehyde inhalations on
oxidative stress in adult and developing rats livers. Exp Anim 56:35-42.
Kuper, CF; Hameleers, DMH; Bruijntjes, JP; et al. (1990) Lymphoid and non-lymphoid cells in
nasal-associated lymphoid tissue (NALT) in the rat. Cell Tissue Res 259:371-377.
Kuppers, R; Klein, U; Hansmann, M-L; et al. (1999) Cellular origin of human B-cell
lymphomas. New Eng J Med 341:1520-1529.
Kuschner, M; Laskin, S; Drew, RT; et al. (1975) Inhalation carcinogenicity of alpha halo ethers.
III. Lifetime and limited period inhalation studies with bis(chloromethyl)ether at 0.1 ppm. Arch
Environ Health 30:73-77.
This document is a draft for review purposes only and does not constitute Agency policy.
R-31 DRAFT—DO NOT CITE OR QUOTE
-------
Laforest, L; Luce, D; Goldberg, P; et al. (2000) Laryngeal and hypopharyngeal cancers and
occupational exposure to formaldehyde and various dusts: a case-control study in France. Occup
Environ Med 57:767-773.
Lam, CW; Casanova, M; Heck, HD. (1985) Depletion of nasal mucosal glutathione by acrolein
and enhancement of formaldehyde-induced DNA-protein cross-linking by simultaneous
exposure to acrolein. Arch Toxicol 58:67-71.
Lang, I; Bruckner, T; Triebig, G. (2008) Formaldehyde and chemosensory irritation in humans:
A controlled human exposure study. Regul Toxicol Pharmacol 50:23-36.
Lash, LH; Parker, JC. (2001) Hepatic and renal toxicities associated with perchloroethylene.
Pharmacol Rev 53:177-208.
Laskin, S; Kuschner, M; Drew, RT; et al. (1971) Tumors of the respiratory tract induced by
inhalation of bis(chloromethyl)ether. Arch Environ Health 23:135-136.
Lazutka, JR; Lekevicius, R; Dedonyte, V, et al. (1999) Chromosomal aberrations and sister-
chromatid exchanges in Lithuanian populations: effects of occupational and environmental
exposures. Mutat Res 445:225-239.
Le Curieux, F; Marzin, D; Erb, F. (1993) Comparison of three short-term assays: results on
seven chemicals. Potential contribution to the control of water genotoxicity. Mutat Res
319:223-236.
Leach, C; Sharma, R; Oberg, S. (1983) Immunotoxic and pathologic effects of inhalation of
various levels of formaldehyde vapor in F-344 rats. Toxicologist 3:73.
Lee, H; Alarie, Y; Karol, M. (1984) Induction of formaldehyde sensitivity in guinea pigs.
Toxicol Appl Pharmacol 75:147-155.
Lee, MH; Kim, YA; Na, TY; et al. (2008) Identification of formaldehyde-responsive genes by
suppression subtractive hybridization. Toxicology 243:224-235.
Lemiere, C; Desjardins, A; Cloutier, Y; et al. (1995) Occupational asthma due to formaldehyde
resin dust with and without reaction to formaldehyde gas. Eur Respir J 8:861-865.
Levine, RJ; Andjelkovich, DA; Shaw, L. (1984) The mortality of Ontario undertakers and a
review of formaldehyde-related mortality studies. J Occup Med 26:740-746.
Levy, S; Nocentini, S; and Billardon, C. (1983) Induction of cytogenetic effects in human
fibroblast cultures after exposure to formaldehyde or X-rays. Mutat Res 119:309-317.
Li, G-Y; Lee, H-Y; Shin, H-S; et al. (2007) Identification of gene markers for formaldehyde
exposure in humans. Environ Health Perspect 115:1460-1466.
Li, R; Lu, ZS; Qiao, Y; et al. (2004) Study on the formaldehyde-induced DNA damage with
comet assay. Shi Yan Sheng Wu Xue Bao 37:262-268.
Liber, HL; Benforado, K; Crosby, RM; et al. (1989) Formaldehyde-induced and spontaneous
alterations in human hprt DNA sequence and mRNA expression. Mutat Res 226:31-37.
Liden, S; Scheynius, A; Fischer, T; et al. (1993) Absence of specific IgE antibodies in allergic
contact sensitivity to formaldehyde. Allergy 48(7):525-529.
This document is a draft for review purposes only and does not constitute Agency policy.
R-32 DRAFT—DO NOT CITE OR QUOTE
-------
Liebling, T; Rosenman, KD; Griffith, RG; et al. (1984) Cancer mortality among workers
exposed to formaldehyde. Am J Ind Med 5:423-428.
Liesivuori, J; Savolainen, H. (1987) Urinary formic acid as an indicator of occupational exposure
to formic acid and methanol. Am Ind Hyg Assoc J 48:32-34.
Lin, Z; Luo, W; Li, H; et al. (2005) The effect of endogenous formaldehyde on the rat aorta
endothelial cells. Toxicol Lett 159:134-143.
Lindbohm, ML; Hemminki, K; Bonhomme, MG; et al. (1991) Effects of paternal occupational
exposure on spontaneous abortions. Am J Public Health 81:1029-1033.
Lino dos Santos Franco, A; Damazo, AS; Beraldo de Souza, HR; et al. (2006) Pulmonary
neutrophil recruitment and bronchial reactivity in formaldehyde-exposed rats are modulated by
mast cells and differentially by neuropeptides and nitric oxide. Toxicol Appl Pharmacol 214:35-
42.
Lino dos Santos Franco, A; Domingos, HV; Damazo, AS; et al. (2009) Reduced allergic lung
inflammation in rats following formaldehyde exposure: long term effects on multiple effector
systems. Toxicology. 256: 157-163.
Liu, KS; Huang, FY; Hay ward, SB; et al. (1991) Irritant effects of formaldehyde exposure in
mobile homes. Environ Health Perspect 94:91-94.
Liu, L; Hausladen, A; Zeng, M; et al. (2001) A metabolic enzyme for S-nitrosothiol conserved
from bacteria to humans. Nature 410:490-494.
Liu, L; Yan, Y; Zeng, M; et al. (2004) Essential roles of S-nitrosothiols in vascular homeostasis
and endotoxic shock. Cell 116:617-628.
Liu, Y; Li, CM; Lu, Z; et al. (2006) Studies on formation and repair of formaldehyde-damaged
DNA by detection of DNA-protein crosslinks and DNA breaks. Front Biosci 11:991-997.
Logue, J; Barrick, M; Jessup, G. (1986) Mortality of radiologists and pathologists in the
radiation registry of physicians. J Occup Med 28:91-99.
Lu, Z; Li, CM; Qiao, Y; et al. (2008) Effect of inhaled formaldehyde on learning and memory in
mice. Indoor Air 18:77-83.
Luce, D; Gerin, M; Leclerc, A; et al. (1993) Sinonasal cancer and occupational exposure to
formaldehyde and other substances. Int J Cancer 53:224-231.
Luce, D; Leclerc, A; Begin, D; et al. (2002) Sinonasal cancer and occupational exposures: a
pooled analysis of 12 case-control studies. Cancer Causes Control 13:147-157.
Luebeck, EG; Grasl-Kraupp, B; Timmermann-Trosiener, I; et al. (1995) Growth kinetics of
enzyme-altered liver foci in rats treated with phenobarbital or alpha-hexachlorocyclohexane.
Toxicol Appl Pharmacol 130:304-315.
Luebeck, EG; Buchmann, A; Stinchcombe, S; et al. (2000) Effects of 2,3,7,8-tetrachlorodibenzo-
p-dioxin on initiation and promotion of GST-P-positive foci in rat liver: A quantitative analysis
of experimental data using a stochastic model. Toxicol Appl Pharmacol 167:63-73.
This document is a draft for review purposes only and does not constitute Agency policy.
R-33 DRAFT—DO NOT CITE OR QUOTE
-------
Luo, W; Li, H; Zhang, Y; et al. (2001) Determination of formaldehyde in blood plasma by high-
performance liquid chromatography with fluorescence detection. J Chromatogr B Biomed Sci
Appl 753:253-257.
Lyapina, M; Zhelezova, G; Petrova, E; et al. (2004) Flow cytometric determination of neutrophil
burst activity in workers exposed to formaldehyde. Int Arch Occup Environ Health 77:335-340.
Ma, TH; Harris, MM. (1988) Review of the genotoxicity of formaldehyde. Mutat Res 196:37-
59.
Mackerer, CR; Angelosanto, FA; Blackburn, GR, et al. (1996) Identification of formaldehyde as
the metabolite responsible for the mutagenicity of methyl tertiary-butyl ether in the activiated
mouse lymphoma assay. Proc. Soc. exp. Biol. Med 212:338-341.
Magana-Schwencke, N; and Ekert, B. (1978) Biochemical analysis of damage induced in yeast
by formaldehyde. II. Induction of cross-links between DNA and protein. Mutat Res 51:11-19.
Main, D; Hogan, TJ. (1983) Health effects of low level exposure to formaldehyde. J Occup Med
25(12):896-900.
Majumder, PK; Kumar, VL. (1995) Inhibitory effects of formaldehyde on the reproductive
system of male rats. Indian J Physiol Pharmacol 39:80-82.
Malaka, T; Kodama, AM. (1990) Respiratory health of plywood workers occupationally exposed
to formaldehyde. Arch Environ Health 45:288-294.
Malek, FA; Moritz, KU; Fanghanel, J. (2003a) A study on the effect of inhalative formaldehyde
exposure on water labyrinth test performance in rats. Ann Anat 185:277-285.
Malek, FA; Moritz, KU; Fanghanel, J. (2003b) Formaldehyde inhalation and open field
behaviour in rats. Indian J Med Res 118:90-96.
Malek, FA; Moritz, KU; Fanghanel, J. (2003c) A study on specific behavioral effects of
formaldehyde in the rat. J Exp Anim Sci 42:160-170.
Malek, FA; Moritz, KU; Fanghanel, J. (2004) Effects of a single inhalative exposure to
formaldehyde on the open field behavior of mice. Int J Hyg Environ Health 207:151-158.
Mannix, RC; Phalen, RF; Walters, RF; et al. (1983) Effects of sulfur dioxide and formaldehyde
on particle clearance in the rat. J Toxicol Environ Health 12:429-440.
Marinari, UM; Ferro, M; Sciaba, L, et al. (1984) DNA-damaging activity of biotic and
xenobiotic aldehydes in Chinese hamster ovary cells. Cell Biochem Funct 2:243-248.
Marks, TA; Worthy, WC; Staples, RE. (1980) Influence of formaldehyde and sonacide
(potentiated acid glutaraldehyde) on embryo and fetal development in mice. Teratology 22:51-
58.
Marnett, LJ; Hurd, HK; Hollstein, MC; et al. (1985) Naturally occurring carbonyl compounds
are mutagens in Salmonella tester strain TA104. Mutat Res 148:25-34.
Maronpot, RR; Miller, RA; Clarke, WJ; et al. (1986) Toxicity of formaldehyde vapor in B6C3F1
mice exposed for 13 weeks. Toxicology 41:253-266.
This document is a draft for review purposes only and does not constitute Agency policy.
R-34 DRAFT—DO NOT CITE OR QUOTE
-------
Marsh, GM. (1982) Proportional mortality patterns among chemical plant workers exposed to
formaldehyde. Br J Ind Med 39:313-322.
Marsh, GM; Youk, AO. (2005) Reevaluation of mortality risks for nasopharyngeal cancer in the
formaldehyde cohort study of the National Cancer Institute. Regul Pharmacol Toxicol 42:275-
283.
Marsh, GM; Stone, RA; Henderson, VL. (1992a) A reanalysis of the National Cancer Institute
study on lung cancer mortality among industrial workers exposed to formaldehyde. J Occup
Med 34:42-44.
Marsh, GM; Stone, RA; Henderson, VL. (1992b) Lung cancer mortality among industrial
workers exposed to formaldehyde: a Poisson regression analysis of the National Cancer Institute
Study. Am Ind Hyg Assoc J 53:681-691.
