United States
Environmental Protection
Agency
Health Effects Support
Document for Boron
-------
Health Effects Support Document
for
Boron
U.S. Environmental Protection Agency
Office of Water (43 04T)
Health and Ecological Criteria Division
Washington, DC 20460
www.epa.gov/safewater/ccl/pdf/boron.pdf
EPA Document Number EPA-822-R-06-005
September, 2006
Printed on Recycled Paper
-------
FOREWORD
The Safe Drinking Water Act (SDWA), as amended in 1996, requires the Administrator
of the U.S. Environmental Protection Agency (U.S. EPA) to establish a list of contaminants to
aid the Agency in regulatory priority setting for the drinking water program. In addition, the
SDWA requires the U.S. EPA to make regulatory determinations for no fewer than five
contaminants by August 2001 and every five years thereafter. The criteria used to determine
whether or not to regulate a chemical on the Contaminant Candidate List (CCL) are the
following:
The contaminant may have an adverse effect on the health of persons.
The contaminant is known to occur or there is a substantial likelihood that the
contaminant will occur in public water systems with a frequency and at levels of
public health concern.
• In the sole judgment of the Administrator, regulation of such contaminant presents a
meaningful opportunity for health risk reduction for persons served by public water
systems.
The Agency's findings for all three criteria are used in making a determination to
regulate a contaminant. The Agency may determine that there is no need for regulation when a
contaminant fails to meet one of the criteria. The decision not to regulate is considered a final
Agency action and is subject to judicial review.
This document provides the health effects basis for the regulatory determination for
boron. In arriving at the regulatory determination, data on toxicokinetics, human exposure, acute
and chronic toxicity to animals and humans, epidemiology, and mechanisms of toxicity were
evaluated. In order to avoid wasteful duplication of effort the Toxicokinetic, Hazard
Identification and Dose-Response Assessment Chapters in the Document are a reproduction of
the comparable Chapters the U. S. EPA Integrated Risk Information System (IRIS)
Toxicological Review for Boron and Compounds (www.epa.gov/iris/toxreviews/0410-tr.pdf)
(U.S. EPA, 2004a). The IRIS assessment was completed in June 2004. The Chapters on
chemical and physical properties, environmental release and fate, exposure from water and
substances other than water was prepared by the Office of Water for the Regulatory
Determination.
A Reference Dose (RfD) from the IRIS Toxicological Review is provided as the
assessment of long-term toxic effects other than carcinogenicity. RfD determination assumes
that thresholds exist for certain toxic effects, such as cellular necrosis, significant body or organ
weight changes, blood disorders, etc. It is expressed in terms of milligrams per kilogram per day
(mg/kg-day). In general, the RfD is an estimate (with uncertainty spanning perhaps an order of
magnitude) of a daily oral exposure to the human population (including sensitive subgroups) that
is likely to be without an appreciable risk of deleterious effects during a lifetime.
Proposal Draft
Boron — September, 2006
-------
The carcinogenicity assessment for boron from the IRIS Toxicological Review includes a
formal hazard identification and a weight-of-evidence judgment of the likelihood that the agent
is a human carcinogen via the oral exposure route.
Development of these hazard identifications and dose-response assessments for boron has
followed the general guidelines for risk assessment as set forth by the National Research Council
(1983). U.S. EPA guidelines that were used in the development of this assessment may include
the following: Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA,
1986a), Guidelines for Mutagenicity Risk Assessment (U.S. EPA, 1986b), Guidelines for
Developmental Toxicity Risk Assessment (U.S. EPA, 1991), Guidelines for Reproductive Toxicity
Risk Assessment (U.S. EPA, 1996), Guidelines for Neurotoxicity Risk Assessment (U.S. EPA,
1998a), Guidelines for Carcinogen Assessment (U.S. EPA, 2005), Recommendations for and
Documentation of Biological Values for Use in Risk Assessment (U.S. EPA, 1988), (proposed)
Interim Policy for Particle Size and Limit Concentration Issues in Inhalation Toxicity (U.S.
EPA, 1994a), Methods for Derivation of Inhalation Reference Concentrations and Application of
Inhalation Dosimetry (U.S. EPA, 1994b), Use of the Benchmark Dose Approach in Health Risk
Assessment (U.S. EPA, 1995), Science Policy Council Handbook: Peer Review (U.S. EPA,
1998b, 2000a), Science Policy Council Handbook: Risk Characterization (U.S. EPA, 2000b),
Benchmark Dose Technical Guidance Document (U.S. EPA, 2000c), Supplementary Guidance
for Conducting Health Risk Assessment of Chemical Mixtures (U.S. EPA, 2000d), and^4 Review
of the Reference Dose and Reference Concentration Processes (U.S. EPA, 2002a).
The chapter on occurrence and exposure to boron through potable water was developed
by the Office of Ground Water and Drinking Water. It is based primarily on the National
Inorganic and Radionuclide Survey (NIRS) data for boron. The NIRS data are supplemented
with ambient water data, as well as data from the States, and published papers on occurrence in
drinking water.
Proposal Draft
Boron — September, 2006
-------
ACKNOWLEDGMENT
This document was prepared under the U.S. EPA Contract No. 68C-02-009, Work
Assignment No.354 with ICF Consulting. The Lead U.S. EPA Scientist is Santhini Ramasamy,
PhD, MPH, DABT, Health and Ecological Criteria Division, Office of Science and Technology,
Office of Water.
Proposal Draft
Boron — September, 2006 iii
-------
TABLE OF CONTENTS
FOREWORD i
ACKNOWLEDGMENT iii
1.0 EXECUTIVE SUMMARY 1-1
2.0 IDENTITY: CHEMICAL AND PHYSICAL PROPERTIES 2-1
3.0 USES AND ENVIRONMENTAL FATE 3-1
3.1 Production and Use 3-1
3.2 Environmental Release 3-3
3.3 Environmental Fate 3-4
3.4 Summary 3-6
4.0 EXPOSURE FROM DRINKING WATER 4-1
4.1 Introduction 4-1
4.2 Ambient Occurrence 4-1
4.2.1 Data Sources and Methods 4-1
4.2.2 Results 4-1
4.3 Drinking Water Occurrence 4-2
4.3.1 Data Sources and Methods 4-2
4.3.2 Derivation of the Health Reference Level 4-3
4.3.3 Results 4-4
4.4 Summary 4-4
5.0 EXPOSURE FROM MEDIA OTHER THAN WATER 5-1
5.1 Exposure from Food 5-1
5.1.1 Concentration in Non-Fish Food Items 5-1
5.1.2 Concentrations in Fish and Shellfish 5-1
5.1.3 Intake of Boron from Food 5-1
5.2 Exposure from Air 5-1
5.3 Exposure from Soil 5-2
5.3.1 Concentration of Boron in Soil 5-2
5.3.2 Intake of Boron from Soil 5-3
5.4 Other Residential Exposures 5-3
5.5 Occupational (Workplace) Exposures 5-3
5.5.1 Description of Industries and Workplaces 5-3
5.5.2 Types of Exposure (Inhalation, Dermal, Other) 5-3
5.5.3 Concentrations of Boron in the Work Environment 5-3
5.6 Summary 5-4
6.0 TOXICOKINETICS 6-1
Proposal Draft
Boron — September, 2006 iv
-------
6.1 Absorption 6-1
6.2 Distribution 6-3
6.3 Metabolism 6-6
6.4 Excretion 6-6
7.0 HAZARD IDENTIFICATION 7-1
7.1 Studies in Humans - Epidemiology and Case Reports 7-1
7.1.1 Oral Exposure 7-1
7.1.2 Inhalation Exposure 7-3
7.2 Prechronic and Chronic Studies and Cancer Bioassays in Animals - Oral and
Inhalation 7-5
7.2.1 Oral Exposure 7-5
7.2.2 Inhalation Exposure 7-10
7.3 Developmental/Reproductive Toxicity 7-11
7.3.1 Developmental Studies 7-11
7.3.2 Reproductive Studies 7-16
7.3.2.1 Male-Only Exposure 7-16
7.3.2.2 Male and Female Exposure 7-19
7.4 Other Studies 7-21
7.4.1 Genotoxicity Studies 7-21
7.4.2 Neurological Studies 7-22
7.4.3 Mechanistic Studies - Testicular Effects 7-23
7.4.4 Mechanistic Studies - Developmental Effects 7-23
7.4.5 Nutrition Studies 7-23
7.5 Synthesis and Evaluation of Major Noncancer Effects and Mode of Action - Oral
and Inhalation 7-24
7.5.1 Oral Exposure 7-24
7.5.2 Inhalation Exposure 7-25
7.6 Weight of Evidence Evaluation and Cancer Characterization - Synthesis of
Human, Animal, and Other Supporting Evidence, Conclusions About Human
Carcinogenicity, and Likely Mode of Action 7-26
7.7 Susceptible Populations 7-26
7.7.1 Possible Childhood Susceptibility 7-26
7.7.2 Possible Gender Differences 7-26
7.7.3 Physiological and Disease Anomalies 7-27
8.0 DOSE-RESPONSE ASSESSMENT 8-1
8.1 Oral Reference Dose 8-1
8.1.1 Choice of Principal Study and Critical Effect — with Rationale and
Justification 8-1
8.1.2 Methods of Analysis — Including Models 8-2
8.1.3 Derivation of the RfD 8-3
8.1.3.1 Derivation of Adjustment Factor Values 8-4
8.1.3.2 Toxicokinetic Modeling Issues for Boron 8-6
Proposal Draft
Boron — September, 2006
-------
8.1.3.3 Summary of Data-Derived Adjustment Factors and RfD
Calculation 8-13
8.1.3.4 Other Uncertainty Factor Approaches 8-14
8.1.4 Previous Oral Assessment 8-16
8.2 Inhalation Reference Concentration (RfC) 8-17
8.3 Cancer Assessment 8-17
8.4 CCL Health Reference Level 8-17
9.0 REGULATORY DETERMINATION AND CHARACTERIZATION OF RISK FROM
DRINKING WATER 9-1
9.1 Regulatory Determination for Chemicals on the CCL 9-1
9.1.1 Criteria for Regulatory Determination 9-1
9.1.2 National Drinking Water Advisory Council Recommendations 9-2
9.2 Health Effects 9-2
9.2.1 Health Criterion Conclusion 9-2
9.2.2 Hazard Characterization and Mode of Action Implications 9-3
9.2.3 Dose-Response Characterization and Implications in Risk Assessment
9-4
9.3 Occurrence in Public Water Systems 9-4
9.3.1 Occurrence Criterion Conclusion 9-5
9.3.2 Monitoring Data 9-5
9.3.3 Use and Fate Data 9-6
9.4 Risk Reduction 9-6
9.4.1 Risk Criterion Conclusion 9-7
9.4.2 Exposed Population Estimates 9-7
9.4.3 Relative Source Contribution 9-7
9.4.4 Sensitive Populations 9-8
9.5 Regulatory Determination Decision 9-8
10.0 REFERENCES 10-1
APPENDIX A: Abbreviations and Acronyms Appendix A-l
Proposal Draft
Boron — September, 2006 vi
-------
LIST OF TABLES
Table 2-1
Table 3-1
Table 3-2
Table 4-1
Table 5-1
Table 6-1
Table 6-2
Table 6-3
Table 6-4
Table 6-5
Table 8-1
Table 8-2
Chemical and Physical Properties of Boron and Related Compounds
2-4
Environmental Releases (in pounds) of Boron Trichloride in the United States,
1995-2002 3-4
Environmental Releases (in pounds) of Boron Trifluoride in the United States,
1995-2002 3-4
Summary Occurrence Statistics for Boron in Ground Water Systems
4-6
Mean Intake of Boron (mg/day) from Food Based on the Continuing Survey of
Food Intake by Individuals 1994-1996 5-2
Tissue Levels of Boron in Male Rats on Day 7 of Exposure to 9000 ppm Boric
Acid (1575 ppm boron) in the Diet (jig boron/g tissue) 6-4
Renal Boron Clearance (mL/rnin/1.73m2) Calculated from Dietary Exposure and
Intravenous Infusion 6-8
Urinary Boron Concentration, Volume, Mean Excretion, and Percent Recovered
in 12 Hours in Nonpregnant and Pregnant Rats Given Boric Acid by Gavage 6-11
Clearance of Boron (Boric Acid), Creatinine and Urea in Nonpregnant and
Pregnant Rats Given Boric Acid by Gavage Expressed as mL/min, mL/min/cm2,
and mL/min/kg 6-13
Urinary Clearance of Boron in Pregnant Women 6-16
Sigma-method Value Calculation 8-13
Default and Data-derived Values for Components of UFA and UFH 8-13
Proposal Draft
Boron — September, 2006
Vll
-------
LIST OF FIGURES
Figure 2-1 Chemical Structure of Boric Acid 2-2
Figure 2-2 Chemical Structure of Borax (Sodium Tetraborate Decahydrate) 2-2
Figure 2-3 Chemical Structure of Boron Oxide 2-3
Proposal Draft
Boron — September, 2006 viii
-------
1.0 EXECUTIVE SUMMARY
The U.S. Environmental Protection Agency (EPA) has prepared this Health Effects
Support Document for Boron to assist in determining whether to regulate boron with a National
Primary Drinking Water Regulation (NPDWR). The available data on occurrence, exposure, and
other risk considerations suggest that boron does not occur in public water systems at a
frequency and at levels of public health concern at the present time. Based on the low
occurrence of boron in the potable water, and on its natural occurrence in the environment, boron
does not present a meaningful opportunity for health risk reduction for persons served by public
water systems. EPA presents its determination and data analysis in the Federal Register Notice
covering the Contaminant Candidate List (CCL) regulatory determinations.