Marsh, GM; Stone, RA; Esmen, NA; et al. (1994) Mortality patterns among chemical plant
workers exposed to formaldehyde and other substances. J Natl Cancer Inst 86:384-386.
Marsh, GM; Stone, RA; Esmen, NA; et al. (1996) Mortality among chemical workers in a
factory where formaldehyde was used. Occup Environ Med 53:613-627.
Marsh, GM; Youk, AO; Buchanich, JM; et al. (2002) Pharyngeal cancer mortality among
chemical plant workers exposed to formaldehyde. Toxicol Ind Health 18:257-268.
Marsh, GM; Youk, AO; Buchanich, JM; et al. (2007a) Work in the metal industry and
nasopharyngeal cancer mortality among formaldehyde-exposed workers. Regul Toxicol
Pharmacol 48:308-319.
Marsh, GM; Youk, AO; Morfeld, P. (2007b) Mis-specified and non-robust mortality risk models
for nasopharyngeal cancer in the National Cancer Institute formaldehyde worker cohort study.
Regul Toxicol Pharmacol 47:59-67.
Martin, CN; McDermid, AC; and Garner, RC. (1978) Testing of known carcinogens and
noncarcinogens for their ability to induce unscheduled DNA synthesis in HeLa cells. Cancer Res
38:2621-2627.
Martin, WJ. (1990) A teratology study of inhaled formaldehyde in the rat. Reprod Toxicol
4:237-239.
Mashford, PM; Jones, AR. (1982) Formaldehyde metabolism by the rat: a re-appraisal.
Xenobiotica 12:119-124.
Matanoski, GM. (1989) Risks of pathologists exposed to formaldehyde. National Institute for
Occupational Safety and Health, Cincinnati, OH. Available from the National Technical
Information Service, Springfield, VA; PB91-173682.
Matsunaga, I; Miyake, Y; Yoshida, T; et al. (2008) Ambient formaldehyde levels and allergic
disorders among Japanese women: baseline data from the Osaka maternal and child health study.
Ann Epidemiol 18:78-84.
Mauderly, JL. (1986) Respiration of F344 rats in nose-only inhalation exposure tubes. J Appl
Toxicol 6:25-30.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 5 DRAFT—DO NOT CITE OR QUOTE
-------
Maurice, F; Rivory, JP; Larsson, PH; et al. (1986) Anaphylactic shock caused by formaldehyde
in a patient undergoing long-term hemodialysis. J Allergy Clin Immunol 77:594-597.
McGhee, JD; von Hippel, PH. (1975a) Formaldehyde as a probe of DNA structure. I. Reaction
with exocyclic amino groups of DNA bases. Biochemistry 14:1281-1296.
McGhee, JD; von Hippel, PH. (1975b) Formaldehyde as a probe of DNA structure. II. Reaction
with endocyclic imino groups of DNA bases. Biochemistry 14:1297-1303.
McGhee, JD; von Hippel, PH. (1977a) Formaldehyde as a probe of DNA structure. 3.
Equilibrium denaturation of DNA and synthetic polynucleotides. Biochemistry 16:3267-3276.
McGhee, JD; von Hippel, PH. (1977b) Formaldehyde as a probe of DNA structure, r.
Mechanism of the initial reaction of Formaldehyde with DNA. Biochemistry 16:3276-3293.
McKone, TE; Daniels, JI. (1991) Estimating human exposure through multiple pathways from
air, water, and soil. Regul Toxicol Pharmacol 13:36-61.
McLean, JH; Shipley, MT. (1988) Postmitotic, postmigrational expression of tyrosine
hydroxylase in olfactory bulb dopaminergic neurons. J Neurosci 8:3658-3669.
McMartin, KE; Martin-Amat, G; Noker, PE; et al. (1979) Lack of a role for formaldehyde in
methanol poisoning in the monkey. Biochem Pharmacol 28:645-649.
Meggs, WJ. (1993) Neurogenic inflammation and sensitivity to environmental chemicals.
Environ Health Perspect 101:234-238.
Meggs, WJ. (1995) Neurogenic switching: a hypothesis for a mechanism for shifting the site of
inflammation in allergy and chemical sensitivity. Environ Health Perspect 103:54-56.
Merk, O; Speit, G. (1998) Significance of formaldehyde-induced DNA-protein crosslinks for
mutagenesis. Environ Mol Mutagen 32:260-268.
Merk, O; and Speit, G. (1999) Detection of crosslinks with the comet assay in relationship to
genotoxicity and cytotoxicity. Environ Mol Mutagen 33:167-172.
Merletti, F; Boffetta, P; Ferro, G; et al. (1991) Occupation and cancer of the oral cavity or
oropharynx in Turin, Italy. Scand J Work Environ Health 17:248-254.
Mery, S; Gross, EA; Joyner, DR; et al. (1994) Nasal diagrams: a tool for recording the
distribution of nasal lesions in rats and mice. Toxicol Pathol 22:353-372.
Metz, B; Kersten, GF; Hoogerhout, P; et al. (2004) Identification of formaldehyde-induced
modifications in proteins: reactions with model peptides. J Biol Chem 279:6235-6243.
Metz, B; Kersten, GF; Baart, GJ; et al. (2006) Identification of formaldehyde-induced
modifications in proteins: reactions with insulin. Bioconjug Chem 17:815-822.
Migliore, L; Ventura, L; Barale, R; et al. (1989) Micronuclei and nuclear anomalies induced in
the gastro-intestinal epithelium of rats treated with formaldehyde. Mutagenesis 4:327-334.
Miretskaia, LV; and Shvartsman, P. (1982) [Chromosome damages in human lymphocytes as
affected by formaldehyde. I. Formaldehyde treatment of lymphocytes in culture], Tsitologiia
24:1056-1060.
This document is a draft for review purposes only and does not constitute Agency policy.
R-36 DRAFT—DO NOT CITE OR QUOTE
-------
Miyachi, T; Tsutsui, T. (2005) Ability of 13 chemical agents used in dental practice to induce
sister-chromatid exchanges in Syrian hamster embryo cells. Odontology 93:24-29.
Molotkov, A; Fan, X; Deltour, L; et al. (2002) Stimulation of retinoic acid production and
growth by ubiquitously expressed alcohol dehydrogenase Adh3. Proc Natl Acad Sci USA
99:5337-5342.
Monteiro-Riviere, NA; Popp, JA. (1986) Ultrastructural evaluation of acute nasal toxicity in the
rat respiratory epithelium in response to formaldehyde gas. Fundam Appl Toxicol 6:251-262.
Monticello, TM; Morgan, KT. (1994) Cell proliferation and formaldehyde-induced respiratory
carcinogenesis. Risk Anal 14:313-319.
Monticello, TM; Morgan, KT. (1997) Chemically induced nasal carcinogenesis and epithelial
cell proliferation: a brief review. Mutat Res 380:33-41.
Monticello, TM; Morgan, KT; Everitt, JI; et al. (1989) Effects of formaldehyde gas on the
respiratory tract of rhesus monkeys. Pathology and cell proliferation. Am J Pathol 134:515-527.
Monticello, TM; Morgan, KT; Hurtt, ME. (1990) Unit length as the denominator for quantitation
of cell proliferation in nasal epithelia. Toxicol Pathol 18:24-31.
Monticello, TM; Miller, FJ; Morgan, KT. (1991) Regional increases in rat nasal epithelial cell
proliferation following acute and subchronic inhalation of formaldehyde. Toxicol Appl
Pharmacol 111:409-421.
Monticello, TM; Swenberg, JA; Gross, EA; et al. (1996) Correlation of regional and nonlinear
formaldehyde-induced nasal cancer with proliferating populations of cells. Cancer Res 56:1012—
1022.
Moolgavkar, SH. (1994) Biological models of carcinogenesis and quantitative cancer risk
assessment. Risk Anal 14:879-882.
Moolgavkar, SH; Venzon, DJ. (1979) Two-event models for carcinogenesis: Incidence curves
for childhood and adult tumors. Math Biosci 47:55-77.
Moolgavkar, SH; Knudson, AJ. (1981) Mutation and cancer: A model for human carcinogenesis.
J Natl Cancer Inst 66:1037-1052.
Moolgavkar, SH; Luebeck, EG. (1992) Interpretation of labeling indices in the presence of cell
death. Carcinogenesis 13:1007-1010.
Moolgavkar, SH; Dewanji, A; Venzon, D. (1988) A stochastic two-stage model for cancer risk
assessment. I. The hazard function and the probability of tumor. Risk Anal 8:383-392.
Moolgavkar, SH; Luebeck, EG; Krewski, D; et al. (1993) Radon, cigarette smoke, and lung
cancer: A reanalysis of the Colorado plateau uranium miner's data. Epidemiology 4:204-217.
Moore, L; Ogrodnik, EC. (1986) Occupational exposure to formaldehyde in mortuaries. J
Environ Health 49:32-35.
Moore, S; Montane-Jaime, K; Shafe, S; et al. (2007) Association of ALDH1 promoter
polymorphisms with alcohol-related phenotypes in Trinidad and Tobago. J Stud Alcohol Drugs
68:192-196.
This document is a draft for review purposes only and does not constitute Agency policy.
R-37 DRAFT—DO NOT CITE OR QUOTE
-------
Morgan, KT. (1997) A brief review of formaldehyde carcinogenesis in relation to rat nasal
pathology and human health risk assessment. Toxicol Pathol 25:291-307.
Morgan, KT; Patterson, DL; Gross, EA (1983) Localization of areas of inhibition of nasal
mucociliary function in rats following in vivo exposure to formaldehyde. Am Rev Respir Dis
127:166-166.
Morgan, KT; Patterson, DL; Gross, EA. (1984a) Frog palate mucociliary apparatus: structure,
function, and response to formaldehyde gas. Fundam Appl Toxicol 4:58-68.
Morgan, KT; Jiang, X-Z; Patterson, DL; et al. (1984b) The nasal mucociliary apparatus. Am
Rev Respir Dis 130:275-281.
Morgan, KT; Gross, EA; Patterson, DL. (1986a) Distribution, progression, and recovery of acute
formaldehyde-induced inhibition of nasal mucociliary function in F-344 rats. Toxicol Appl
Pharmacol. 86(3):448-456.
Morgan, KT; Jiang, XZ; Starr, TB; et al. (1986b) More precise localization of nasal tumors
associated with chronic exposure of F-344 rats to formaldehyde gas. Toxicol Appl Pharmacol
82:264-271.
Morgan, KT; Patterson, DL; Gross, EA. (1986c) Responses of the nasal mucociliary apparatus of
F-344 rats to formaldehyde gas. Toxicol Appl Pharmacol 82:1-13.
Morgan, K; Patterson, DL; Gross, E. (1986d) Responses of the nasal mucociliary apparatus to
airborne irritants. In: Barrow, CS; ed. Toxicology of the nasal passages. Washington, DC:
Hemisphere Publishing Corporation; pp. 123-133.
Morgan, KT; Kimbell, JS; Monticello, TM; et al. (1991) Studies of inspiratory airflow patterns
in the nasal passages of the F344 rat and rhesus monkey using nasal molds: relevance to
formaldehyde toxicity. Toxicol Appl Pharmacol 110:223-240.
Moser, B; Bodrogi, F; Eibl, G; et al. (2005) Mass spectrometric profile of exhaled breath-field
study by PTR-MS. Respir Physiol Neurobiol 145:295-300.
Muller, W; Engelhart, G; Herbold, B; et al. (1993) Evaluation of mutagenicity testing with
Salmonella typhimurium TA102 in three different laboratories. Environ Health Perspect 101
Suppl 3:33-36.
Natarajan, AT; Darroudi, F; Bussman, CJ; et al. (1983) Evaluation of the mutagenicity of
formaldehyde in mammalian cytogenetic assays in vivo and vitro. Mutat Res 122:355-360.
NCHS (National Center for Health Statistics). (2002) United States Life Tables, 1999. National
Vital Statistics Reports, vol. 50, no. 6, March 21, pp. 7-8.