Boron is a metalloid element from Group IIIA of the periodic table. Naturally-occurring
boron usually is found in sediments and sedimentary rock formations and rarely exists in
elemental form. Other forms of boron include boric acid, borax, borax pentahydrate, anhydrous
borax, and boron oxide. The principal uses for boron compounds in the United States include
glass and ceramics, soaps and detergents, algicides in water treatment, fertilizers, pesticides,
flame retardants, and reagents for production of other boron compounds.
The major sources of free boron in the environment are exposed minerals containing
boron, boric acid volatilization from seawater, and volcanic material. Global releases of
elemental boron through weathering, volcanism, and other geothermal processes are estimated at
approximately 360,000 metric tons annually (Moore, 1991). Anthropogenic inputs of boron to
natural environments are considered smaller than inputs from natural processes. The following
human activities release boron to the environment: agriculture, waste and wood burning, power
generation using coal and oil, glass product manufacture, use of borates/perborates in the home
and industry, borate mining/processing, leaching of treated wood, and sewage/sludge disposal.
Contamination of water can come directly from industrial wastewater and municipal sewage, as
well as indirectly from air deposition and soil runoff. Borates in detergents, soaps, and personal
care products can also contribute to the presence of boron in water.
The available data for boron support its ubiquitous presence in the ambient environment.
TRI data for the years 1995 to 2002 on total releases for boron trichloride (on- and off-site) have
fluctuated within the range of hundreds of pounds per year. Boron trifluoride releases for the
years 1995 to 2002 are similarly dominated by on-site air emissions, with releases ranging in the
tens of thousands of pounds annually. In drinking water, approximately 81.9% of groundwater
public water systems (PWSs) had detections of boron (>minimum reporting level (MRL) of
0.005 mg/L). These detections affected about 88.1% of the population served by the public
water systems, equivalent to approximately 75.5 million people served by ground water
nationally. Detections at a concentration greater than one-half the health reference level (HRL)
of 0.7 mg/L occurred in 4.3% of surveyed PWSs, affecting 2.9% of the population served,
equivalent to approximately 2.5 million people nationally. Concentrations greater than the HRL
(1.4 mg/L) were found in approximately 1.7% of surveyed PWSs, affecting 0.4% of the
population served, equivalent to approximately 0.4 million people nationally. Supplementary
data from an AWWARF-sponsored study indicate that boron contamination of surface water is
Proposal Draft
Boron — September, 2006 1-1
-------
less significant than boron contamination of ground water. Of 228 ground water and 113 surface
water samples analyzed, boron was detected in 99.1% of the ground water samples and 97.3% of
the surface water samples. Boron was detected at a concentration greater then one-half the
health reference level (>/^HRL or >0.7 mg/L) in 8.8% of the ground water samples and none of
the surface water samples. Boron was detected at concentrations greater than the HRL (>HRL or
>1.4 mg/L) in 3.1 % of the ground water samples and in none of the surface water samples.
Studies in both humans and animals show that boron is readily absorbed from the
gastrointestinal tract (the absorption evidence is weak from the respiratory tract as described).
Boric acid and borate compounds in the body exist primarily as undissociated boric acid, which
distributes evenly throughout the soft tissues, but shows some accumulation in bone. In several
animal studies, boron levels in all tissues, except adipose, increased rapidly after the start of
dietary exposure, with the greatest increase in bone. In one study, boron showed a 2- to 3-fold
increase over plasma levels after 7 days. In another study, concentrations of boron in bone in
exposed animals were 5- to 6-fold higher than in unexposed controls after eight weeks of
recovery; thirty-two weeks after recovery bone boron concentrations remained 3-fold higher in
treated groups than in controls.
Inorganic borate compounds are present as boric acid in the body. Boric acid is the only
boron compound that has been identified in urine, and it has repeatedly been found to account
for >90% of the ingested boron dose. There is no evidence that boric acid is degraded in the
body. Metabolism may not be feasible because a large amount of energy (523 kJ/Mol) is
apparently required to break the boron-oxygen bond. Boric acid can form complexes with
various biomolecules. It has an affinity for hydroxyl, amino, and thiol groups. Complex
formation is concentration dependent and reversible. The primary route of excretion of boron is
in the urine.
Boron is a trace element for which essentiality is suspected but has not been directly
proven in humans. The National Academy of Science Institute of Medicine categorizes boron as
a possible trace mineral nutrient for humans. Boron is essential for plant growth. Deficiency
studies in animals and humans have provided some evidence that low intakes of boron affect
cellular function and the activity of other nutrients. It may interact with Vitamin D and calcium,
influence estrogen metabolism, and play a role in cognitive function. The average dietary intake
for male adults is about 1.5 mg/day.
Some human oral data are available from cases where boron was ingested for medical
reasons. When the amount ingested was less than 3.68 mg/kg, subjects were asymptomatic,
while doses of 20 and 25 mg/kg resulted in nausea and vomiting. Case reports and surveys of
accidental poisonings indicate that the lethal doses of boron range from 15 to 20 grams
(approximately 200 to 300 mg/kg) for adults, 5 to 6 grams (approximately 70 to 85 mg/kg) for
children, and 2 to 3 grams (approximately 30 to 45 mg/kg) for infants.
The primary adverse effects seen in animals after chronic exposure to low doses of boron
generally occur in testes and fetuses. Chronic effects of dietary boron exposure in two-year
studies included the following: testicular atrophy and spermatogenic arrest in dogs, decreased
Proposal Draft
Boron — September, 2006 1 -2
-------
food consumption, suppressed growth, and testicular atrophy in rats, and decreased survival,
testicular atrophy, and interstitial cell hyperplasia in mice. Although researchers observed some
increases in tumor incidences in the liver and in subcutaneous tissues in mice, based on
comparisons to historic controls these tumors were determined not to be associated with
exposure to boron from boric acid. Boron is not considered mutagenic. EPA has determined
that there are inadequate data to assess the human carcinogenic potential for boron.
In developmental studies with rats, mice, and rabbits, oral exposure to boric acid resulted
in decreased pregnancy rates, increased prenatal mortality, decreased fetal weights, and
increased malformations in fetuses and pups. These reproductive effects were associated with
maternal toxicity, including changes in maternal organ weights, body weights, weight gain, and
increased renal tubular dilation and/or regeneration. Reproductive effects in males were noted in
the subchronic and chronic studies described above.
The EPA reference dose (RfD) for boron is 0.2 mg/kg/day based on developmental
effects in rats from two studies. The RfD was derived using the benchmark dose (BMD) method
and a data-derived uncertainty factor of 66. EPA established the Health Reference Level (HRL)
for boron (1.4 mg/L or 1400 |ig/L) using the RfD of 0.2 mg/kg-day and a 20 percent relative
source contribution.
EPA evaluated whether health information is available regarding the potential effects on
children and other sensitive populations. Studies in rats, mice, and rabbits identify the
developing fetus as potentially sensitive to boron. Price et al. (1996a) identified a lowest
observed adverse effect level (LOAEL) of 13.3 mg/kg-day and a no observed adverse effect
level (NOAEL) of 9.6 mg/kg-day in the developing rat fetus, based on decreased fetal body
weight. Accordingly, boron at concentrations greater than the HRL might affect prenatal
development. Individuals with impaired kidney function might be more sensitive to boron
exposure than the general population since the kidney is the main route for excretion.
Based on the concentrations of boron in the potable water where it occurs relative to the
HRL, boron does not present a meaningful opportunity for health risk reduction for persons
served by public water systems.
Proposal Draft
Boron — September, 2006 1 -3
-------
2.0 IDENTITY: CHEMICAL AND PHYSICAL PROPERTIES
Boron is a metalloid element from Group IIIA of the periodic table with an atomic
number of 5, atomic weight of 10.81, and oxidation state of+3. Boron exists naturally as 19.78%
10B isotope and 80.22% UB isotope (WHO, 1998a). It is a polymorphic element that exists in a
variety of different crystalline forms: a-rhombohedral (clear red crystals); p-a-rhombohedral
(black); a-tetragonal (black, opaque crystals with metallic luster); amorphous (black or dark
brown powder); and yellow monoclinic crystals or brown amorphous powder (O'Neil et al.,
2001; Weast, 1988-1989). Elemental boron is insoluble in water, but if finely divided, it is
soluble in boiling sulfuric acid and in most molten metals, such as copper, iron, magnesium,
aluminum, and calcium. Elemental boron undergoes an oxidation reaction upon exposure to
oxygen which is limited by the formation of a protective boric oxide film. This film evaporates
at temperatures above 1000°C. At room temperature, boron is a poor conductor of electricity,
but its conductivity increases at higher temperatures (O'Neil et al., 2001). Technical grade
boron has 90-92% boron content (Sax and Lewis, 1987) and can include impurities such as
carbon, oxygen, hydrogen, and nitrogen. Impurities in ultrapure boron are usually below the
0.5% range (Kroschwitz and Howe-Grant, 1992).
Boron is electron-deficient, possessing a vacant p-orbital; it does not form ionic bonds,
but does form stable covalent bonds. Compounds of boron often behave as Lewis acids, readily
bonding with electron-rich substances.
Boric acid (Figure 2-1) exists as odorless, colorless, translucent crystals or white granules
or powder at ambient temperatures (O'Neil et al., 2001). It is a weak acid with a pKa of 9.2 (pH
5.1 when in a 0.1 molar solution) and exists primarily as the undissociated acid (H3BO3) in
aqueous solution at physiological pH, as do borate salts (Woods, 1994). Three grades of
granular and powdered boric acid are manufactured in the United States, i.e., technical grade
(99.9%), NF grade, and special quality grade. The principal impurities in technical grade boric
acid are sulfate (0.1%) and various minor metallic impurities present in borate ore (Kirk-Othmer,
1984).
Borax (Figure 2-2) is an odorless substance that exists in the form of white or colorless
monoclinic crystals. Its solutions have alkaline properties, but do not cause corrosion to ferrous
metals (HSDB, 2003c). Borax is produced as crystals, granules, and powder (Sax and Lewis,
1987). Technical borax is an herbicide, also known as "Nippon" insecticide, while refined borax
is known as sodium tetraborate decahydrate (99% purity). Mixtures include brocil (borax +
bromacil), ureabor (borax + monuron), and borax + sodium chlorate (Worthing, 1987; Weed
Science Society of America, 1983). Anhydrous borax is an odorless, hygroscopic substance that
exists as white to gray powder or as glass-like plates (HSDB, 2003d). It is produced from its
hydrated forms by fusion, usually through an intermediate step involving calcining
(Kirk-Othmer, 1984).
Technical grade anhydrous borate (borax) contains 99% sodium tetraborate and comes in
fine, granular form, as glass (fused; Sax and Lewis, 1987).
Proposal Draft
Boron — September, 2006 2-1
-------
Boron oxide (Figure 2-3) is an odorless, slightly bitter substance, which at ambient
temperatures exists in the form of colorless, semitransparent lumps or hard, white crystals.
These solids are brittle and hygroscopic and they slowly react with water to form boric acid.
Boron oxide is soluble in alcohol and glycerol; it is corrosive to metals in the presence of oxygen
(O'Neil et al., 2001; Kirk-Othmer, 1984). Both technical and high-purity (99.99%) grades of
boron oxide are manufactured in a glass (fused) or powdered form (Sax and Lewis, 1987).
Figure 2-1 Chemical Structure of Boric Acid
Source: Chemfinder.com (2004)
Figure 2-2 Chemical Structure of Borax (Sodium Tetraborate Decahydrate)
Source: Chemfmder.com (2004)
Proposal Draft
Boron — September, 2006 2-2
-------
Figure 2-3 Chemical Structure of Boron Oxide
O
o
Source: Chemfinder.com (2004)
The chemical structures of boric acid, borax, and boron oxide are shown above (Figures
2-1 through 2-3); the chemical structure of elemental boron, borax pentahydrate, and anhydrous
borax were not available (Chemfmder.com, 2004). The physical and chemical properties, and
other reference information on boron, boric acid, borax, borax pentahydrate, anhydrous borax,
and boron oxide are listed in Table 2-1.
Proposal Draft
Boron — September, 2006 2-3
-------
Table 2-1 Chemical and Physical Properties of Boron and Related Compounds
Property
Chemical Abstracts
Registry (CAS) No.
U.S. EPA Pesticide
Chemical Code
Synonyms
Chemical Formula
Molecular Weight
Physical State
Boiling Point
Melting Point
Density (at 20 °C)
Vapor Pressure:
At 20 °C
At 25 °C
Log Kow
Log Knr
Solubility in:
Water
Other Solvents
Boron
7440-42-8
128945
none identified
B
10.81
Solid; black
crystal or yellow-
brown amorphous
powder
2,550°C
2,300°C
2.34
1.56xlO-5atm(at
2,140°C)
none identified
none identified
none identified
Insoluble
none identified
Boric Acid
10043-35-3
011001
boron trihydroxide;
trihydroxy borate;
orthoboric acid;
boracic acid
LLBO,
61.83
Solid; white or
colorless crystalline
granules or powder
30°C
171 °C (closed space)
450°C (anhydrous,
crystal form)
1.51
none identified
none identified
none identified
none identified
55.6 g/L cold water3
250 g/L boiling water3
methanol, acetone,
alcohol, glycerol
Borax
1303-96-4
029601 or 01 1102
disodium tetraborate
decahydrate, borax
decahydrate, borax 10
Na,B407.10H,0
381.43
Solid; white or colorless
crystalline granules or
powder
none identified
>62°C (closed space)
1.73
none identified
none identified
none identified
none identified
62.5g/Lat25°C
glycerol
Borax Pentahydrate
12179-04-3
11130-12-4
011110
Sodium tetraborate
pentahydrate; Borax 5
Na,B407.5H,0
291.35
Solid; white or colorless
crystalline granules or
powder
none identified
<200°C (closed space)
1.81
none identified
none identified
none identified
none identified
35.9 g/L at 20°C
482.4 g/L at 100°C
glycerol
Anhydrous Borax
1330-43-4
011112
Sodium tetraborate;
borax glass; disodium
tetraborate; fused
borax
Na,B407
201.27
Solid; white or
colorless vitreous
granules
1,575°C (decomposes)
742°C
2.37
none identified
none identified
none identified
none identified
24.8 g/L at 20°C
331.2 g/L at 100°C
ethylene glycol
Boron Oxide
1303-86-2
011002
Boric oxide; boron
trioxide; anhydrous
boric acid
B,0,
69.62
Solid; white or
colorless vitreous
granules
1,860°C
450°C
2.46
none identified
none identified
none identified
none identified
rapidly hydrates to
boric acid
alcohol, glycerol
Source(s): HSDB (2003a-e); Weast (1988-1989); O'Neil et al. (2001)
a Water temperature was not defined.
Proposal Draft
Boron September, 2006
2-4
-------
3.0 USES AND ENVIRONMENTAL FATE
3.1 Production and Use
Elemental boron occurs naturally and is found in borax ore or tincal (Na2B4O7.10H2O),
boric acid (H3BO3), colemanite (CaB3O4(OH)3-H2O), kernite or rasorite (Na2B4O7-4H2O), ulexite
(NaCaB5Cy8H2O), and borates (salt or ester of boric acid). Boric acid is sometimes found in
volcanic spring waters. Ulexite is a borate mineral that naturally has fiber optic properties.
In 2003, the United States was the world's largest producer of refined boron compounds.
About one-half of the domestic production (1060 x 103 metric tons) was exported. Domestic
production of boron minerals, primarily as sodium borates, by four companies was centered in
southern California (USGS, 2004). The largest company produced and processed ore from an
open pit mine; the second company produced boron, sodium carbonate, and sodium sulfate from
brines; the third company has since ceased production; the fourth operates an underground mine
in California and processes the ore in Nevada for overseas export. U.S. processed products had
fewer impurities and lower emissions than products from other countries (USGS, 2004).
Elemental boron production methods include chemical reduction with reactive elements,
nonaqueous electrolytic reduction, or thermal decomposition of the oxide. Purification to
ultrapure boron is accomplished by zone-refining or other thermal techniques. Another method
for boron production is by electrolysis of fused melts with a boron carbide anode (Kirk-Othmer,
1984).
Boric acid is produced by reacting borax or other borates with hydrochloric or sulfuric
acid (Osol, 1980). An alternative method employs extraction from weak borax brines with a
kerosine solution or a chelating agent, such as 2-ethyl-l,3-hexanediol or other polyols. The
chelates are subsequently removed by sulfuric acid (Sax and Lewis, 1987).
Commercial production of borax involves the processing of sodium borate ores by
crushing, heating, mechanical separation, selective crystallization, and then flotation of borax
decahydrate or pentahydrate from the resultant concentrated borax liquor (HSDB, 2003c).
Boron is used in nuclear chemistry as a radiation shield and for neutron-detecting
instrumentation (Weast, 1988-1989). It is a deoxidizer in nonferrous metallurgy and ignition
rectifiers for rockets and radio tubes (O'Neil et al., 2001). Boron also is used in aluminum as a
grain refiner for delayed action fuses, solar battery coatings (Clayton and Clayton, 1994), iron
cementation, wire-coatings for semiconductors, and high temperature abrasive alloys (Sax and
Lewis, 1987). Boron is a catalyst in olefm polymerization and alcohol dehydration (Kroschwitz
and Howe-Grant, 1992).
Borax (hydrous or anhydrous) and boric acid are widely used for a wide range of
industrial purposes. Major applications are in the manufacture of porcelain enamel, ceramic
glazes, and metal alloys, and to enhance thermal properties of glass and durability of fiberglass
insulation. These compounds also are commonly used in fire retardants in cellulose insulation,
wood and textiles, laundry additives, herbicides, fertilizers (boron is an essential element for
Proposal Draft
Boron — September, 2006 3-1
-------
plants), and insecticides (HSDB, 2003c,d; Woods, 1994). Boric acid, borates, and perborates
have been used as ingredients in mild antiseptics or bacteriostats in eyewashes, mouthwashes,
burn dressings, and diaper rash powders, although their effectiveness has largely been
discredited (Seller et al., 1988).
Borax is used in the following diverse applications: tanning, artificial aging of wood,
fireproofing fabrics and woods, curing and preserving hides, soldering metals, and inhibition of
wood fungus rot. It also is used in antiseptics, detergents, and astringents, antifreeze, plant
fertilizers, nonselective herbicides, and soil sterilants (Kroschwitz and Howe-Grant, 1992;
O'Neil et al., 2001; Sax and Lewis, 1987; U.S. FDA, 1988).
Borax decahydrate is a commercial fungicide for citrus (Spencer, 1982), an ingredient in
household germicidal cleaning products, a chemical intermediate in the productions of
perborates and other boron derivatives, a flux in the nonferrous metallurgy, and an additive in
ferrous and nonferrous boron alloys production (HSDB, 2003d).
Commercial anhydrous borax is an industrial water algicide, corrosion inhibitor,
emulsifying agent in cosmetics, and a buffer component in a variety of products (Oilman et al.,
1990; Kroschwitz and Howe-Grant, 1992; O'Neil et al., 2001; Sax and Lewis, 1987; U.S. FDA,
1988).
Boric acid is used in printing, dyeing, painting, leather making, and hard-steel
production. It is used in the manufacture of soaps, artificial gems, electric condensers (O'Neil et
al., 2001), paper products for food packaging, adhesives, sizes, and coatings (U.S. FDA, 1988).
It is the key raw ingredient in the manufacture of synthetic inorganic borate salts, boron
phosphate, fluoborate, borate esters, and metal alloys such as ferroboron (Kroschwitz and Howe-
Grant, 1992). Boric acid is a component of high contrast lith-type film developer formula (e.g.,
Kodak D-85), an additive in nuclear-reactor cooling water, and a catalyst for alcohol production
from air oxidation of hydrocarbons (Kroschwitz and Howe-Grant, 1992), and a constituent of
insect baits, repellants, and poisons (Rossoff, 1974; Meister, 1989).