NCI (National Cancer Institute). (2003) SEER Cancer Statistics Review, 1975-2000. Bethesda,
MD. http://seer, cancer.gov/csr/1975_2000/.
NEG (Nordic Expert Group for Criteria Documentation of Health Risks from Chemicals);
DECOS (Dutch Expert Committee on Occupational Standards). (2003) 132. Formaldehyde.
Stockholm, Sweden: National Institute for Working Life. Available from
. 2003: 11.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 8 DRAFT—DO NOT CITE OR QUOTE
-------
Neri, M; Bonassi, S; Knudsen, LE, et al. (2006) Children's exposure to environmental pollutants
and biomarkers of genetic damage. I. Overview and critical issues. MutatRes 612:1-13.
Nielsen, GD. (1991) Mechanisms of activation of the sensory irritant receptor by airborne
chemicals. CritRev Toxicol 21:183-208.
Nielsen, GD; Hougaard, KS; Larsen, ST; et al. (1999) Acute airway effects of formaldehyde and
ozone in Balb/c mice. Hum Exp Toxicol 18:400-409.
Niinimaa, V; Cole, P; Mintz, S; et al. (1981) Oronasal distribution of respiratory airflow. Respir
Physiol 43:69-75.
Nilsson, JA; Hedberg, JJ; Vondracek, M; et al. (2004) Alcohol dehydrogenase 3 transcription
associates with proliferation of human oral keratinocytes. Cell Mol Life Sci 61:610-617.
Norback, D; Bjornsson, E; Janson, C; et al. (1995) Asthmatic symptoms and volatile organic
compounds, formaldehyde, and carbon dioxide in dwellings. Occup Environ Med 52(6):388-
395.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing
the process. Washington, DC: National Academy Press.
NRC (1988) Health risks of radon and other internally deposited alpha-emitters: BEIRIV.
Committee on the Biological Effects of Ionizing Radiations. Washington, DC: National
Academy Press.
NRC. (1993) Understanding child abuse and neglect. Panel on Research on Child Abuse and
Neglect. National Research Council, National academy Press, Washington, DC.
NTP (National Toxicology Program). (2002) Report on Carcinogens. Tenth Edition. Research
Triangle Park, NC: National Institute of Environmental Health Services, Public Health Service,
U.S. Department of Health and Human Services. Available from:
.
NTP. (2005) Control data from NTP Technical Report Series Nos. 210, 267, 272, 306, 311,314,
329, 346, 363, 371, 375, 377, 379, 380, 386, 421, 437, 440, 447, 448, 450, 451, 453, 454, 461,
462, 466, 467, 468, 471, 472, 475, 482, 486, 487, 490, 492, 499, 400, 507, 513, 515, 519 and
control data from the wollastonite calcium silicates study. U.S. Department of Health and
Human Services, NTP, Research Triangle Park, NC. Available at http://ntp-
app s. ni ehs. nih. gov/ntp. tox/index. cfm.
Neuss, S; and Speit, G. (2008) Further characterization of the genotoxicity of formaldehyde in
vitro by the sister chromatid exchange test and co-cultivation experiments. Mutagenesis 23:355-
7.
Nunn, A; Craigen, A; Darby shire, J; et al. (1990) Six year follow up of lung function in men
occupationally exposed to formaldehyde. Br J Ind Med 47:747-752.
Obe, G; and Beek, B. (1979) Mutagenic activity of aldehydes. Drug Alcohol Depend 4:91-94.
O'Connor, PM; and Fox, BW. (1987) Comparative studies of DNA cross-linking reactions
following methylene dimethanesulphonate and its hydrolytic product, formaldehyde. Cancer
Chemother Pharmacol 19:11-5.
This document is a draft for review purposes only and does not constitute Agency policy.
R-39 DRAFT—DO NOT CITE OR QUOTE
-------
Odeigah, PGC. (1997) sperm head abnormalities and dominant lethal effects of formaldehyde in
albino rats. Mutat Res 389:141-148.
O'Donovan, MR; Mee, CD. (1993) Formaldehyde is a bacterial mutagen in a range of
Salmonella and Escherichia indicator strains. Mutagenesis 8:577-581.
Oerstavik, D; and Hongslo, JK. (1985) Mutagenicity of endodontic sealers. Biomaterials 6:129-
132.
Ohba, Y; Morimitsu, Y; Watarai, A. (1979) Reaction of formaldehyde with calf-thymus
nucleohistone. Eur JBiochem 100:285-293.
Ohmichi, K; Komiyama, M; Matsuno, Y; et al. (2006) Relationship between exposure to
formaldehyde and immunoglobulin E (IgE) production during the gross anatomy laboratory. J
Health Sci 52/5:642-647.
Ohta, T; Watanabe-Akanuma, M; Tokishita, S; et al. (1999) Mutation spectra of chemical
mutagens determined by Lac+ reversion assay with Escherichia coli WP3101P-WP3106P tester
strains. Mutat Res 440:59-74.
Ohta, T; Watanabe-Akanuma, M; Yamagata, H. (2000) A comparison of mutation spectra
detected by the Escherichia coli lac(+) reversion assay and the Salmonella typhimurium his(+)
reversion assay. Mutagenesis 15:317-23.
Ohtani, T; Mizuashi, M; Nakagawa, S; et al. (2004a) Diesel exhaust particles and formaldehyde
deviate human immune system to Th2-dominant pattern by affecting both T cells and dendritic
cells. J Invest Dermatol 122: A121.
Ohtani, T; Kurosawa, W; Mizuashi, M; et al. (2004b) Effects of diesel exhaust particles (DEP)
and formaldehyde on human monocyte-derived dendritic cells (MoDCs) and T cells. J Dermatol
Sci 34:134.
Ohtsuka, R; Shuto, Y; Fujie, H; et al. (1997) Response of respiratory epithelium of BN and F344
rats to formaldehyde inhalation. Exp Anim 46:279-286.
Ohtsuka, R; Shutoh, Y; Fujie, H; et al. (2003) Rat strain difference in histology and expression
of Thl- and Th2-related cytokines in nasal mucosa after short-term formaldehyde inhalation.
Exp Toxicol Pathol 54:287-291.
Olin, KL; Cherr, GN; Rifkin, E, et al. (1996) The effects of some redox-active metals and
reactive aldehydes on DNA-protein cross-links in vitro. Toxicology 110:1-8.
Olsen, JH; Dossing, M. (1982) Formaldehyde induced symptoms in day care centers. Am Ind
Hyg Assoc J 43:366-370.
Olsen, JH; Asnaes, S. (1986) Formaldehyde and the risk of squamous cell carcinoma of the
sinonasal cavities. Br J Ind Med 43:769-774.
Olsen, JH; Jensen, SP; Hink, M; et al. (1984) Occupational formaldehyde exposure and
increased nasal cancer risk in man. Int J Cancer 34:639-644.
Olsen, GW; Lacy, SE; Bodner, KM; et al. (1997) Mortality from pancreatic and lymphopoietic
cancer among workers in ethylene and propylene chlorohydrin production. Occup Environ Med
54:592-598.
This document is a draft for review purposes only and does not constitute Agency policy.
R-40 DRAFT—DO NOT CITE OR QUOTE
-------
Orsiere, T; Sari-Minodier, I; Iarmarcovai, G; et al. (2006) Genotoxic risk assessment of
pathology and anatomy laboratory workers exposed to formaldehyde by use of personal air
sampling and analysis of DNA damage of peripheral lymphocytes. Mutat Res 605:30-41.
Orstavik, D; Hongslo, JK. (1985) Mutagenicity of endodontic sealers. Biomaterials 6:129-132.
Oshino, N; Chance, B; Sies, H; et al. (1973) The role of H202 generation in perfused rat liver
and the reaction of catalase compound I and hydrogen donors. Arch Biochem Biophys 154:117-
131.
Ostojic, L; Bradaric, A; Mise, K; et al. (2006) Pulmonary function in persons who are
professionally exposed to formaldehyde fumes. Coll Antropol 30(3):507-511.
Overman, DO. (1985) Absence of embryotoxic effects of formaldehyde after percutaneous
exposure in hamsters. Toxicol Lett 24:107-110.
Overton, JH; Kimbell, JS; Miller, FJ. (2001) Dosimetry modeling of inhaled formaldehyde: the
human respiratory tract. Toxicol Sci 64:122-134.
0vreb0, S; Haugen, A; Skaug, V. (2002) Biotransformation of formaldehyde in cultured human
bronchus. Environ Res 89:38-42.
Owen, BA; Dudney, CS; Tan, E; et al. (1990a) Comparative Hazard Evaluation, An Approach
To Regulation:Formaldehyde in Drinking Water. Health and Safety Research Division, Oak
Ridge National Laboratory
Owen, BA; Dudney, CS; Tan, EL; et al. (1990b) Formaldehyde in drinking water: comparative
hazard evaluation and an approach to regulation. Regul Toxicol Pharmacol 11:220-236.
Ozen, OA; Yaman, M; Sarsilmaz, M; et al. (2002) Testicular zinc, copper and iron
concentrations in male rats exposed to subacute and subchronic formaldehyde gas inhalation. J
Trace Elem Med Biol 16:119-122.
Ozen, OA; Songur, A; Sarsilmaz, M; et al. (2003a) Changes of zinc, copper, and iron levels in
the lung of male rats after subacute (4-week) and subchronic (13-week) exposure to
formaldehyde. J Trace Elem Exp Med 16:67-74.
Ozen, OA; Songur, A; Sarsilmaz, M; et al. (2003b) Zinc, copper and iron concentrations in
cerebral cortex of male rats exposed to formaldehyde inhalation. J Trace Elem Med Biol
17:207-209.
Ozen, OA; Akpolat, N; Songur, A; et al. (2005) Effect of formaldehyde inhalation on Hsp70 in
seminiferous tubules of rat testes: An immunohistochemical study. Toxicol Ind Health 21:249-
254.
Palczynski, C; Krakowiak, A; Hanke, W; et al. (1999) Indoor formaldehyde exposure and airway
allergic diseases. Int Rev Allergol Clin Immunol 5:65-69.
Park, H-S; Francis, KP; Yu, J; et al. (2003) Membranous cells in nasal-associated lymphoid
tissue: a portal of entry for the respiratory mucosal pathogen group A streptococcus. J Immunol
171:2532-2537.
Partanen, T; Kauppinen, T; Hernberg, S; et al. (1990) Formaldehyde exposure and respiratory
cancer among woodworkers: an update. Scand J Work Environ Health 16:394-400.
This document is a draft for review purposes only and does not constitute Agency policy.
R-41 DRAFT—DO NOT CITE OR QUOTE
-------
Passegue, E; Jamieson, CH; Ailles, LE; et al. (2003) Normal and leukemic hematopoiesis: are
leukemias a stem cell disorder or a reacquisition of stem cell characteristics? Proc Natl Acad Sci
U.S.A. 100(Suppl 1): 11842-11849.
Pauluhn, J. (1996) Predictive testing for respiratory sensitization Toxicol Lett 86:177-185.
Pauluhn, J. (2003) Issues of dosimetry in inhalation toxicity. Toxicol Lett 140-141:229-238.
Pazdrak, K; Gorski, P; Krakowiak, A; et al. (1993) Changes in nasal lavage fluid due to
formaldehyde inhalation. Int Arch Occup Environ Health 64:515-519.
Peddada, SD; Dinse, GE; Kissling, GE. (2007) Incorporating historical control data when
comparing tumor incidence rates. J Am Stat Assoc 102:1212-1220.
Pinkerton, LE; Hein, MJ; Stayner, LT. (2004) Mortality among a cohort of garment workers
exposed to formaldehyde: an update. Occup Environ Med 61:193-200.
Pitten, FA; Kramer, A; Herrmann, K; et al. (2000) Formaldehyde neurotoxicity in animal
experiments. Pathol Res Pract 196:193-198.