Boric oxide is used as a chemical intermediate for obtaining elemental boron, boron
master alloys, borides, boron carbide, nitrides and halides. It is a fire resistant ingredient in
paints and electronic products. It also is used in liquid encapsulation techniques and blowpipe
analysis, and protocols used to determine silicon dioxide and alkalide presence in silicates
(National Fire Protection Association, 1997; O'Neil et al., 2001).
The principal uses for boron compounds in the United States in 2001 were estimated as
follows: 78% glass and ceramics; 6% soaps and detergents; 4% agriculture; 3% flame retardants;
and 9% as other boron compounds (USGS, 2004). The use pattern for borax (decahydrate,
pentahydrate, and anhydrous) is: 23% in insulation glass fibers; 20% in household cleaning
products as germicide; 11% in borosilicate glasses; 11% as algicide in water treatment; 8% in
enamel flux, frits, and glazes; 8% as chemical intermediate for perborates; 7% in fertilizers; 5%
as antifreeze corrosion inhibitor; 4% as a chemical intermediate for other boron compounds; 3%
in herbicides; 1% as flame retardant and metallurgical flux; and 10% in other miscellaneous
Proposal Draft
Boron — September, 2006 3-2
-------
applications (HSDB, 2003a). Borate consumer uses in 1985 were estimated as follows: 18%
glass fiber insulation; 11% textile glass fiber; 15% chemical fire retardants; 5% borosilicate
glass; 4% soap and detergents; 13% miscellaneous; and 44% exports (HSDB, 2003a).
3.2 Environmental Release
The United States, Turkey, and Russia are the leading producers of boron compounds,
followed by Argentina, Chile, and China (USGS, 2004). In 2003, Turkey produced the greatest
quantity of ore, while the U.S. led in production of refined boron compounds. U.S. boron
resources, mostly sediments and brines, are primarily located in California. U.S. production of
boron compounds between 1999 and 2003 ranged from between 518,000 metric tons and
618,000 metric tons (measured as boric oxide). In 2003, the U.S. imported approximately
174,000 metric tons of boron compounds and exported approximately 244,000 metric tons
(USGS, 2004).
Boron is a naturally occurring compound, usually found in various inorganic forms in
sediments and sedimentary rocks. The richest known boron-containing deposits in the U.S. are
found in California. Boron presents in water, soil, and air originates from both natural and
anthropogenic sources.
Natural weathering of boron-containing rocks is thought to be the primary source of
boron compounds in water and soil (Butterwick et al., 1989). Releases to air from oceans,
volcanos, and geothermal steam are other natural sources of boron in the environment (Graedel,
1978). Global releases of elemental boron through weathering, volcanic, and geothermal
processes are estimated at approximately 360,000 metric tons annually (Moore, 1991).
Human causes of boron contamination include releases to air from power plants,
chemical plants, and manufacturing facilities. Fertilizers, herbicides, and industrial wastes are
among the sources of soil contamination. Contamination of water can come directly from
industrial wastewater and municipal sewage, as well as indirectly from air deposition and soil
runoff (ATSDR, 1992). Borates in detergents, soaps, and personal care products can also
contribute to the presence of boron in water.
Boric acid and its sodium salts are registered for use as pesticides. Data from the U.S.
Bureau of Mines, cited in the U.S. EPA's 1994 reregi strati on eligibility document for boron
pesticides (U.S. EPA, 1994c), suggests that approximately 293,000 pounds of boron minerals
were used annually for "agricultural purposes" during a period around 1990. In the re-
registration eligibility document, the U.S. EPA stated that the amount of boron used specifically
as pesticide is somewhat less than the amount used for other agricultural purposes, and that boric
acid use declined significantly during the 1980s (U.S. EPA, 1994c).
Two anthropogenic boron compounds, boron trichloride and boron trifluoride, are listed
as Toxic Release Inventory (TRI) chemicals.
Proposal Draft
Boron — September, 2006 3-3
-------
TRI data for boron trichloride (see Table 3-1) are reported for the years 1995 to 2002
(U.S. EPA, 2004b). For boron trichloride, on-site air emissions constitute all of the total releases
(on- and off-site), and these have generally fluctuated in the range of hundreds of pounds per
year during the period of record. TRI releases for boron trichloride were reported from Arizona,
California, Indiana, New Mexico, Pennsylvania, and Wisconsin.
Table 3-1 Environmental Releases (in pounds) of Boron Trichloride in the United
States, 1995-2002
Year
2002
2001
2000
1999
1998
1997
1996
1995
On-Site Releases
Air
Emissions
258
626
605
350
750
754
37
5
Surface Water
Discharges*
0
0
-
-
0
0
-
-
Underground
Injection
0
0
0
0
0
0
0
0
Releases
to Land
0
0
0
0
0
0
0
0
Off-Site
Releases
0
0
0
0
0
0
0
0
Total On- &
Off-site
Releases
258
626
605
350
750
754
37
5
Source: U.S. EPA (2004b)
* "-" denotes blank cells on reporting forms. "0" is entered when the reporting forms contained only zeros or NAs.
Boron trifluoride releases, also for the years 1995 to 2002 (see Table 3-2), are similarly
dominated by on-site air emissions. Boron trifluoride releases ranged in the tens of thousands of
pounds annually. TRI releases for boron trifluoride were reported from 14 States (AL, AR, DE,
FL, KY, LA, MD, NY, OH, OK, PA, SC, TN, and TX) (U.S. EPA, 2004b).
Table 3-2 Environmental Releases (in pounds) of Boron Trifluoride in the United
States, 1995-2002
Year
2002
2001
2000
1999
1998
1997
1996
1995
On-Site Releases
Air
Emissions
10,114
11,496
11,595
16,725
37,802
21,290
29,881
25,019
Surface Water
Discharges
0
0
0
0
5
0
0
0
Underground
Injection
0
0
0
0
0
0
0
0
Releases
to Land
0
0
0
0
0
0
0
0
Off-Site
Releases
0
0
250
0
0
5
0
929
Total On- &
Off-site
Releases
10,114
11,496
11,845
16,725
37,807
21,295
29,881
25,948
Source: U.S. EPA (2004b)
3.3 Environmental Fate
Boron in the environment primarily derives from the weathering of minerals containing
boron, seawater volatilization producing boric acid, and volcanic activity. Anthropogenic
sources of boron are considered to contribute a lesser amount to the environment than natural
processes. Anthropogenic sources of boron are as follows: agricultural, waste and fuel wood
burning, power generation using coal and oil, glass product manufacture, use of
Proposal Draft
Boron — September, 2006
3-4
-------
borates/perborates in the home and industry, borate mining/processing, leaching of treated
wood/paper, and sewage/sludge disposal of boron (HSDB, 2003a).
Atmospheric boron occurs as particulate matter or aerosols, as borides, boron oxides,
borates, boranes, organoboron compounds, trihalide boron compounds, or borazines. Borates are
relatively soluble in water and will probably be removed from the atmosphere by precipitation
and dry deposition (U.S. EPA, 1987). The half-life of boron and boron containing compounds in
the atmosphere was estimated to be on the order of days (Nriagu, 1979) with particle size
determining the length of time in the atmosphere. Transformation or degradation of boron
particulates in the atmosphere has not been studied.
Most boron compounds are soluble in water while the solubility of elemental boron is
very low. Due to the high water-solubility of the environmentally-relevant boron minerals, Rai
et al. (1986) concluded that it is unlikely that mineral equilibria will control the fate of boron in
water. Boron compounds such as borax rapidly hydrolyze to form a boric acid-borate mixture.
Boric acid is a weak acid that exists primarily in its unionized for at pHs below 7. Borate and
boric acid establish an equilibrium reaction in water that is dependant on the pH.
B(OH)3 + H2O * B(OH)4- + H+ pKa = 9.14
The extent of boron adsorption depends on the pH of the water and the chemical
composition of the soil or sediment. The greatest adsorption is generally observed at pH 7.5-9.0
(Keren et al., 1981; Keren and Mezuman, 1981; Waggott, 1969) with amorphous aluminum
oxide (Bingham et al., 1971), iron oxide (Sakata, 1987), and, to a lesser extent, organic matter
present in the soil (Parks and White, 1952). Boron is adsorbed mainly on the edge surfaces of
2:1 clay minerals (Keren and Bingham, 1985; Keren and Sparks, 1994; Keren and Talpaz, 1984).
Some clay materials, e.g., montmorillonite, have a negative electric field, which makes them less
accessible to approaching borate anions (Secor and Radke, 1985).
Boron in the soil may adsorb onto iron and aluminum hydroxy compounds and clay
minerals. Boron sorption by clay minerals and iron and aluminum oxides is pH dependent, with
maximum sorption in the range 7-9. The amount of boron adsorbed depends on the surface area
of the clay or oxide, and this sorption is only partially reversible (Brown et al., 1983). Finer
textured soils retain boron longer than do coarse, sandy soils. Keren and Mezuman (1981)
determined that the amount of organic matter present in water systems was not as important in
adsorption of boron as the inorganic minerals present.
Borax may be persistent for one or more years, depending on soil type and rainfall.
Borax is less persistent in acidic soils and in high rainfall areas, with it leaching rapidly under
high rainfall conditions (Weed Science Society of America, 1983). Boron is thought to
accumulate in plants. Boron content of lentil and barley plants from soil treated with 8 ppm
boron was approximately 7- and 8-fold that of control plants, respectively (Singh and Singh,
1984).
Proposal Draft
Boron — September, 2006 3-5
-------
3.4 Summary
Boron enters the environment primarily by the weathering of rock strata containing boron
minerals, boric acid volatilization from seawater, and volcanic activity. Anthropogenic inputs
are lower than natural inputs (HSDB, 2003a). Atmospheric boron usually exists as particulates,
which can be deposited at a relatively rapid rate; therefore, particle size and weight determine
the half-life of airborne particulates. Boron and boron-containing compounds in aqueous
environments adsorb onto iron and aluminum hydroxy compounds and clay minerals, and is pH-
dependent, with basic conditions favoring the sorption. Borate and boric acid equilibria in water
are pH-dependent, with borate predominating at higher pH (>9.3); therefore, pH determines
which boron-containing species is available. Boron adsorbs onto particulates in the water and
soil that are high in amorphous aluminum oxide, iron oxide, clay, and to a lesser extent, organic
matter. These interactions are pH-driven as well and adsorption of boron is greatest at basic
conditions (pH 7-9); this is based on boron's need for electron rich environments to form
covalently bonds.
Proposal Draft
Boron — September, 2006 3-6
-------
4.0 EXPOSURE FROM DRINKING WATER
4.1 Introduction
EPA used data from several sources to evaluate the potential for occurrence of boron in
Public Water Systems (PWS) and exposure to boron through drinking water. The primary
source for the drinking water occurrence data is the National Inorganic and Radionuclide Survey
(MRS). In addition to this primary source, the Agency evaluated supplemental sources of
occurrence information, including United State Geological Survey groundwater and surface
water data, American Water Works Association Research Foundation data, and published
literature.