Plesner, BH; and Hansen, K. (1983) Formaldehyde and hexamethylenetetramine in Styles' cell
transformation assay. Carcinogenesis 4:457-459.
Plopper, CG; Weir, AJ; Nishio, SJ; et al. (1994) Elevated susceptibility to 4-ipomeanol
cytotoxicity in immature Clara cells of neonatal rabbits. J Pharmacol Exp Ther 269:867-880.
Poortmans, JR; Kumps, A; Duez, P; et al. (2005) Effect of oral creatine supplementation on
urinary methylamine, formaldehyde, and formate. Med Sci Sports Exerc 37:1717-1720.
Portier, CJ; Hedges, JC; Hoel, DG. (1986) Age-specific models of mortality and tumor onset for
historical control animals in the National Toxicology Program's carcinogenicity experiments.
Cancer Res 46:4372-4378.
Pottern, LM; Heineman, EF; Olsen, JH; et al. (1992) Multiple myeloma among Danish women:
employment history and workplace exposures. Cancer Causes Control 3:427-432.
Pourmahabadian, M; Azam, K; Ghasemkhani, M. (2006) Pulmonary function study between
formaldehyde exposed and non-exposed staffs at some of the Tehran Educational Hospitals. J
Med Sci 6/4:621-625.
Priha, E; Liesivuori, J; Santa, H; et al. (1996) Reactions of hydrated formaldehyde in nasal
mucus. Chemosphere 32:1077-1082.
Proctor, BL; Gaulden, ME; Dowd, MA. (1986) Reactivity and fate of benzene and formaldehyde
in culture medium with and without fetal calf serum; relevance to in vitro mutagenicity testing.
MutatRes 160:259-266.
Pross, HF; Day, JH; Clark, RH; et al. (1987) Immunologic studies of subjects with asthma
exposed to formaldehyde and urea-formaldehyde foam insulation (UFFI) off products. J Allergy
Clin Immunol 79:797-810.
Pushkina, N; Gofmekler, V; Klevtsova, G. (1968) Changes in ascorbic-acid and nucleic-acid
content during benzol and formaldehyde action rat [translation]. Bull Exp Biol Med 66:51-53.
This document is a draft for review purposes only and does not constitute Agency policy.
R-42 DRAFT—DO NOT CITE OR QUOTE
-------
Pyatt, D; Natelson, E; Golden, R. (2008) Is inhalation exposure to formaldehyde a biologically
plausible cause of lymphohematopoietic malignancies? Regul Toxicol Pharmacol 51:119-133.
Que, LG; Liu, L; Yan, Y; et al. (2005) Protection from experimental asthma by an endogenous
bronchodilator. Science 308:1618-1621.
Quievryn, G; Zhitkovich, A. (2000) Loss of DNA-protein crosslinks from formaldehyde-exposed
cells occurs through spontaneous hydrolysis and an active repair process linked to proteosome
function. Carcinogenesis 21:1573-1580.
Ragan, DL; and Boreiko, CJ. (1981) Initiation of C3H/10T1/2 cell transformation by
formaldehyde. Cancer Lett 13:325-331.
Rao, GN; Piegorsch, WW; Haseman, JK. (1987) Influence of body weight on the incidence of
spontaneous neoplasms in rats and mice of long-term studies. Am J Clin Nutr. 45:252-260.
Rapoport, IA. (1946) Carbonyl compounds and the chemical mechanisms of mutation. Compt
Rend Acad Sci (U.S.S.R.) 54:65-67.
Recio, L; Sisk, S; Pluta, L, et al. (1992) p53 mutations in formaldehyde-induced nasal squamous
cell carcinomas in rats. Cancer Res 52:6113-6116.
Report of the Federal Panel on Formaldehyde. (1982) Environ Health Perspect 43:139-168.
Reuzel, PGJ; Wilmer, JWGM; Woutersen, RA; et al. (1990) Interactive effects of ozone and
formaldehyde on the nasal respiratory lining epithelium in rats. J Toxicol Environ Health
29:279-292.
Riedel, F; Hasenauer, E; Barth, PJ; et al. (1996) Formaldehyde exposure enhances inhalative
allergic sensitization in the guinea pig. Allergy 51:94-99.
Ritchie, IM; Lehnen, RG. (1987) Formaldehyde-related health complaints of residents living in
mobile and conventional homes. Am J Public Health 77:323-328.
Robins, J; Ambrun, M; Hute, C; et al. (1988) Estimating the effect of formaldehyde exposure on
long cancer and non-malignant respiratory disease (NMRD) mortality using a new method to
control for the health worker survivor effect. In: Hogstedt, C; Reuterwall, C; eds. Progress in
occupational epidemiology. Amsterdam: Elsevier Science; pp. 75-78.
Roemer, E; Anton, HJ; Kindt, R. (1993) Cell proliferation in the respiratory tract of the rat after
acute inhalation of formaldehyde or acrolein. J Appl Toxicol 13:103-107.
Rogers, J; Mole, M; Chernoff, N; et al. (2002) Developmental toxicity of inhaled methanol in the
CD-I mouse, with application of quantitative dose-response modeling for estimation of
benchmark dosages. U.S. Environmental Protection Agency, Washington, D.C., EPA/600/J-
92/350 (NTIS PB93106979).
Rogers, JM; Brannen, KC; Barbee, BD; et al. (2004) Methanol exposure during gastrulation
causes holoprosencephaly, facial dysgenesis, and cervical vertebral malformations in C57BL/6J
mice Birth Defects B Dev Reprod Toxicol 71:80-88.
Ross, WE; Shipley, N. (1980) Relationship between DNA damage and survival in formaldehyde-
treated mouse cells. Mutat Res 79:277-283.
This document is a draft for review purposes only and does not constitute Agency policy.
R-43 DRAFT—DO NOT CITE OR QUOTE
-------
Ross, WE; McMillan, DR; and Ross, CF. (1981) Comparison of DNA damage by
methylmelamines and formaldehyde. J Natl Cancer Inst 67:217-221.
Rotstein, J; Sarma, DSR; Farber, E. (1986) Sequential alterations in growth control and cell
dynamics of rat hepatocytes in early precancerous steps in hepatocarcinogenesis. Cancer Res
46:2377-2385.
Roush, GC; Walrath, J; Stayner, LT; et al. (1987) Nasopharyngeal cancer, sinonasal cancer, and
occupations related to formaldehyde: a case-control study. J Natl Cancer Inst 79:1221-1224.
Rumchev, KB; Spickett, JT; Bulsara, MK; et al. (2002) Domestic exposure to formaldehyde
significantly increases the risk of asthma in young children. Eur Respir J 20:403-408.
Rumchev, K; Spickett, J; Bulsara, M; et al. (2004) Association of domestic exposure to volatile
organic compounds with asthma in young children. Thorax 59: 746-751.
Rusch, GM; Bolte, HF; Rinehart, WE. (1983a) A 26-week inhalation toxicity study with
formaldehyde in the monkey, rat, and hamster. In: Gibson, JE; ed. Formaldehyde toxicity.
Washington, DC: Hemisphere Publishing Corporation; pp. 98-110.
Rusch, GM; Clary, JJ; Rinehart, WE; et al. (1983b) A 26-week inhalation toxicity study with
formaldehyde in the monkey, rat and hamster. Toxicol Appl Pharmacol 68:329-343.
Ryden, E; Ekstrom, C; Hellmer, L; et al. (2000) Comparison of the sensitivities of Salmonella
typhimurium strains TA102 and TA2638A to 16 mutagens. Mutagenesis 15:495-502.
Sadakane, K; Takano, H; Ichinose, T; et al. (2002) Formaldehyde enhances mite allergen-
induced eosinophilic inflammation in the murine airway. J Environ Pathol Toxicol Oncol
21:267-276.
Saillenfait, AM; Bonnet, P; de Ceaurriz, J. (1989) The effects of maternally inhaled
formaldehyde on embryonal and foetal development in rats. Food Chem Toxicol 27:545-548.
Saito, Y; Nishio, K; Yoshida, Y, et al. (2005) Cytotoxic effect of formaldehyde with free radicals
via increment of cellular reactive oxygen species. Toxicology 210:235-245.
Saladino, AJ; Willey, JC; Lechner, JF, et al. (1985) Effects of formaldehyde, acetaldehyde,
benzoyl peroxide, and hydrogen peroxide on cultured normal human bronchial epithelial cells.
Cancer Res 45:2522-2526.
Salkie, ML. (1991) The prevalence of atopy and hypersensitivity to formaldehyde in
pathologists. Arch Pathol Lab Med 115(6):614—616
Sallmen, M; Lindbohm, ML; Kyyronen, P; et al. (1995) Reduced fertility among women
exposed to organic solvents. Am J Ind Med 27(5):699-713.
Salthammer T, S Mentese and R Marutzky (2010). Formaldehyde in the Indoor Environment.
Chem. Rev., DOI: 10.1021/cr800399g
Sandikci, M; Eren, U; Kum, S. (2007) Effects of formaldehyde and xylene on CD4- and CD8-
positive T cells in bronchus-associated lymphoid tissue in rats. Toxicol Ind Health 23:471-477.
Sanghani, PC; Stone, CL; Ray, BD; et al. (2000) Kinetic mechanism of human glutathione-
dependent formaldehyde dehydrogenase. Biochemistry 39:10720-10729.
This document is a draft for review purposes only and does not constitute Agency policy.
R-44 DRAFT—DO NOT CITE OR QUOTE
-------
Sanotskii, IV; Fomenko, VN; Sheveleva, GA; et al. (1976) Study of the effect of pregnancy of
animals to chemical agents. Gig Tr Prof Zabol 1:25-28.
Santiago, LY; Hann, MC; Ben-Jebria, A; et al. (2001) Ozone absorption in the human nose
during unidirectional airflow. J Appl Physiol 91:725-732.
Sari-Minodier, I; Orsiere, T; Auquier, P; et al. (2001) Use of the micronucleus assay in the
mutagenic risk assessment: A case of 10 workers occupationally exposed to formaldehyde. Arch
Mai Prof 62:75-82.
Sarsilmaz, M; Ozen, O; Akpolat, N; et al. (1999) The histopathologic effects of inhaled
formaldehyde on Ley dig cells of the rats in subacute period [translation], Firat Univ Med Sci J
13:37-40.
Sarsilmaz, M; Kaplan, S; Songur, A; et al. (2007) Effects of postnatal formaldehyde exposure on
pyramidal cell number, volume of cell layer in hippocampus and hemisphere in therat: A
stereological study. Brain Res 1145:157-167.
Sarrif, AM; Krahn, DF; Donovan, SM, et al. (1997) Evaluation of hexamethylphosphoramide for
gene mutations in Salmonella typhimurium using plate incorporation, preincubation, and
suspension assays. Mutat Res 380:167-177.
Sauder, LR; Chatham, MD; Green, DJ; et al. (1986) Acute pulmonary response to formaldehyde
exposure in healthy nonsmokers. J Occup Med 28:420-424.
Sauder, LR; Green, DJ; Chatham, MD; et al. (1987) Acute pulmonary response of asthmatics to
3.0 ppm formaldehyde. Toxicol Ind Health 3:569-578.
Schachter, EN; Witek, TJ, Jr; Tosun, T; et al. (1986) A study of respiratory effects from
exposure to 2 ppm formaldehyde in healthy subjects. Arch Environ Health 41:229-239.
Schachter, EN; Witek, TJ, Jr; Brody, DJ; et al. (1987) A study of respiratory effects from
exposure to 2.0 ppm formaldehyde in occupationally exposed workers. Environ Res 44:188-
205.
Schauenstein, E; Esterbauer, H; Zollner, H. (1977) Aldehydes in biological systems. London:
Pion Limited.
Schlosser, PM. (1999) Relative roles of convection and chemical reaction for the disposition of
formaldehyde and ozone in nasal mucus. Inhal Toxicol 11:967-980.
Schlosser, PM; Lilly, PD; Conolly, RB; et al. (2003) Benchmark dose risk assessment for
formaldehyde using airflow modeling and a single-compartment, DNA-protein cross-link
dosimetry model to estimate human equivalent doses. Risk Anal 23:473-487.
Schmid, E; Goggelmann, W; Bauchinger, M. (1986) Formaldehyde-induced cytotoxic, genotoxic
and mutagenic response in human lymphocytes and Salmonella typhimurium. Mutagenesis
1:427-431.
Schmid, O; Speit, G. (2007) Genotoxic effects induced by formaldehyde in human blood and
implications for the interpretation of biomonitoring studies. Mutagenesis 22:69-74.
Schoenberg, JB; Beck, GJ; Bouhuys, A. (1978) Growth and decay of pulmonary function in
healthy blacks and whites. Respir Physiol 33:367-393.
This document is a draft for review purposes only and does not constitute Agency policy.
R-45 DRAFT—DO NOT CITE OR QUOTE
-------
Schreiber, H; Bibbo, M; Wied, GL; et al. (1979) Bronchial metaplasia as a benign or
premalignant lesion. I. Cytologic and ultrastructural discrimination between acute carcinogen
effects and toxin-induced changes. Acta Cytol 23:496-503.
Schuck, EA; Stephens, ER; Middleton, JT. (1966) Eye irritation response at low concentrations
of irritants. Arch Environ Health 13(5):570-575.
Schulte-Hermann, R; Bursch, W; Grasl-Kraupp, B; et al. (1997) Concepts of cell death and
application to carcinogenesis. Toxicol Pathol 25:89-93.
Schulte-Hermann, R; Bursch, W; Marian, B; et al. (1999) Active cell death (apoptosis) and
cellular proliferation as indicators of exposure to carcinogens. IARC Sci.Publ 273-285.
Seidenberg, J; Anderson, DG; Becker, R. (1986) Validation of an in vivo developmental toxicity
screen in the mouse. Teratog Carcinog Mutagen 6:361-374.
Seidenberg, J; Becker, R. (1987) A summary of the results of 55 chemicals screened for
developmental toxicity in mice. Teratog Carcinog Mutagen 7:17-28.
Seitz, T; Baron, S. (1990) Health hazard evaluation report: Rockcastle Manufacturing, Mount
Vernon, Kentucky. National Institute for Occupational Safety and Health, U.S. Department of
Health and Human Services, Cincinnati, OH; HETA 87-349-2022.
Selevan, SG; Kimmel, CA; Mendola, P. (2000) Identifying critical windows of exposure for
children's health. Environ Health Perspect 108 Suppl 3:451-455.
Sellakumar, AR; Snyder, CA; Solomon, JJ; et al. (1985) Carcinogenicity of formaldehyde and
hydrogen chloride in rats. Toxicol Appl Pharmacol 81:401-406.
Senichenkova, IN. (1991) The embryotoxic effect of industrial environmental pollutants
formaldehyde and gasoline [translation]. Gig Sanit 9:35-38.
Senichenkova, IN; Chebotar, NA. (1996) Effects of gasoline and formaldehyde on prenatal
development of rats with induced iron microelementosis [translation], Ontogenez 27:108-113.
Shaham, J; Bomstein, Y; Meltzer, A; et al. (1996a) DNA-protein crosslinks, a biomarker of
exposure to formaldehyde—in vitro and in vivo studies. Carcinogenesis 17:121-125.
Shaham, J; Bomstein, Y; Meltzer, A; et al. (1996b) Response to Letter to Editor by Casanova et
al. (1996) in the same issue. Carcinogenesis 17:2098-2101.
Shaham, J; Bomstein, Y; Melzer, A; et al. (1997) DNA-protein crosslinks and sister chromatid
exchanges as biomarkers of exposure to formaldehyde. Int J Occup Environ Health 3:95-104.
Shaham, J; Gurvich, R; Kaufman, Z. (2002) Sister chromatid exchange in pathology staff
occupationally exposed to formaldehyde. Mutat Res 514:115-123.
Shaham, J; Bomstein, Y; Gurvich, R; et al. (2003) DNA-protein crosslinks and p53 protein
expression in relation to occupational exposure to formaldehyde. Occup Environ Med 60:403-
409.
Shangina, O; Brennan, P; Szeszenia-Dabrowska, N; et al. (2006) Occupational exposure and
laryngeal and hypopharyngeal cancer risk in central and eastern Europe. Am J Epidemiol
164(4): 3 67-3 75.
This document is a draft for review purposes only and does not constitute Agency policy.
R-46 DRAFT—DO NOT CITE OR QUOTE
-------
Sharpe, JA; Hostovsky, M; Bilbao, JM; et al. (1982) Methanol optic neuropathy. A
histopathological study. Neurology 32:1093-1100.
Shelby, M; Portier, C; Goldman, L; et al. (2004) NTP-CERHR expert panel report on the
reproductive and developmental toxicity of methanol. Reprod Toxicol 18:303-390.
Sheppard, D; Eschenbacher, W; Epstein, J. (1984) Lack of bronchomotor response to up to 3
ppm formaldehyde in subjects with asthma. Environ Res 35:133-139.
Sheveleva (1971) In Toxicology of New Industrial Chemicals. 12:78-86.
Shibamoto, T. (2006) Analytical methods for trace levels of reactive carbonyl compounds
formed in lipid peroxidation systems. J Pharm Biomed Anal 41:12-25.
Shin, HS; Ahn, HS; Lee, BH. (2007) Determination of thiazolidine-4 carboxylates in urine by
chloroformate derivatization and gas chromatography-electron impact mass spectrometry. J
Mass Spectrom 42:1225-1232.
Shipkovitz, HD. (1966) Formaldehyde vapor in the permanent-press fabrics industry.
Occupational Health Program, Public Health Service, U.S. Department of Health, Education, and
Welfare, Cincinnati, OH; Report-TR-52.
Singh, GD; McNamara, JA Jr; Lozanoff, S. (1998) Morphometry of the midfacial complex in
subjects with class III malocclusions: Procrustes, Euclidean, and cephalometric analyses. Clin
Anat 11:162-170.
Skrzydlewska, E. (1996) Decreased antioxidant status and increased lipid peroxidation in rats
after methanol intoxication. Rocz Akad Med Bialymst 41:397-404.
Slater, TF. (1984) Free-radical mechanisms in tissue injury. Biochem J 222:1-15.
Slikker, WJ; Andersen, M; Bogdanffy, M; et al. (2004) Dose-dependent transitions in
mechanisms of toxicity: Case studies. Toxicol Appl Pharmacol 201:226-294.
Smedje, G; Norback, D. (2001) Incidence of asthma diagnosis and self-reported allergy in
relation to the school environment—a four-year follow-up study in schoolchildren. Int J Tuberc
Lung Dis 5:1059-1066.
Smith, AE; Coit, DW. (1995) Penalty Functions. In Environmental Computation 2. Eds: Back,
T; Fogel, DB; Michalewicz., Z. 41-48. Institute of Physics Publishing and Oxford University
Press, New York, NY.
Kusch, I; Schwarz, K; Schwentner, L; et al. (2008) Compounds enhanced in a mass
spectrometric profile of smokers' exhaled breath versus non-smokers as determined in a pilot
study using PTR-MS. J. Breath Res. 2: 1-26.
Snedecor, GW; Cochran, WG. (1980) Statistical Methods, Seventh Edition. Iowa State
University Press, Ames, IA.
Snyder, RD; van Houten, B. (1986) Genotoxicity of formaldehyde and an evaluation of its
effects on the DNA repair process in human diploid fibroblasts. Mutat Res 165:21-30.
Soffritti, M; Maltoni, C; Maffei, F; et al. (1989) Formaldehyde: an experimental multipotential
carcinogen. Toxicol Ind Health 5:699-730.
This document is a draft for review purposes only and does not constitute Agency policy.
R-47 DRAFT—DO NOT CITE OR QUOTE
-------
Soffritti, M; Belpoggi, F; Lambertin, L; et al. (2002) Results of long-term experimental studies
on the carcinogenicity of formaldehyde and acetaldehyde in rats. Ann NY Acad Sci 982:87-
105.
Songur, A; Kus, I; Sahin, S; Sogut, S; Ozen, OA; Yaman, M; Sarsilmaz, M. (2005) The changes
of zinc, copper, and iron levels in lung tissue after formaldehyde inhalation during the early
postnatal period of rats. Eur J Gen Med 2(2):62-68.
Songur, A; Sarsilmaz, M; Ozen, OA. (2008) The effects of inhaled formaldehyde on oxidant and
antioxidant systems of rat cerebellum during the postnatal development process. Toxicol
Mechan Methods 18:569-574.
Sorg, BA; Hochstatter, T. (1999) Behavioral sensitization after repeated formaldehyde exposure
in rats. Toxicol Ind Health 15:346-355.
Sorg, BA; Willis, JR; See, RE; et al. (1996) Repeated low-level formaldehyde exposure produces
long-term cross-sensitization to cocaine: relevance to an animal model for multiple chemical
sensitivity. SocNeurosci Abstr 22:2064
Sorg, BA; Willis, JR; See, RE; et al. (1998) Repeated low-level formaldehyde exposure produces
cross-sensitization to cocaine: Possible relevance to chemical sensitivity in humans.
Neuropsychopharm 18:385-394.
Sorg, BA; Bailie, TM; Tschirgi, ML; et al. (2001) Exposure to repeated low-level formaldehyde
in rats increases basal corticosterone levels and enhances the corticosterone response to
subsequent formaldehyde. Brain Res 898:314-320.
Sorg, BA; Swindell, S; Tschirgi, ML. (2004) Repeated low level formaldehyde exposure
produces enhanced fear conditioning to odor in male, but not female, rats. Brain Res 1008:11-
19.
Spanel, P and Smith, D. (2008) Quantification of trace levels of the potential cancer biomarkets
formaldehyde, acetaldehyde and propanol in breath by SIFT-MS. J. Breath Res. 2: 1-10.
Speit, G; Merk, O. (2002) Evaluation of mutagenic effects of formaldehyde in vitro: detection of
crosslinks and mutations in mouse lymphoma cells. Mutagenesis 17:183-187.
Speit, G; Schmid, O. (2006) Local genotoxic effects of formaldehyde in humans measured by the
micronucleus test with exfoliated epithelial cells. Mutat Res 613:1-9.
Speit, G; Schutz, P; Merk, O. (2000) Induction and repair of formaldehyde-induced DNA-
protein crosslinks in repair-deficient human cell lines. Mutagenesis 15:85-90.
Speit, G; Schmid, O; Frohler-Keller, M; et al. (2007a) Assessment of local genotoxic effects of
formaldehyde in humans measured by the micronucleus test with exfoliated buccal mucosa cells.
Mutat Res 627:129-135.
Speit, G; Schutz, P; Hogel, J, et al. (2007b) Characterization of the genotoxic potential of
formaldehyde in V79 cells. Mutagenesis 22:387-394.
Speit, G; Schmid, O; Neuss, S, et al. (2008) Genotoxic effects of formaldehyde in the human
lung cell line A549 and in primary human nasal epithelial cells. Environ Mol Mutagen 49:300-
307.
This document is a draft for review purposes only and does not constitute Agency policy.
R-48 DRAFT—DO NOT CITE OR QUOTE
-------
Speit, G; Zeller, J; Schmid, O, et al. (2009) Inhalation of formaldehyde does not induce systemic
genotoxic effects in rats. Mutat Res. (In Press)
Squire, RA; Cameron, LL. (1984) An analysis of potential carcinogenic risk from formaldehyde.
Regul Toxicol Pharmacol 4:107-129.
Staab, CA; Hellgren, M; Hoog, J-O. (2008a) Dual functions of alcohol dehydrogenase 3:
Implications with focus on formaldehyde dehydrogenase and S-nitrosoglutathione reductase
activities. Cell Mol Life Sci 65:3950-3960.
Staab, CA; Alander, J; Brandt, M; et al. (2008b) Reduction of S-nitrosoglutathione by alcohol
dehydrogenase 3 is facilated by substrate alcohols via direct cofactor recycling and leads to
GSH-controlled formation of glutathione transferase inhibitors. Biochem J 41:493-504.