4.2 Ambient Occurrence
4.2.1 Data Sources and Methods
The U.S. Geological Survey (USGS) instituted the National Water Quality Assessment
(NAWQA) program in 1991 to examine ambient water quality status and trends in the United
States. Between 1991 and 2001 the program study units included aquifers and watersheds
covering source water areas for more than 60% of the nation's drinking water and water used for
agriculture and industry. NAWQA monitors the occurrence of contaminants, e.g., pesticides,
nutrients, volatile organic compounds (VOCs), trace elements, and radionuclides, as well as the
condition of aquatic ecosystems (Hamilton et al., 2004). However, no national NAWQA data
are available on the occurrence of boron in ambient waters.
4.2.2 Results
Boron was among the analytes in USGS ground water monitoring in the Sacramento
Valley in California in 1996 (Dawson, 2001) and the lower Illinois River Basin from 1984 to
1991 (Warner, 1999). Boron also was an analyte in NAWQA studies of bed sediments and/or
fish tissues from the Tualatin River Basin of Oregon from 1992 and 1996 (Bonn, 1999), the
Lower Snake River Basin of Idaho and Oregon in 1997 (Clark and Maret, 1998), and the
Yellowstone River Basin in Montana, North Dakota, and Wyoming from 1976 to 1979 (Peterson
andZelt, 1999).
In ground water from the Sacramento Valley aquifer, boron was detected in all thirty-one
samples; concentrations ranged from 12 |ig/L to 1100 |ig/L. The median concentration was 42
|ig/L. Two of the thirty-one samples had concentrations in excess of the early Health Advisory
Level of 600 |ig/L (Dawson, 2001). The lifetime Health Advisory Level changed with the U.S.
EPA revision of the RfD in June 2004.
In ground water from the lower Illinois River Basin, 71% of samples collected between
1984 and 1991 contained boron concentrations higher than the minimum reporting level (50
|ig/L. The highest detected concentration was 2100 |ig/L. Higher boron concentration samples
generally were from deeper aquifers (Warner, 1999).
Proposal Draft
Boron — September, 2006 4-1
-------
In all often fish tissue samples from Oregon's Tualatin River Basin, boron
concentrations exceeded the minimum reporting level of 0.2 |ig/g dry weight. The median
concentration was 1.2 |ig/g and the maximum concentration was 3.5 |ig/g (Bonn, 1999).
In most or all of twenty-five fish tissue samples from the Lower Snake River Basin,
boron concentrations exceeded the minimum reporting level; the highest reported concentration
in this study was 1.8 |ig/g (a minimum reporting level of 0.1 |ig/g dry weight; Clark and Maret,
1998).
In bed sediment samples from the Yellowstone River Basin, boron was detected in 98%
of samples, with a median concentration of 28 mg/kg; the 95th percentile concentration was 57
mg/kg (reporting limit of 10 mg/kg; Peterson and Zelt, 1999).
4.3 Drinking Water Occurrence
4.3.1 Data Sources and Methods
In the mid-1980s, the U.S. EPA conducted the National Inorganic and Radionuclide
Survey (NIRS) to collect national occurrence data on a select set of radionuclides and inorganic
chemicals being considered for National Primary Drinking Water Regulation. NIRS analytes
included 26 unregulated inorganic compounds (lOCs) and 4 unregulated radionuclides, as well
as 10 regulated lOCs and 2 regulated radionuclides.
NIRS collected contaminant occurrence data from 989 public water systems (PWSs)
served by ground water. NIRS did not include surface water systems. The statistical selection of
PWSs was designed to be geographically representative of national occurrence in ground water.
NIRS data were collected from PWSs in 49 states. Data were not available for the state of
Hawaii. In addition, sampling of PWSs was designed so that the stratification of different sized
water systems used in the study represented as best as possible the stratification of the nation's
ground water systems. Consequently, within the study the proportion of any particular size of
PWSs to the total number PWSs in the study was comparable to the proportion of all PWSs of
corresponding size relative to all PWS nationally, e.g., 92% of NIRS PWSs serve small or very
small populations (less than 3,300 persons) and only 2.5% of NIRS PWSs serve populations
greater than 10,000 (65 FR 21576).
Each PWS included in the survey was sampled once between 1984 and 1986. Uniform
detection limits were employed; therefore, NIRS data can be used directly for national
contaminant occurrence analyses with very few, if any, data quality, completeness, or
representativeness issues. There has not been a comparable national survey of inorganic
chemicals and radionuclides since NIRS (65 FR 21576).
Because NIRS did include surface water systems, EPA consulted another study as well, a
boron survey funded by the American Water Works Research Foundation (Frey et al., 2004).
The AWWARF study recruited 189 PWSs representing 407 source waters in 41 States. Of the
407 source water sample kits distributed in 2003, approximately 342 were returned. Of these 342
Proposal Draft
Boron — September, 2006 4-2
-------
samples, 341 were analyzed for boron. Approximately 67 percent (or 228) represented ground
water sources and 33 percent (or 113) represented surface water sources.
4.3.2 Derivation of the Health Reference Level
To evaluate the systems and populations exposed to boron through drinking water from
PWSs, the monitoring data were analyzed against the Minimum Reporting Level (MRL) and a
benchmark value for health that is termed the Health Reference Level (HRL). Two different
approaches were used to derive the HRL, one for chemicals that cause cancer and exhibit a linear
response to dose and the other applies to noncarcinogens and carcinogens evaluated using a
nonlinear approach.
For those contaminants considered to be likely or probable human carcinogens, EPA
evaluated data on the mode of action of the chemical to determine the method of low dose
extrapolation. When the mode of action analysis indicates that a linear low dose extrapolation is
needed, or when data on the mode of action are lacking, a default low dose linear extrapolation
was used to calculate the risk-specific dose equivalent to a one cancer in a million (10"6) risk.
The risk-specific dose was combined with adult body weight and drinking water consumption
data to estimate the drinking water concentration equivalent to a one-in-a-million (10"6) cancer
risk and this value was used as the HRL for likely or probable carcinogens.
For those chemicals not considered to be carcinogenic to humans, EPA generally
calculates a reference dose (RfD). An RfD is an estimate (with uncertainty spanning perhaps an
order of magnitude) of a daily oral exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.
The RfD can be derived from either a "no observed adverse effect level" (NOAEL), a "lowest
observed adverse effect level" (LOAEL), or a benchmark dose, with uncertainty factors applied
to reflect limitations of the data used. EPA derived the HRLs for noncarcinogens using the RfD
approach as follows:
HRL = [(RfD x BW)/DWI] x RSC
Where:
RfD = Reference Dose
BW = Body Weight for an adult, assumed to be 70 kilograms (kg)
DWI = Drinking Water Intake, assumed to be 2 L/day (90th percentile)
RSC = Relative Source Contribution, or the level of exposure believed to result from
drinking water when compared to other sources (e.g., food, ambient air). In all
cases a 20 percent RSC is used for HRL derivation because it is the lowest and
most conservative RSC used in the derivation of an MCLG for drinking water.
The EPA RfD for boron is 0.2 mg/kg/day (U.S. EPA, 2004d) based on developmental
effects in rats from two studies (Price et al., 1996a; Heindel et al., 1992). The RfD was derived
using the benchmark dose (BMD) method (BMDL05 from Allen et al., 1996). EPA established
the HRL for boron using the RfD of 0.2 mg/kg-day and a 20 percent relative source contribution.
Proposal Draft
Boron — September, 2006 4-3
-------
The HRL is calculated to be 1.4 mg/L or 1,400 |ig/L. Further discussion of the RfD derivation
may be found in Section 8.
4.3.3 Results
Nationally, approximately 81.9% of groundwater PWSs had detections of boron
(>minimum reporting level, >MRL or >0.005 mg/L [Table 4-1]). Therefore, about 88.1% of the
population served by the surveyed groundwater PWSs is exposed to boron in drinking water; this
population is equivalent to approximately 75.5 million people. Detections at a concentration
greater than one-half the health reference level (>/^HRL or >0.7 mg/L) occurred in 4.3% of
surveyed groundwater PWSs, indicating that 2.9% of the population served, equivalent to
approximately 2.5 million people, are exposed to this level of boron. Concentrations greater
than the HRL (>HRL or >1.4 mg/L) were found in approximately 1.7% of surveyed groundwater
PWSs, indicating that exposure at this level occurs in 0.4% of the population served, equivalent
to approximately 0.4 million people.
In the AWWARF study, samples were analyzed for boron with a method detection limit
of 0.002 mg/L, or 2.0 Fg/L (Frey et al., 2004). Boron was detected with concentrations equal or
greater than the method detection limit in 226 of 228 ground water samples (99.1%) and 110 of
113 surface water samples (97.3%). Boron concentrations greater than /fflRL or >0.7 mg/L
were found in 20 of 228 ground water samples (8.8%) and no surface water samples (0%).
Boron concentrations greater than the HRL or > 1.4 mg/L were found in 7 of 228 ground water
samples (3.1%) and no surface water samples (0%). The highest concentration detected in
ground water was approximately 3.32 Fg/L (Seidel, 2006). The median concentrations were
0.0514 mg/L in ground water and 0.029 mg/L in surface water (Frey et al., 2004). Although the
survey was not statistically representative, it indicates some general trends. On the whole, boron
contamination of surface water is less significant than contamination of ground water. No
geographic trends were evident in ground water results, but surface water contamination
appeared to be more prevalent in the western U.S. than the eastern U.S. Longitudinal sampling at
15 systems revealed that a wide variety of treatment techniques were largely ineffective at
removing boron, so boron concentrations in source water (such as those collected in this study)
are likely to be indicative of concentrations in finished water (Frey et al., 2004).
4.4 Summary
The limited data used in this report suggests boron could be ubiquitous in the
environment, including ground water, fish tissues, and stream bed sediments. The Reference
Dose (RfD) for boron is 0.2 mg/kg/day and the Health Reference Level (HRL) based on the RfD
was determined to be 1.4 mg/L. According to the U.S. EPA's National Inorganic and
Radionuclide Survey (MRS), approximately 81.9% of groundwater PWSs had detections of
boron (>minimum reporting level, >MRL, or >0.005 mg/L). These detections affected about
88.1% of the population served by the PWSs, equivalent to approximately 75.5 million people
served by ground water nationally. Detections at a concentration greater then one-half the health
reference level (>/^HRL or >0.7 mg/L) occurred in 4.3% of surveyed PWSs, affecting 2.9% of
the population served and equivalent to approximately 2.5 million people nationally.
Concentrations greater than the HRL (>HRL or > 1.4 mg/L) were found in approximately 1.7%
Proposal Draft
Boron — September, 2006 4-4
-------
of surveyed PWSs, affecting 0.4% of the population served and equivalent to approximately 0.4
million people nationally.
Supplementary data from an AWWARF-sponsored study indicate that boron
contamination of surface water is less significant than boron contamination of ground water.