Staab, CA; Hellgren, M; Grafstrom, RC; et al. (2009) Medium-chain fatty acids and glutathione
derivatives as inhibitors of S-nitroglutathione reduction mediated by alcohol dehydrogenase 3.
Chem Biol Interact 180:113-118.
Stayner, LT; Elliot, L; Blade, L; et al. (1988) A retrospective cohort mortality study of workers
exposed to formaldehyde in the garment industry. Am J In Med 13:667-681.
Stedman's Medical Dictionary. (2005) 28th edition. Lippincott Williams, and Wilkins, New
York, NY.
Steinhagen, WH; Barrow, CS. (1984) Sensory irritation structure-activity study of inhaled
aldehydes in B6C3F1 and Swiss-Webster mice. Toxicol Appl Pharmacol 72:495-503.
Stellman, SD; Demers, PA; Colin, D; et al. (1998) Cancer mortality and wood dust exposure
among participants in the American cancer Society cancer prevention study-II (CPS-II). Am J
Ind Med 34:229-237.
Sterling, TD; Weinkam, JJ. (1988) Reanalysis of lung cancer mortality in a National Cancer
Institute study on mortality among industrial workers exposed to formaldehyde. J Occup Med
30:895-901.
Sterling, TD; Weinkam, JJ. (1989a) Reanalysis of lung cancer mortality in a National Cancer
Institute study on "Mortality among industrial workers exposed to formaldehyde." Exp Pathol
37:128-132.
Sterling, TD; Weinkam, JJ. (1989b) Reanalysis of lung cancer mortality in a National Cancer
Institute study of "Mortality among industrial workers exposed to formaldehyde:" additional
discussion. J Occup Med 31:881-884.
Sterling, TD; Weinkam, JJ. (1994) Mortality from respiratory cancers (including lung cancer)
among workers employed in formaldehyde industries. Am J Ind Med 25:593-602.
Stewart, PA; Blair, A; Cubit, DA; et al. (1986) Estimating historical exposures to formaldehyde
in a retrospective mortality study. Appl Ind Hyg 1:34-41.
Stewart, PA; Cubit, D; Blair, A. (1987) Formaldehyde levels in seven industries. Appl Ind Hyg
2:231-236.
This document is a draft for review purposes only and does not constitute Agency policy.
R-49 DRAFT—DO NOT CITE OR QUOTE
-------
Stewart, PA; Herrick, RF; Feigley, CE; et al. (1992) Study design for assessing exposures of
embalmers for a case-control study. Part I. Monitoring results. Appl Occup Environ Hyg 7:532-
540.
Stone, DM; Grillo, M; Margolis, FL; et al. (1991) Differential effect of functional olfactory bulb
deafferentation on tyrosine hydroxylase and glutamic acid decarboxylase messenger RNA levels
in rodent juxtaglomerular neurons. J Comp Neurol 311:223-233.
Stroup, NE; Blair, A; Erikson, GE. (1986) Brain cancer and other causes of death in anatomists.
J Natl Cancer Inst 77:1217-1224.
Subramaniam, RP; Richardson, RB; Morgan, KT; et al. (1998) Computational fluid dynamics
simulations of inspiratory airflow in the human nose and nasopharynx. Inhal Toxicol 10:92-120.
Subramaniam, RP; Crump, KS; Van Landingham, C; et al. (2007) Uncertainties in the CUT
model for formaldehyde-induced carcinogenicity in the rat: A limited sensitivity analysis-I. Risk
Anal 27:1237-1254.
Subramaniam, RP, Chen, C; Crump, KS; et al. (2008) Uncertainties in biologically-based
modeling of formaldehyde-induced respiratory cancer risk: Identification of key issues. Risk
Anal 28:907-923.
Sul, D; Kim, H; Oh, E; et al. (2007) Gene expression profiling in lung tissues from rats exposed
to formaldehyde. Arch Toxicol 81:589-597.
Suruda, A; Schulte, P; Boeniger, M; et al. (1993) Cytogenetic effects of formaldehyde exposure
in students of mortuary science. Cancer Epidemiol Biomarkers Prev 2:453-460.
Suskov, II; and Sazonova, LA. (1982) Cytogenetic effects of epoxy, phenolformaldehyde and
polyvinylchloride resins in man. Mutat Res 104:137-140.
Sweeney, MH; Beaumont, JJ; Waxweiler, RJ; et al. (1986) An investigation of mortality from
cancer and other causes of death among workers employed at an East Texas chemical plant.
Arch Environ Health 41:23-28.
Swenberg, JA; Kerns, WD; Mitchell, RI; et al. (1980) Induction of squamous cell carcinomas of
the rat nasal cavity by inhalation exposure to formaldehyde vapor. Cancer Res 40:3398-3402.
Swenberg, JA; Gross, EA, Martin, J; et al. (1983) Mechanisms of formaldehyde toxicity. In:
Formaldehyde toxicity, Ed: Gibson, JE. Hemisphere Publishing Corp. Washington, D.C.
Swenberg, JA; Gross, EA; Randall, HW. (1986) Localization and quantitation of cell
proliferation following exposure to nasal irritants. In Toxicology of Nasal Passages, ed: Barrow,
CS. Hemisphere Publishing Corp. Washington, D.C.
Swiecichowski, AL; Long, KJ; Miller, ML; et al. (1993) Formaldehyde-induced airway
hyperreactivity in vivo and ex vivo in guinea pigs. Environ Res 61:185-199.
Szarvas, T; Szatloczky, E; Volford, J; et al. (1986) Determination of endogenous formaldehyde
level in human blood and urine by dimedone-14C radiometric method. J Radioanaly Nucl Chem
Letters 106:357-367.
Takahashi, K; Morita, T; Kawazoe, Y. (1985) Mutagenic characteristics of formaldehyde on
bacterial systems. Mutat Res 156:153-161.
This document is a draft for review purposes only and does not constitute Agency policy.
R-50 DRAFT—DO NOT CITE OR QUOTE
-------
Takahashi, M; Hasegawa, R; Furukawa, F; et al. (1986) Effects of ethanol, potassium
metabisulfite, formaldehyde and hydrogen peroxide on gastric carcinogenesis in rats after
initiation with N-methyl-N'-nitro-N-nitrosoguanidine. Jpn J Cancer Res 77:118-124.
Takahashi, S; Tsuji, K; Fujii, K; et al. (2007) Prospective study of clinical symptoms and skin
test reactions in medical students exposed to formaldehyde gas. J Dermatol 34(5):283-289.
Takigawa, T; Usami, M; Yamasaki, Y; et al. (2005) Reduction of indoor formaldehyde
concentrations and subjective symptoms in a gross anatomy laboratory. Bull Environ Contam
Toxicol 74:1027-1033.
Tarkowski, M; Gorski, P. (1995) Increased IgE antiovalbumin level in mice exposed to
formaldehyde. Int Arch Allergy Immunol 106:422-424.
Tarone, RE. (1975) Tests for trend in life table analysis. Biometrika 62:679-682. (as cited in
Battelle Columbus Laboratories [1981])
Tarone, R; McLaughlin, JK. (2005) Re: "mortality from solid cancers among workers in
formaldehyde industries." Am JEpidemiol 161:1089-1091.
Taskinen, H; Kyyronen, P; Hemminki, K. (1994) Laboratory work and pregnancy outcome. J
Occup Med 36:311-319.
Taskinen, HK; Kyyronen, P; Sallmen, M; et al. (1999) Reduced fertility among female wood
workers exposed to formaldehyde. Am J Ind Med 36:206-212.
Tavernier, G; Fletcher, G; Gee, I; et al. (2006) IPEADAM study: Indoor endotoxin exposure,
family status, and some housing characteristics in English children. J Allergy Clin Immunol
117:656-662.
Temcharoen, P; Thilly, WG. (1983) Toxic and mutagenic effects of formaldehyde in Salmonella
typhimurium. Mutat Res 119:89-93.
Teng, S; Beard, K; Pourahmad, J; et al. (2001) The formaldehyde metabolic detoxification
enzyme systems and molecular cytotoxic mechanism in isolated rat hepatocytes. Chem Biol
Interact 130-132:285-296.
Thomas, RS; Allen, BC; Nong, A; et al. (2007) A method to integrate benchmark dose estimates
with genomic data to assess the functional effects of chemical exposure. Toxicol Sci 98(1):240-
248.
Thompson, CM; Grafstrom, RC. (2008) Mechanistic considerations for formaldehyde-induced
bronchoconstriction involving S-nitrosoglutathione reductase. J Toxicol Environ Health A
71:244-248.
Thompson, CM; Subramaniam, RP; Grafstrom, RC. (2008) Mechanistic and dose considerations
for supporting adverse pulmonary physiology in response to formaldehyde. Toxicol Appl
Pharmacol 233:355-359.
Thompson, CM; Sonawane, B; Grafstrom, RC. (2009) The ontogeny, distribution and regulation
of alcohol dehydrogenase 3: implications for pulmonary physiology. Drug Metab Dispos. [Epub
ahead of print]
This document is a draft for review purposes only and does not constitute Agency policy.
R-51 DRAFT—DO NOT CITE OR QUOTE
-------
Thomson, EJ; Shackleton, S; Harrington, JM. (1984) Chromosome aberrations and sister-
chromatid exchange frequencies in pathology staff occupationally exposed to formaldehyde.
MutatRes 141:89-93.
Thorndike, J; Beck, WS. (1977) Production of formaldehyde from N5-methyltetrahydrofolate by
normal and leukemic leukocytes. Cancer Res 37:1125-1132.
Thrasher, JD; Wojdani, A; Cheung, G; et al. (1987) Evidence for formaldehyde antibodies and
altered cellular immunity in subjects exposed to formaldehyde in mobile homes. Arch Environ
Health 42:347-350.
Thrasher, JD; Broughton, A; Gard, Z. (1988) Indoor formaldehyde and the elderly. Clin
Gerontol 7(3-4):63-66.
Thrasher, JD; Broughton, A; Madison, R. (1990) Immune activation and autoantibodies in
humans with long-term inhalation exposure to formaldehyde. Arch Environ Health 45:217-223.
Thun, MJ; Lakat, M; Altman, R. (1982) Symptom survey of residents of homes insulated with
urea formaldehyde foam. Environ Res 29(2):320-334.
Til, HP; Woutersen, RA; Feron, VJ; et al. (1988) Evaluation of the oral toxicity of acetaldehyde
and formaldehyde in a 4-week drinking-water study in rats. Food Chem Toxicol 26:447-452.
Til, HP; Woutersen, RA; Feron, VJ; et al. (1989) Two-year drinking-water study of
formaldehyde in rats. Food Chem Toxicol 27:77-87.
Titenko-Holland, N; Levine, AJ; Smith, MT; et al. (1996) Quantification of epithelial cell
micronuclei by fluorescence in situ hybridization (FISH) in mortuary science students exposed to
formaldehyde. MutatRes 371:237-248.
Tobe, M; Kaneko, Y; Uchida, Y; et al. (1985) Studies of the inhalation toxicity of formaldehyde.
National Sanitary and Medical Laboratory Service, Toxicity, Tokyo, Japan; T-85-0236. (cited in
IPCS, 2002)
Tobe, M; Naito, K; Kurokawa, Y. (1989) Chronic toxicity study on formaldehyde administered
orally to rats. Toxicology 56:79-86.
Togias, A. (1999) Mechanisms of nose-lung interaction. Allergy 54:94-105.
Trezl, L; Hullan, L; Jaszay, ZM; et al. (2003) Antagonistic reactions of arginine and lysine
against formaldehyde and their relation to cell proliferation, apoptosis, folate cycle and
photosynthesis. Mol Cell Biochem 244:167-176.
Tsubone, H; Kawata, M. (1991) Stimulation to the trigeminal afferent nerve of the nose by
formaldehyde, acrolein, and acetaldehyde gases. Inhal Toxicol 3:211-222.