Table 4-1 Summary Occurrence Statistics for Boron in Ground Water Systems
Frequency Factors
Total Number of Samples/Systems
99* Percentile Concentration (all samples)
Health Reference Level (HRL)
Minimum Reporting Level (MRL)
99* Percentile Concentration of Detections
Median Concentration of Detections
Total Population Served
Occurrence by Sample/System
Ground Water PWSs with Detections (^MRL)
Range ofNIRS States
Ground Water PWSs > 1/2 HRL
Range ofNIRS States
Ground Water PWSs > HRL
Range ofNIRS States
Occurrence by Population Served
Population Served by GW PWSs with Detections
Range ofNIRS States
Population Served by GW PWSs > 1/2 HRL
Range ofNIRS States
Population Served by GW PWSs > HRL
Range ofNIRS States
NIRS Data on
Boron
989
2.44 mg/L
1.4 mg/L
0.005 mg/L
2.6 mg/L
0.047 mg/L
1,482,153
Systems/
Population %
81.9%
0 - 100%
4.3%
0 - 37%
1.7%
0 - 26%
88.1%
0 - 100%
2.9%
0 - 34%
0.4%
0 - 34%
National System
& Population
Numbers1
59,440
--
--
--
--
--
85,681,696
National
Extrapolation
48,682
N/A
2,584
N/A
1,022
N/A
75,501,000
N/A
2,469,000
N/A
372,000
N/A
1. Total PWS and population numbers are from U.S. EPA (2000e), Water Industry Baseline Handbook, 2nd Edition. National
extrapolations are generated by multiplying the system/population percentages and the national Baseline Handbook
system/population numbers.
Abbreviations:
PWS = Public Water Systems; GW = Ground Water; N/A = Not Applicable; Total Number of Samples/Systems = total number
of samples/systems on record for the contaminant; 99th Percentile Concentration = the concentration in the 99th percentile
sample (out of either all samples or just samples with detections); Median Concentration of Detections = the concentration in
the median sample (out of samples with detections); Total Population Served = the total population served by PWSs for which
sampling results are available; Ground Water PWSs with Detections, PWSs >1/2HRL, and PWSs >HRL = percentages of GW
PWSs with at least one sampling result greater than or equal to the MRL, exceeding the '/JHIRL benchmark, or exceeding the
HRL benchmark; Population Served by GW PWSs with Detections, by PWSs >1/2HRL, and by PWSs >HRL = percentages of
the population served by GW PWSs with at least one sampling result greater than or equal to the MRL, exceeding the 'AHRL
benchmark, or exceeding the HRL benchmark.
Notes: Only results at or above the MRL were reported as detections. Concentrations below the MRL are considered nondetects.
The HRL used in this analysis is a draft value for working review only.
Proposal Draft
Boron — September, 2006
4-5
-------
5.0 EXPOSURE FROM MEDIA OTHER THAN WATER
5.1 Exposure from Food
5.1.1 Concentration in Non-Fish Food Items
Levels of boron in food products are related to boron in the soils where they are grown
and, accordingly, show some geographic fluctuations. Product categories having high levels
have been identified as tubers, legumes, fruits and fruit-based beverages (IOM, 2001). In one
dietary study, coffee, milk, apples, dried beans and potatoes accounted for 27 percent of the
boron in the diet (Rainey et al., 1999). In the 1994 Total Diet Study from the United Kingdom,
the food groupings with the highest boron concentrations were nuts (14 mg/kg fresh weight),
fruits and fruit products (2.4-3.4 mg/kg), green vegetables (2.0 mg/kg), and potatoes and other
vegetables (1.2-1.4 mg/kg). The levels were below 1 mg/kg for other food categories (Ysart et
al., 1999). Most foods contain less than 6 mg boron/kg of food. Some individual foods may
contain more than 20 mg B/kg of food (Seller et al., 1988).
5.1.2 Concentrations in Fish and Shellfish
The data on the presence of boron in fish and shellfish are very limited. The average
concentration measured in fish in the United Kingdom 1994 Total Diet Study was 0.5 mg/kg
fresh weight (Ysart et al., 1999). Boron has been detected in shrimp by inductively coupled
plasma spectroscopy (Mann, 1988 ).
As mentioned in Section 4.2.1, the presence of boron in fish tissues was measured by the
USGS in several surveys. In ten fish tissue samples from Oregon's Tualatin River Basin the
median concentration was 1.2 |ig/g and the maximum concentration was 3.5 |ig/g (Bonn, 1999).
In fish tissue samples from the Lower Snake River Basin, the highest reported boron
concentration was 1.8 |ig/g (a minimum reporting level of 0.1 |ig/g dry weight; Clark and Maret,
1998).
5.1.3 Intake of Boron from Food
Dietary intake data from the Continuing Survey of Food Intakes (CSFII) during 1994-
1996 (IOM, 2001) are displayed in Table 5-1. Average values for adults range from 0.87 to 1.34
and 90 percentile intakes are about 1.5 to 2 mg/day. Findings from the NHANES III survey
(1988-1994) are similar (IOM, 2001).
5.2 Exposure from Air
Bertine and Goldberg (1971) estimated that approximately 11,600 tons of boron are
injected into the atmosphere as a component of fly ash produced by coal combustion; the fly ash
was estimated to contain an average of about 75 mg/kg boron. There are insufficient data to
estimate the intake of boron from ambient air.
Proposal Draft
Boron — September, 2006 5-1
-------
Table 5-1 Mean Intake of Boron (mg/day) from Food Based on the Continuing Survey
of Food Intake by Individuals 1994-1996 (IOM, 2001)
Age
N
mean
standard error
Both sexes
0-6 mo
7-12 mo
1-3 yr
4-8 yr
195
130
1834
1650
0.75
0.99
0.86
0.80
0.14
0.12
0.02
0.01
Males
9-13 yr
14-18 yr
19-30 yr
31-50yr
51-70yr
71+yr
552
446
853
1684
1606
674
0.90
.02
.15
.33
.34
.25
0.03
0.04
0.03
0.03
0.02
0.03
Females
9-13 yr
14-18 yr
19-30 yr
31-50yr
51-70yr
71+yr
Pregnant
Lactating
All excluding P&La woman
All including P&L women
560
436
760
1614
1539
623
70
41
15,156
15,267
0.83
0.78
0.87
.00
.11
0.98
.16
.39
.06
.06
0.03
0.01
0.03
0.02
0.02
0.03
0.09
0.16
0.01
0.01
Adapted from Rainey et al., 1999
a. P&L, pregnant and lactating
5.3 Exposure from Soil
5.3.1 Concentration of Boron in Soil
Boron occurs in the earth's crust at a concentration of about 0.001%, generally as
compounds, and rarely as a pure element (O'Neil et al., 2001). Widely distributed boron
compounds include borax, kernite, and tourmaline, the three most commonly mined boron
minerals (Seller et al., 1988). High levels of boron occur predominantly in soil originating from
marine sediments and arid regions (Brown et al., 1983). Boric acid naturally occurs as the
mineral sassolite (O'Neil et al., 2001). Sodium tetraborate, Na2B4O7, usually occurs as a
decahydrate mineral known as borax and is found largely in ancient dry lake beds from the
tertiary period (Clayton and Clayton, 1994). One report indicated that the average concentration
of boron in soil is 10 mg/kg (Weast, 1988-1989). Another report indicated a geometric mean
background concentration of 26 mg/kg, with a maximum concentration of 300 mg/kg, for boron
in U.S. soils (Eckel and Langley, 1988). Boron was detected in soils in Idaho at geometric mean
concentrations of 4.6-9.8 mg/kg (Rope et al., 1988). Malins et al. (1984) reported on boron in
sediments of Puget Sound.
Proposal Draft
Boron — September, 2006
5-2
-------
5.3.2 Intake of Boron from Soil
Human exposure to contaminants in soils is usually from dust that infiltrates homes and
automobiles, and incidental soil ingestion. Estimates of soil intake often assume an ingestion
rate of 100 mg/day for children and 50 mg/day for adults (U.S. EPA, 1997a). Using the average
concentration of boron in soil from Weast (1988-1989), 10 mg boron/kg soil, and the assumption
that infants and adults ingest 0.0001 and 0.00005 kg/soil per day (100 mg and 50 mg),
respectively, exposure of children to boron from soils would be about 1.0 |_ig/day and 0.5 |_ig/day
for adults.
10 mg/kg soil x 0.0001 kg soil (children) = 0.001 mg/day (1.0 \ig)
10 mg/kg soil x 0.00005 kg soil (adults) = 0.0005 mg/day (0.05 [ig)
5.4 Other Residential Exposures
Some human exposures to borates are linked to insecticide use. Typically, borate-based
insecticides are powders or dust used to control cockroaches. Children who, relative to adults,
have greater hand-to-mouth contact and exposure to floor boards, where the insecticides usually
are applied, are more likely to ingest them. Medicinals and personal care products containing
boron may be absorbed through mucous membranes and/or damaged skin. Populations living in
areas of California and other western states with boron-rich mineral deposits potentially have
high exposure to boron from drinking water and locally grown foods (Butterwick et al., 1989).
5.5 Occupational (Workplace) Exposures
5.5.1 Description of Industries and Workplaces
Industries and workplaces where boron compounds are found in abundance include
borate mines and processing plants. Manufacture of fiberglass and other glass products, cleaning
and laundry products, fertilizers, pesticides, and cosmetics constitute industries where boron
compounds can commonly be found in the workplace (U.S. Borax and Chemical Corporation,
1991).
5.5.2 Types of Exposure (Inhalation, Dermal, Other)
Exposure of boron in the workplace is expected to be mainly through inhalation and
dermal contact.
5.5.3 Concentrations of Boron in the Work Environment
Boron in its various forms is classified under the "nuisance" category (Clayton and
Clayton, 1994). Reported concentrations of borax dust in different areas of a large borax mining
and refining facility ranged from 1.1-14.6 mg/m3 (Garabrant et al., 1985); the mean boric
acid/boron oxide dust concentration in one boric acid manufacturing plant was 4.1 mg/m3
(Garabrant et al., 1984).
5.6 Summary
Proposal Draft
Boron — September, 2006 5-3
-------
The boron exposure for the general population is mostly through the ingestion of food
and, to a lesser extent, water. Populations with the greatest risk of exposure are those from
boron-rich regions of the western United States, especially California, children having frequent
hand-to-mouth contact and greater exposure to floor boards where the insecticides containing
boron usually are applied, and workers in industries that use boron.
Proposal Draft
Boron — September, 2006 5-4
-------
6.0 TOXICOKINETICS
6.1 Absorption
Oral Exposure
Studies in both humans and animals show that boron is well absorbed from the
gastrointestinal tract. Schou et al. (1984) administered approximately 131 mg B as boric acid in
both water (750 mg) and water-emulsifying ointment (740-1473 mg, approximately 130-258 mg
B) to six volunteers and found that an average of 92-94% of administered boron was excreted in
the urine within 96 hours, indicating that at least that much had been absorbed in that time.
Although there was no significant difference in cumulative excretion for the two different
vehicles, it was noted that excretion in the first 2-hour sampling period was lower after exposure
to the ointment, suggesting delayed absorption of boron from the ointment in comparison to the
water vehicle. Similarly, the two women who ingested approximately 62 mg B as boric acid (in
addition to 80-140 mg of boron in food) excreted greater than 90% of ingested boron in the urine
in the first week after dosing (Kent and McCance, 1941). Volunteers (n=10) who drank spa
waters containing approximately 100 mg daily dose of boron for 2 weeks had over 90%
absorption of boron based on urinary excretion data (Job, 1973). Naghii and Samman (1997)
studied the effect of boron supplementation (10 mg B/day) into the normal diet of male
volunteers (n=8). Supplementation of the 10 mg B/day for 4 weeks resulted in 84% recovery in
the urine.
Studies in animals have shown that boron is readily absorbed following oral exposure in
rats (Ku et al., 1991; Usuda et al., 1998), rabbits (Draize and Kelley, 1959), sheep (Brown et al.,
1989) and cattle (Owen, 1944; Weeth et al., 1981). Using mass spectrometry and the boron-10
isotope, Vanderpool et al. (1994) showed that fasted rats fed 20 ug of 10B in the diet eliminated
95% of the 10B in the urine and 4% in the feces within 3 days of dosing, producing a 77%
increase in the ratio of 10B to UB in the urine. Moreover, 10B in the liver peaked within 3 hours
of dosing with over 90% recovery and a 56% increase in 10B:nB ratio, which returned to normal
within 24 hours. This result suggests that >90% of orally administered boron is absorbed from
the gastrointestinal tract within 3 hours and that absorption is complete within 24 hours.
Dermal Exposure
Human and animal studies show that boron is not absorbed across intact skin. However,
there is evidence that boron can be absorbed through more severely damaged skin, especially
from an aqueous vehicle. Draize and Kelley (1959) found no increase in urinary boron in a
volunteer given topical application of powdered boric acid (15 g) to the forearm and held under
occlusion for 4 hours. Friis-Hansen et al. (1982) reported no evidence of boron absorption in 22
newborn infants treated dermally with ointment containing 3% boric acid for 4-5 days (total dose
of approximately 16 mg B); plasma boron levels fell over the 5-day study period, as expected for
neonates, and did not differ from 10 untreated controls. Vignec and Ellis (1954) found minimal
difference in blood or urinary boron levels in twelve 1- to 10-month-old infants exposed to
talcum powder containing 5% boric acid 7-10 times per day for at least 1 month (estimated daily
dose of 2.33 g boric acid or 407 mg B) compared with an equal number of untreated controls.
An additional group of 12 infants with mild to moderate diaper rash during the test period was
continued on the powder regimen for 48-72 hours after rashes appeared. Their boron blood
levels were similar to controls. However, blood and urinary boron levels were increased in six
Proposal Draft
Boron — September, 2006 6-1
-------
male volunteers with severe skin conditions (e.g., psoriasis, eczema, urticaria) following topical
application of an aqueous jelly containing 3% boric acid (Stuttgen et al., 1982). Urinary boron
levels did not increase in skin-damaged volunteers given 3% boric acid in an emulsifying
ointment.
Studies in laboratory animals have produced similar results. Boron was not absorbed
across intact or mildly abraded skin in rabbits topically administered boric acid as the undiluted
powder or at 5% in talc or aqueous solution (1.5 hr/day under occlusion for 4 days; 10-15% of
body surface exposed) (Draize and Kelley, 1959). However, boron was readily absorbed across
severely damaged skin in rabbits in proportion to the exposure concentration. Rats with intact
skin treated topically with 3% boric acid (ointment or aqueous jelly) did not absorb boron, but
urinary boron was increased 4- to 8-fold (to 1% of dose) following exposure to boric acid
oleaginous ointment and 34-fold (to 23% of dose) following exposure to aqueous boric acid in
rats with damaged skin (Nielsen, 1970).
Inhalation Exposure
Boron is absorbed during inhalation exposure. Culver et al. (1994) monitored boron
levels in the blood and urine of male workers exposed to borate dust (borax, borax pentahydrate
and anhydrous borax) at a borax production facility. The workers were divided into three groups
according to borate exposure. Workers in both the medium- and high-exposure categories had
significantly increased levels of boron in the blood after working Monday (about 0.25 |ig/g) in
comparison to pre-shift Monday morning values (about 0.1 |ig/g). Similarly, workers in the high
exposure category had significantly higher urinary boron levels Monday post-shift (about 12
|ig/mg creatinine) than pre-shift (about 2 |ig/mg creatinine). Boron in the diets (which were
assigned by the researchers to ensure uniformity among workers) and workplace air also was
monitored during this study. A higher proportion of total boron intake was from air than from
diet, and both blood and urine boron were best modeled based on air concentration of boron
alone (i.e., inclusion of dietary boron as an independent variable did not increase the predictive
power of the models). These data show that boron was absorbed during the work day, and that
borate dust in the air was the source of the additional boron in the blood and urine. However, it
is not clear what amount of the inhaled boron was actually absorbed through the respiratory tract.
The researchers speculated that due to the large size of the dust particles in the work area, most
of the inhaled borate would have been deposited in the upper respiratory tract, where it could
have been absorbed directly through the mucous membranes or could have been cleared by
mucociliary activity and swallowed.
Similar evidence of absorption of airborne boron in rats was obtained by Wilding et al.
(1959), who monitored urinary boron levels in rats exposed to aerosols of boron oxide (average
concentration of 77 mg/m3). Urinary boron was much higher in exposed rats than controls
throughout the 22-week exposure period (average of 11.90 vs. 0.24 mg B/kg-day) and quickly
reverted to control levels following cessation of exposure. These data show that inhalation
exposure to boron oxide particulate produced high levels of urinary boron, but do not rule out a
contribution by gastrointestinal absorption of particles transported from the upper respiratory
tract by mucociliary activity. No toxic effects were observed.
Proposal Draft
Boron — September, 2006 6-2
-------
6.2 Distribution
Studies suggest that boric acid and borate compounds in the body exist primarily as
undissociated boric acid, which distributes evenly throughout the soft tissues of the body, but
shows some accumulation in bone. Ku et al. (1991) studied tissue distribution in male rats fed
9000 ppm of boric acid (1575 ppm boron) for 7 days. The authors estimated the 9000 ppm dose
to be 93-96 mg B/kg-day. The tissue levels of boron on day 7 of exposure are listed in Table
6-1. Boron levels in all tissues except adipose increased rapidly after the start of exposure (2- to
20-fold increase over controls after 1 day). The greatest increase (20-fold) was in bone. Levels
in adipose tissue increased only 1.3-fold above controls. Boron levels in plasma and soft tissues
other than adipose tissue reached steady-state (12-30 |ig/g) within 3-4 days. Variability in levels
of boron among soft tissues (adipose and kidney excluded) was minimal, with tissue
concentrations at 60% of plasma levels on day 1 and 30-40% of plasma levels on days 2, 3, 4,
and 7. Levels in bone and adipose continued to increase throughout the 7-day study period. In
comparison to plasma levels, there was no appreciable accumulation of boron in any soft tissue.
However, boron did accumulate in bone, showing a 2- to 3-fold increase over plasma levels after
7 days. Boron levels in adipose tissue remained at 20% of plasma levels after 7 days. Other
investigators provided support for these findings: (1) accumulation of boron in bone in rats
(Forbes and Mitchell, 1957); (2) lack of appreciable accumulation of boron in the testis (Lee et
al.,1978; Treinen and Chapin, 1991); and (3) lack of appreciable accumulation of boron in the
epididymis (Treinen and Chapin, 1991).
In a follow-up to Ku et al. (1991), Chapin et al. (1997) monitored bone boron
concentrations in rats fed 200-9000 ppm of boric acid for 9-12 weeks. Bone boron was
significantly increased over controls at 200 ppm and increased proportionally up to 6000 ppm,
above which the increase in bone was slightly less than the increase in the feed. Bone boron
levels reached steady state within 1 week at doses up to 3000 ppm and after approximately 4
weeks at higher doses. Steady-state bone boron levels were approximately 4-fold greater than
serum boron levels. Chapin et al. (1997) also monitored bone (tibia) boron levels for 32 weeks
following cessation of exposure in rats that had been fed boron in the diet for 9 weeks. Levels of
boron in the bone declined slowly. After 8 weeks of recovery, bone levels of boron were
reduced to roughly 10% of levels at the end of exposure (e.g., at 9000 ppm: about 6 ug B/g bone
from about 60 ug B/g bone) but still remained 5- to 6-fold higher than bone levels in unexposed
controls (about 1 ug B/g bone). Even after 32 weeks of recovery (and about 31.5 weeks after the
return of blood boron levels to normal, which took only 4 days), bone boron concentrations
remained 3-fold higher in treated groups than bone concentrations in controls.
In a drinking water study using multiple dose levels of boric acid in rats, Naghii and
Samman (1996) found, like Ku et al. (1991), that levels of boron in soft tissues were very similar
to levels in plasma (the only exception being a 1.5- to 2-fold increase in the kidney that may
have been due to contamination with urine because the organ was not perfused prior to analysis).
Proposal Draft
Boron — September, 2006 6-3
-------
Table 6-1 Tissue Levels of Boron in Male Rats on Day 7 of Exposure to 9000 ppm Boric
Acid (1575 ppm boron) in the Diet (ug boron/g tissue)
Tissue
Plasma
Liver
Kidney
Adipose
Muscle
Bone
Large intestine"
Brain
Hypothalamus
Testes
Epididymisa
Seminal vesicles"
Seminal vesicle fluidb
Adrenalsb
Prostateb
Control
1.94±0.17
0.66±0.10
1.55±0.03
1.71±0.17
3.69±0.54
1.17±0.19
3.08±0.17
0.76±0.02
0.91
0.97±0.10
0.81±0.15
1.64±0.23
2.05
7.99
1.20
Day 7
16.00±0.71
13.13±0.54
19.80±1.65
3.78±0.13
14.23±0.19
47.40±1.14
14.90±0.7
13.50±0.86
14.30
16.00±1.19
16.81±3.7
23.70±6.56
19.20
21.90
14.80
Source: Kuetal. (1991)
Note: Values are means +/- SE: N = 3 animals unless indicated by footnote
a Mean +/- SE. N = 3 samples, each sample represents a pool of tissue from two animals
b A single sample was analyzed representing a pool from six animals
Proposal Draft
Boron — September, 2006
6-4
-------
After 3 and 6 weeks of exposure to boric acid in drinking water at doses of 0, 2, 12.5, and
25 mg/rat/day, solid tissues (kidney excluded) demonstrated boron contents which varied less
than 25% within any given dose time group. In boric acid-exposed rats, maximally observed
differences in boron concentrations between plasma and solid tissues (kidney excluded) were
less than 28%, while most differences noted were less than 10% at any dose or time. The
researchers also found that boron plasma and tissue levels increased proportionally with dose.
Bone was not analyzed in this study. WHO (1998a) reported a preliminary comparison of blood
boron levels across species in rats exposed to boron in the diet or drinking water and humans
exposed in the diet, drinking water, or accidental ingestion. Rat and human blood boron levels
had a good overlap in the dose range of 0.01-100 mg B/kg body weight. Locksley and Sweet
(1954) found that concentration of boron in the tissues was directly proportional to dose over a
range of 1.8-71 mg B/kg in mice given borax by intraperitoneal (ip) injection.