Tsukahara, S; Yamamoto, S; Tin-Tin-Tin-Shwe; et al. (2006) Inhalation of low-level
formaldehyde increases the Bcl-2/Bax expression ratio in the hippocampus of immunologically
sensitized mice. Neuroimmunomodulation 13:63-68.
Turner, C; Parekh, B; Walton, C; et al. (2008) An exploratory comparative study of volatile
compounds in exhaled breath and emitted by skin using selected ion flow tube mass
spectrometry. Rapid Commun Mass Spectrom. 22(4):526-532.
This document is a draft for review purposes only and does not constitute Agency policy.
R-52 DRAFT—DO NOT CITE OR QUOTE
-------
Tuthill, RW. (1984) Woodstoves, formaldehyde, and respiratory disease. Am J Epidemiol
120:952-955.
Tyihak, E; Albert, L; Nemeth, ZI; et al. (1998) Formaldehyde cycle and the natural
formaldehyde generators and capturers. Acta Biol Hung 49:225-238.
Tyihak, E; Bocsi, J; Timar, F; et al. (2001) Formaldehyde promotes and inhibits the proliferation
of cultured tumour and endothelial cells. Cell Prolif 34:135-141.
U.S. EPA (U.S. Environmental Protection Agency). (1985) Health and environmental effects
profile for formaldehyde. Prepared by the U.S. Environmental Protection Agency, Office of
Research and Development, Cincinnati, OH for the Office of Solid Waste and Emergency
Response, Washington, D.C.; EPA 600/X-85/362.
U.S. EPA (Environmental Protection Agency). (1986a) Guidelines for the health risk assessment
of chemical mixtures. Federal Register 51(185):34014-34025. Available from:
.
U.S. EPA. (1986b) Guidelines for mutagenicity risk assessment. Federal Register
51(185):34006-34012. Available from: .
U.S. EPA. (1987) Assessment of health risks to garment workers and certain home residents
from exposure to formaldehyde. Office of Pesticides and Toxic Substances, Washington, DC.
U.S. EPA. (1988) Recommendations for and documentation of biological values for use in risk
assessment. Prepared by the Environmental Criteria and Assessment Office, Office of Health and
Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and Emergency
Response, Washington, DC; EPA 600/6-87/008. Available from:
.
U.S. EPA. (1989) Review of the national ambient air quality standard for ozone, assessment for
scientific and technical information. U.S. Environmental Protection Agency, RTP, NC, EPA-
450/2-92.
U.S. EPA. (1991a) Guidelines for developmental toxicity risk assessment. Federal Register
56(234):63798-63826. Available from: .
U.S. EPA. (1991b) Formaldehyde Risk Assessment Update. Prepared by the Office of Toxic
Substances, Washington, DC.
U.S. EPA. (1991c) Formaldehyde. Integrated Risk Information System (IRIS). National Center
for Environmental Assessment, Washington, DC. Available online at http://www.epa.gov/iris.
U.S. EPA. (1994a) Interim policy for particle size and limit concentration issues in inhalation
toxicity studies. Federal Register 59(206):53799. Available from:
.
U.S. EPA. (1994b) Methods for derivation of inhalation reference concentrations and application
of inhalation dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-
90/066F. Available from: .
This document is a draft for review purposes only and does not constitute Agency policy.
R-53 DRAFT—DO NOT CITE OR QUOTE
-------
U.S. EPA. (1995) Use of the benchmark dose approach in health risk assessment. Risk
Assessment Forum, Washington, DC; EPA/630/R-94/007. Available from:
.
U.S. EPA. (1996) Guidelines for reproductive toxicity risk assessment. Federal Register
61(212):56274-56322. Available from: .
U.S. EPA. (1998) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-
26954. Available from: .
U.S. EPA. (2000a) Science policy council handbook: risk characterization. Office of Science
Policy, Office of Research and Development, Washington, DC; EPA 100-B-00-002. Available
from: .
U.S. EPA. (2000b) Benchmark dose technical guidance document [external review draft]. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/001. Available from:
.
U.S. EPA. (2000c) Supplementary guidance for conducting for health risk assessment of
chemical mixtures. Risk Assessment Forum, Washington, DC; EPA/630/R-00/002. Available
from: .
U.S. EPA. (2002a) A review of the reference dose and reference concentration processes. Risk
Assessment Forum, Washington, DC; EPA/630/P-02/0002F. Available from:
.
U.S. EPA. (2002b) Toxicological Review and IRIS Summary of 1,3-Butadiene. Integrated Risk
Information System, National Center for Environmental Assessment, Washington, DC.
Accessed at http://www.epa.gov/EPA/iris/subst/0139.htm.
U.S. EPA. (2002c) Memorandum: Interim genomics policy. Office of the Administrator Science
Policy Council.
U.S. EPA. (2004) Potential implications of genomics for regulatory and risk assessment
applications at EPA. Prepared by the Genomics Task Force Workgroup. EPA 100/B-04/002.
U.S. EPA. (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/P-03/001B. Available from: .
U.S. EPA. (2005b) Supplemental guidance for assessing susceptibility from early-life exposure
to carcinogens. Risk Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available from:
.
U.S. EPA. (2006a) Science policy council handbook: peer review. Third edition. Office of
Science Policy, Office of Research and Development, Washington, DC; EPA/100/B-06/002.
Available from: .
U.S. EPA. (2006b) A framework for assessing health risk of environmental exposures to
children. National Center for Environmental Assessment - Washington, DC, EPA/600/R-
05/093F. Available from: .
This document is a draft for review purposes only and does not constitute Agency policy.
R-54 DRAFT—DO NOT CITE OR QUOTE
-------
U.S. EPA. (2006c) Technology transfer network 1999 national-scale air toxics assessment.
Available from http://www.gov/ttn/atw/nata/.
U.S. EPA. (2006d) Air quality criteria for ozone and related photochemical oxidants (Final).
U.S. Environmental Protection Agency, Washington, DC, EPA/600/R-05/004aF-cF.
U.S. EPA (2008) Reregi strati on eligibility decision for formaldehyde and paraformaldehyde.
Office of Pesticide Programs. Available from:
http://www.epa.gov/pesticides/reregistration/formaldehyde/.
U.S. EPA (2009a) Formaldehyde: Chemical Report. TRI Explorer. Available from
http://www.epa.gov/triexplorer/chemical.htm.
U.S. EPA (2009b) AirData. Available from http://www.epa.gov/air/data/index.html.
Uba, G; Pachorek, D; Bernstein, J; et al. (1989) Prospective study of respiratory effects of
formaldehyde among healthy and asthmatic medical students. Am J Ind Med 15(1):91—101.
Uotila, L; Koivusalo, M. (1974a) Formaldehyde dehydrogenase from human liver. Purification,
properties, and evidence for the formation of glutathione thiol esters by the enzyme. J Biol
Chem 249:7653-7663.
Uotila, L, Koivusalo, M. (1974b) Purification and properties of S-formylglutathione hydrolase
from human liver. J Biol Chem 249:7664-7672.
Ura, H; Nowak, P; Litwin, S; et al. (1989) Effects of formaldehyde on normal xenotransplanted
human tracheobronchial epithelium. Am J Pathol 134:99-106.
Usanmaz, SE; Akarsu, ES; Vural, N. (2002) Neurotoxic effects of acute and subacute
formaldehyde exposures in mice. Environ Toxicol Pharmacol 11:93-100.
USDA (U.S. Department of Agriculture). (1979) Food intakes: Individuals in 48 States, Year
1977-78, Nationwide Food Consumption Survey 1977-78, Rep No 1-1. Technical report no.
PB91-103523INZ.
van Birgelen, APJM; Chou, BJ; Renne, RA; et al. (2000) Effects of glutaraldehyde in a 2-year
inhalation study in rats and mice. Toxicol Sci 55:195-205.
Van Schoor, J; Joos, GF; Pauwels, RA. (2000) Indirect bronchial hyperresponsiveness in asthma:
mechanisms, pharmacology and implications for clinical research. Eur Respir J 16:514-533.
Vandenplas, O; Fievez, P; Delwiche, JP; et al. (2004) Persistent asthma following accidental
exposure to formaldehyde. Allergy 59:115-116.
Vargova, M; Janota, S; Karelova, J; et al. (1992) Analysis of the health risk of occupational
exposure to formaldehyde using biological markers. Analysis 20:451-454.
Vasudeva, N; Anand, C. (1996) Cytogenetic evaluation of medical students exposed to
formaldehyde vapor in the gross anatomy dissection laboratory. J Am Coll Health 44:177-179.
Vaughan, TL; Strader, C; Davis, S; et al. (1986a) Formaldehyde and cancers of the pharynx,
sinus and nasal cavity: I. Occupational exposures. Int J Cancer 38:677-683.
Vaughan, TL; Strader, C; Davis, S; et al. (1986b) Formaldehyde and cancers of the pharynx,
sinus and nasal cavity: II. Residential exposures. Int J Cancer 38:685-688.
This document is a draft for review purposes only and does not constitute Agency policy.
R-55 DRAFT—DO NOT CITE OR QUOTE
-------
Vaughan, TL; Stewart, PA; Teschke, K; et al. (2000) Occupational exposure to formaldehyde
and wood dust and nasopharyngeal carcinoma. Occup Environ Med 57:376-384.
Venkatakrishna-Bhatt, H; Panchal, GM. (1992) Behavioural change in rats due to chronic oral
and systemic formaldehyde. Ind J Physiol Pharmacol 36:270-272.
Venkatakrishna-Bhatt, H and Panchal, GM (1997). Chronic exposure to formaldehyde in
drinking water and behavioral changes in rats. Ind. J. Physioil. & Allied Sci. 51(4): 181-184.
Vogelstein, B; Fearon, ER; Hamilton, SR; et al. (1988) Genetic alterations during colorectal-
tumor development. N Engl J Med 319:525-532.
Von Hippel, PH; and Wong, KY. (1971) Dynamic aspects of native DNA structure: kinetics of
the formaldehyde reaction with calf thymus DNA. J Mol Biol 61:587-613.
Wahlberg, JE. (1993) Measurement of skin-fold thickness in the guinea pig. Assessment of
edema-inducing capacity of cutting fluids, acids, alkalis, formalin and dimethyl sulfoxide.
Contact Dermatitis 28:141-145.
Walker, JF. (1975) Formaldehyde. 3rd edition. Huntington, NY: R.E. Krieger Publishing
Company.
Walrath, J; Fraumeni, JF, Jr. (1983) Mortality patterns among embalmers. Int J Cancer 31:407-
411.
Walrath, J; Fraumeni, JF, Jr. (1984) Cancer and other causes of death among embalmers. Cancer
Res 44:4638-4641.
Wang, B; and Liu, DD. (2006) [Detection of formaldehyde induced developmental toxicity
assessed with single cell gel electrophoresis]. Fen Zi Xi Bao Sheng Wu Xue Bao 39:462-466.
Wang, M; Cheng, G; Villalta, PW; et al. (2007) Development of liquid chromatography
electrospray ionization tandem mass spectrometry methods for analysis of DNA adducts of
formaldehyde and their application to rats treated with N-nitrosodimethylamine or 4-
(methylnitrosamino)-l-(3-pyridyl)-l-butanone. Chem Res Toxicol 20:1141-1148.
Wang TS, Pysanenko A, Dryahina K, Spanel P, Smith D. (2008) Analysis of breath, exhaled via
the mouth and nose, and the air in the oral cavity. J. Breath Res. 2:1-13.
Wangenheim, J; and Bolcsfoldi, G. (1988) Mouse lymphoma L5178Y thymidine kinase locus
assay of 50 compounds. Mutagenesis 3:193-205.
Wantke, F; Demmer, CM; Tappler, P; et al. (1996a) Exposure to gaseous formaldehyde induces
IgE-mediated sensitization to formaldehyde in school-children. Clin Exp Allergy 26:276-280.
Wantke, F; Focke, M; Hemmer, W; et al. (1996b) Formaldehyde and phenol exposure during an
anatomy dissection course: a possible source of IgE-mediated sensitization. Allergy 51:837-
841.
Ward, Jr., JB; Hokanson, JA; Smith, ER; et al. (1984) Sperm count, morphology and fluorescent
body frequency in autopsy service workers exposed to formaldehyde. Mutat Res 30:417-424.
This document is a draft for review purposes only and does not constitute Agency policy.
R-56 DRAFT—DO NOT CITE OR QUOTE
-------
Watanabe, K; Sakamoto, K; Sasaki, T. (1996) Comparisons on chemically-induced mutagenicity
among four bacterial strains, Salmonella typhimurium TA102 and TA2638, and Escherichia coli
WP2/pKM101 and WP2 uvrA/pKMlOl: collaborative study I. Mutat Res 361:143-155.
Waydhas, C; Weigl, K; Sies, H. (1978) The disposition of formaldehyde and formate arising
from drug N-demethylations dependent on cytochrome P-450 in hepatocytes and in perfused rat
liver. Eur JBiochem 89:143-150.
Weber-Tschopp, A; Fischer, T; Grandjean, E. (1977) Irritating effects of formaldehyde on man.
Int Arch Occup Environ Health 39(4):207-218.
Weibel, ER. (1963) Morphometry of the human lung. Springer-Verlag, Berlin.
Weiler, E; Apfelbach, R. (1992) Age-related differences in olfactory sensitivity in Wistar rats
after low-level formaldehyde gas exposure. Chem Senses 17:717-718.
Weiss, B; Stern, S; Soderholm, SC; et al. (1996) Developmental neurotoxicity of methanol
exposure by inhalation in rats. Research Report (Health Effects Institute) 73:1-64.
Weisskopf, M; Morozova, N; O'Reilly, EJ; et al. (2009) Prospective study of chemical exposures
and amyotrophic lateral sclerosis mortality. J Neurol Neurosurg Psychiatry 80:558-561.
West, S; Hildesheim, A; Dosemeci, M. (1993) Non-viral risk factors for nasopharyngeal
carcinoma in the Philippines: results from a case-control study. Int J Cancer 55:722-727.
Westerlund, M; Gaiter, D; Carmine, A; et al. (2005) Tissue- and species-specific expression
patterns of class I, III, and IV ADH and ALDH1 mRNAs in rodent embryos. Cell Tissue Res
322:227-236.
WHO (World Health Organization). (1987) Laboratory manual for the examination of human
semen and sperm-cervical mucus interaction. Cambridge, UK: Cambridge University Press.
Wilcox, AJ; Horney, LF. (1984) Accuracy of spontaneous abortion recall. Am J Epidemiol
120:727-733.
Wilcox, P; Naidoo, A; Wedd, DJ; et al. (1990) Comparison of Salmonella typhimurium TA102
with Escherichia coli WP2 tester strains. Mutagenesis 5:285-291.
Wilhelmsson, B; Holmstrom, M. (1992) Possible mechanisms of formaldehyde-induced
discomfort in the upper airways. Scand J Work Environ Health 18:403-407.
Wilkins, RJ; and Macleod, HD. (1976) Formaldehyde induced DNA-protein crosslinks in
Escherichia Coli. Mutat Res 36:11-6.
Williams, GM; Mori, H; and McQueen, CA. (1989) Structure-activity relationships in the rat
hepatocyte DNA-repair test for 300 chemicals. Mutat Res 221:263-286.
Wilmer, JW; Woutersen, RA; Appelman, LM; et al. (1987) Subacute (4-week) inhalation
toxicity study of formaldehyde in male rats: 8-hour intermittent versus 8-hour continuous
exposures. J Appl Toxicol 7:15-16.
Wilmer, JW; Woutersen, RA; Appelman, LM; et al. (1989) Subchronic (13-week) inhalation
toxicity study of formaldehyde in male rats: 8-hour intermittent versus 8-hour continuous
exposures. Toxicol Lett 47:287-293.
This document is a draft for review purposes only and does not constitute Agency policy.
R-57 DRAFT—DO NOT CITE OR QUOTE
-------
Witek, TJ, Jr; Schachter, EN; Tosun, T. (1986) Controlled human studies on the pulmonary
effects of indoor air pollution: experiences with sulfur dioxide and formaldehyde. Environ Int
12:129-135.
Witek, TJ, Jr; Schachter, EN; Tosun, T; et al. (1987) An evaluation of respiratory effects
following exposure to 2.0 ppm formaldehyde in asthmatics: lung function, symptoms, and
airway reactivity. Arch Environ Health 42:230-237.
Wolf, DC; Gross, EA; Lyght, O, et al. (1995) Immunohistochemical localization of p53, PCNA,
and TGF-alpha proteins in formaldehyde-induced rat nasal squamous cell carcinomas. Toxicol
Appl Pharmacol 132:27-35.
Wolfe, RK. (1986) Effects of airborne pollutants on mucociliary clearance. Environ Health
Perspect 66: 223-237.
Wong, O. (1983) An epidemiologic mortality study of a cohort of chemical workers potentially
exposed to formaldehyde, with a discussion on SMR and PMR. In: Gibson, J; ed. Formaldehyde
toxicity. Washington, DC: Hemisphere Publishing Corporation; pp. 256-272.
Wong, EY; Ray, R; Gao, DL; et al. (2006) Reproductive history, occupational exposures, and
thyroid cancer risk among women textile workers in Shanghai, China. Int Arch Occup Environ
Health 79:251-158.
Wood, RW; Coleman, JB. (1995) Behavioral evaluation of the irritant properties of
formaldehyde. Toxicol Appl Pharmacol 130:67-72.
Wortley, P; Vaughan, TL; Davis, S; et al. (1992) A case-control study of occupational risk
factors for laryngeal cancer. Br J Ind Med 49:837-844.
Woutersen, RA; Appelman, LM; Wilmer, JW; et al. (1987) Subchronic (13-week) inhalation
toxicity study of formaldehyde in rats. J Appl Toxicol 7:43-49.
Woutersen, RA; Garderen-Hoetmer, A; Bruijntjes, JP; et al. (1989) Nasal tumours in rats after
severe injury to the nasal mucosa and prolonged exposure to 10 ppm formaldehyde. J Appl
Toxicol 9:39-46.
Wu, H-Y; Nguyen, HH; Russell, MW. (1997) Nasal lymphoid tissue (NALT) as a mucosal
immune inductive site. Scand J Immunol 46:506-513.
Wu, H; Romieu, I; Sienra-Monge, J-J; et al. (2007) Genetic variation in S-nitrosoglutathione
reductase (GSNOR) and childhood asthma. J Allergy Clin Immunol 120:322-328.
Xing S-Y; Ye. L; Wang, N-N. (2007) Toxic effect of formaldehyde on reproduction and heredity
in male mice. J Jilin Univ 33:716-718.
Xu, B; Aoyama, K; Takeuchi, M; et al. (2002) Expression of cytokine mRNAs in mice
cutaneously exposed to formaldehyde. Immunol Lett 84:49-55.
Yager, JW; Cohn, KL; Spear, RC; et al. (1986) Sister-chromatid exchanges in lymphocytes of
anatomy students exposed to formaldehyde-embalming solution. Mutat Res 174:135-139.
Yang, X; Zhang, YP; Chen, D; et al. (2001) Eye irritation caused by formaldehyde as an indoor
air pollution—a controlled human exposure experiment. Biomed Environ Sci 14:229-236.
This document is a draft for review purposes only and does not constitute Agency policy.
R-58 DRAFT—DO NOT CITE OR QUOTE
-------
Yang, YH; Xi, ZG; Chao, FH; et al. (2005) Effects of formaldehyde inhalation on lung of rats.
Biomed Environ Sci 18:164-168.
Ye, X; Yan, W; Xie, H; et al. (2005) Cytogenetic analysis of nasal mucosa cells and
lymphocytes from high-level long-term formaldehyde exposed workers and low-level short-term
exposed waiters. Mutat Res 588:22-27.
Yi C, Ke K, Xiaohua L and Xu Y (2007). Up-regulation of GSNO reductase in mice lungs by
formaldehyde inhalation. Bioinform Biomed Engineering 6-8: 294-297.
Ying, CJ; Yan, WS; Zhao, MY; et al. (1997) Micronuclei in nasal mucosa, oral mucosa and
lymphocytes in students exposed to formaldehyde vapor in anatomy class. Biomed Environ Sci
10:451-455.
Ying, CJ; Ye, XL; Xie, H; et al. (1999) Lymphocyte subsets and sister-chromatid exchanges in
the students exposed to formaldehyde vapor. Biomed Environ Sci 12:88-94.
Yokley, KA; Tran, H; Schlosser, PM. (2008) Sensory irritation response in rats: modeling,
analysis and validation. Bull Math Biol 70:555-588.
Young, JT. (1981) Histopathologic examination of the rat nasal cavity. Fundam Appl Toxicol
1:309-312.
Yu, L-Q; Jiang, S-F; Leng, S-G; et al. (2005) Early genetic effects on workers occupationally
exposed to formaldehyde. Chin J Prevent Med 39:392-395.
Yu, PH; Zuo, DM. (1996) Formaldehyde produced endogenously via deamination of
methylamine. A potential risk factor for initiation of endothelial injury. Atherosclerosis 120:189-
197.
Yu, PH; Lai, CT; Zuo, DM. (1997) Formation of formaldehyde from adrenaline in vivo; a
potential risk factor for stress-related angiopathy. Neurochem Res 22:615-620.
Yu, PH. (1998) Increase of formation of methylamine and formaldehyde in vivo after
administration of nicotine and the potential cytotoxicity. Neurochem Res 23:1205-1210.
Zhang, L; Steinmaus, C; Eastmond, DA; et al. (2009) Formaldehyde exposure and leukemia: A
new meta-analysis and potential mechanisms. Mutat Res 681:150-168.
Zhao, Z; Zhang, Z; Wang, Z; et al. (2008) Asthmatic symptoms among pupils in relation to
winter indoor and outdoor air pollution in schools in Taiyaun, China. Environ Health Perspect
116:90-97.
Zhitkovich, A; and Costa, M. (1992) A simple, sensitive assay to detect DNA-protein crosslinks
in intact cells and in vivo. Carcinogenesis 13:1485-1489.
Zhong, W; Que Hee, S. (2004a) Quantitation of normal and formaldehyde-modified
deoxynucleosides by high-performance liquid chromatography/UV detection. Biomed
Chromatogr 18:462-469.
Zhong, W; Que Hee, SS. (2004b) Formaldehyde-induced DNA adducts as biomarkers of in vitro
human nasal epithelial cell exposure to formaldehyde. Mutat Res 563:13-24.
This document is a draft for review purposes only and does not constitute Agency policy.
R-59 DRAFT—DO NOT CITE OR QUOTE
-------
Zhou, DX; Qiu, SD; Zhang, J; et al. (2006) Reproductive toxicity of formaldehyde to adult male
rats and the functional mechanism concerned. Sichuan Da Xue Xue Bao Yi Xue Ban 37:566-
569.
Zielenska, M; Guttenplan, JB. (1988) Mutagenic activity and specificity of N-
nitrosomethylaniline and N-nitrosodiphenylamine in Salmonella. Mutat Res 202:269-76.
Zuercher, AW; Cebra, JJ. (2002) Structural and functional differences between putative mucosal
inductive sites of the rat. Eur J Immunol 32:3191-3196.
Zuercher, AW; Coffin, SE; Thurnheer, MC; et al. (2002) Nasal-associated lymphoid tissue is a
mucosal inductive site for virus-specific humoral and cellular immune responses. J Immunol
168:1796-1803.
Zwart, A; Woutersen, RA; Wilmer, JW; et al. (1988) Cytotoxic and adaptive effects in rat nasal
epithelium after 3-day and 13-week exposure to low concentrations of formaldehyde vapour.
Toxicology 51:87-99.
- End of Volume III -
This document is a draft for review purposes only and does not constitute Agency policy.
R-60 DRAFT—DO NOT CITE OR QUOTE
------- |