Magour et al. (1982) examined the levels of distribution of boron in blood and tissues of
3-week- and 3-month-old female Wistar rats administered one time intraperitoneally with 42 mg
B/kg as sodium borate. Boron levels in kidney, brain, liver, heart, and blood of 3-week-old rats
were examined, and demonstrated peak concentrations at 30 minutes following intraperitoneal
injection (brain excluded). Concentrations in blood, liver, and heart differed by approximately
30% at 30 minutes, and declined in parallel fashion, with concentration differences among
tissues diminishing out to 4 hours post-administration. Boron tissue concentration-time profiles
were somewhat different when observed in 3-month-old rats. In contrast to the younger rats,
blood boron concentrations continued to rise to 1 hour post-administration, and brain
concentrations were maximal at 30 minutes post-administration. Boron concentrations in blood,
liver, and heart reached concentrations which differed by approximately 10% at 3 hours
post-administration and remained similar at 4 hours post-administration. Concentration decay
profiles of boron in kidney, heart and liver appeared parallel 1 to 4 hours post-administration,
with concentrations in kidney being approximately 70% higher than those in blood, liver, and
heart. Similar to findings in 3-week-old rats, the highest concentrations were attained in kidney,
and maximal concentrations in tissues other than blood were reached at 30 minutes following
injection. In another experiment, 3-week-old rats received 20 mg B/kg in their drinking water
for 21 days. Boron levels in the kidney, liver, and brain increased steadily during the first 9 days
of treatment and returned to control levels 7 days following cessation of exposure. Blood boron
levels continued to rise up to day 21 of treatment while levels in the liver and brain returned
rapidly to control levels during that time frame. The authors stated that the data suggest the
development of a hemostatic mechanism which eliminates any excess of boron from liver and
brain against its own concentration gradient because the concentration in the blood was
significantly higher than in the liver and brain between days 13 and 21. The authors also state
that boron will be completely eliminated if the animals consume drinking water without added
boron from days 21-28 which suggests boron is not firmly bound to any tissue components.
Data concerning the distribution of boron in humans is more limited than in experimental
animals. Evidence that boron does not accumulate in the blood in humans was obtained by
Culver et al. (1994). These researchers found no progressive accumulation of boron across the
work week as measured by blood and urine levels in mine workers. Accumulation of boron in
skeletal bones of human cadavers has also been reported by Alexander et al. (1951) and Forbes
etal. (1954).
Proposal Draft
Boron — September, 2006 6-5
-------
6.3 Metabolism
Overview of Metabolic Pathways
Boron is a trace element for which essentiality is suspected but has not been directly
proven in humans (Nielsen, 1991, 1992, 1994; NRC, 1989; Hunt, 1994; Mertz, 1993; Devirian
and Volpe, 2003). Boron deprivation studies with animals and three human clinical studies have
shown that boron affects macromineral and cellular metabolism of other substances that affect
life processes such as calcium and magnesium.
Inorganic borate compounds are present as boric acid in the body. Boric acid is the only
boron compound that has been identified in urine, and it has repeatedly been found to account
for >90% of the ingested boron dose (WHO, 1998a). There is no evidence that boric acid is
degraded in the body. Metabolism may not be feasible because a large amount of energy (523
kJ/Mol) is apparently required to break the boron-oxygen bond (Emsley, 1989). Boric acid can
form complexes with various biomolecules (IEHR, 1997; WHO, 1998a). It has an affinity for
hydroxyl, amino, and thiol groups. Complex formation is concentration dependent and
reversible.
6.4 Excretion
The elimination and excretion of boron have been evaluated in humans and rodents, in
oral studies only. No studies were summarized that addressed excretion after dermal or
inhalation exposures (U.S. EPA, 2004a).
Studies have demonstrated that more than 90% of an orally administered dose of boric
acid is excreted unchanged in the urine a short time after treatment (Section 6.1 under oral
exposure). In humans, Jansen et al. (1984a) and Schou et al. (1984) reported that boron's
primary route of elimination was in the urine. Jansen et al. (1984b) reported that approximately
60-75% of a dose of 750 mg boric acid (131 mg B) in a water solution or 740-1473 mg boric
acid (129.5-261.3 mg B) in a water emulsifying ointment administered orally to humans is
eliminated in urine over the initial 24 hours, with the urinary route of elimination accounting for
93% of the dose at 96 hours after oral administration. Graphically, Jansen et al. (1984b)
demonstrated cumulative boron elimination, as percentage of dose, from six adult males who
consumed an aqueous solution of boric acid. Results indicate that at 12 hours, the urinary
elimination accounted for 52.7 ± 4.9% (mean ±S.D.) of the dose (range 46.4-58.9%); at 24
hours, the cumulative urinary elimination accounted for 66.9 ± 6.4% of the dose (mean ± S.D.),
with a range of 57.1-75.0%. These data demonstrate a marked similarity among this limited
sample of adult men in the renal elimination of boric acid. In a clinical report of an acute,
uncontrolled intoxication with boric acid, Astier et al. (1988) estimated the dose as 45 g boric
acid (7.9 g B), and reported that renal elimination accounted for 50% of the dose in the first 24
hours. Regression analysis of plasma B concentrations revealed a clearance of 0.77 L/hour.
While no methods of analysis were presented, the authors concluded that tubular reabsorption
affected 80% of the dose. Kent and McCance (1941) also reported that 92-93% of an
administered oral dose (352 mg as boric acid) in humans was eliminated in urine during the first
week following administration. Additional minor elimination pathways include saliva, sweat,
and feces (Jansen et al., 1984a).
Proposal Draft
Boron — September, 2006 6-6
-------
Jansen et al. (1984a) evaluated boron clearance daily in seven adult males exposed
through dietary intake over 3 days and in the same subjects after 20-minute intravenous infusion
of 28.52 mg boric acid (5-5.6 mg B) per minute, or a total dose per subject of 570-620 mg boric
acid (91-108.5 mg B). In the dietary intake phase, urine was collected at 12-hour intervals, and
blood was sampled twice per day to determine basal levels of boron. There were no restrictions
on diet during this period. For the infusion phase, subjects stayed in a metabolic ward for 12
hours after receiving the intravenous dose. Each subject was catheterized with a Venflon
catheter in the right arm for boric acid administration. Another Venflon catheter was placed in
the left arm for blood sampling. Blood samples were drawn at 0, 0.42, 0.67, 2, 4, 6, 8, 10, and 12
hours, for a total of nine blood samples from each subject during the 12-hour period. After
release from the metabolic ward, each subject had a blood sample drawn at 9 a.m. and 4 p.m.
daily for 5 days. Renal clearance was calculated as the total amount of boron excreted per
minute in the urine, divided by the area under the plasma boron concentration-time curve (mg
Burine/AUC-min), normalized to body-surface area.
For the dietary exposure phase, the urinary excretion of boric acid during any 12-hour
period ranged from 1.52 to 18.1 mg, consistent with large variations in dietary intake of boron.
Plasma concentrations during this 72-hour period ranged from <0.10-0.46 mg boric acid/L
(<0.018-0.081 mg B/L). In contrast, following boric acid infusion, plasma boron rose to peak
concentrations 25 minutes after the start of the infusion at 10-20 mg/mL, approximately 100
times the basal concentration. Virtually the entire dose (99%) was eliminated in the urine over
120 hours.
Jansen et al. (1984a) did not calculate boron clearances for dietary exposure but
published the individual data, from which clearances can be calculated using the following
formula (Murray, 2002):
Renal Clearance = Amount of boron excreted/min in urine over 24 hours
Average of same day plasma boron at 9 a.m. and 4 p.m.
The results are shown in Table 6-2, along with the infusion-phase clearances published
by Jansen et al. (1984a). Boron clearance at dietary exposure levels was characterized by a high
coefficient of variation (CV, standard deviation/mean) of 0.78, but the mean value was
remarkably consistent (39-42 mg/min/1.73 m2) for each day of the 3-day baseline measurement
period. Boron clearance following boric acid infusion was 60.5 mL/min/1.73, with a CV of only
0.09 (Table 6-2). The interindividual variability in renal boron clearance was much greater when
clearance was calculated from the subjects receiving exposure to boron in the diet alone
compared to the values calculated in the same individuals receiving a single intravenous
infusion.
Proposal Draft
Boron — September, 2006 6-7
-------
Table 6-2 Renal Boron Clearance (mL/min/1.73m2) Calculated from Dietary Exposure
and Intravenous Infusion
Subject3
1
2
3
5
6
7
8
Mean±S.D.
Boron Clearance (mL/min/1.73m2)
Dietary Boron Exposure Onlyb (mg B/day)
Day 1: 1.79±1.23
47.7
58.3
12.0
83.0
62.8
15.6
16.4
42.3 ±27.8
Day 2: 1.45±0.47
113.4
14.5
20.2
66.8
29.6
15.2
20.8
40.1 ±37.1
Day 3: 1.52±0.44
83.4
42.6
19.6
77.2
17.3
13.3
22.2
39.4±29.5
Intravenous
Infusion0
(mg B/day)
105
55.9
65.8
63.8
62.7
65.0
51.2
58.9
60.5±5.4
Source: Adapted from Jansen et al. (1984a)
a Subject No. 4 was excluded due to increasing excretion in urine during the period
b Dose estimated from total urinary excretion of boron during 24 hours of normal dietary exposure
c Dose administered by 20-minute intravenous infusion
Proposal Draft
Boron — September, 2006
6-8
-------
The variance of dietary-exposure boron clearance was 66 times greater than for
intravenous infusion. The mean boron clearance estimated by this method was lower than the
mean boron clearance estimated from the intravenous infusion by a factor of 1.5. There are a
number of possible reasons for both the higher variability and lower absolute clearance values as
outlined in the following paragraphs.
Any analytical error that overestimated plasma boron would have led to an underestimate
of boron clearance. The detection limit of the spectrophotometric method used by Jansen et al.
(1984a) to determine plasma boron was 0.1 mg/L of boric acid. The precision of the method was
degraded substantially at low boric acid concentrations, with a CV of 0.71 at 0.14 mg/L versus a
CV of 0.055 at 4.93 mg/L. At the plasma boron levels found on the first three days of the study
(0.10-0.46 mg/L), the precision of the analytical method was a potential source of significant
error. In addition, more than 25% of the plasma boron samples measured during the
dietary-exposure phase were below the limit of detection, and were entered as half the limit of
detection in the calculations. If the actual plasma boron concentration was lower (i.e., less than
0.05 mg/L of boric acid), the estimated boron clearance would have been higher. The plasma
boron levels in the intravenous infusion study were orders of magnitude higher, so that analytical
error and detection limit problems were less likely to be factors.
Another factor that would lead to an underestimate of boron clearance in the
dietary-exposure phase would be missed or incomplete urine samples. In the Jansen study, the
subjects did not stay in the clinic for the 3-day dietary-exposure phase. As urine was collected at
12-hour intervals during this phase, urine samples may not have been 100% complete. Because
the subjects remained in the clinic for the first 12 hours of the infusion phase, complete urine
collection was more likely.
Although less likely, biological factors could play a role in the relative magnitude and
variability of boron clearance in the two phases. Some of the variability may have its basis in
interindividual differences in the rate, pattern, and extent of absorption from the gut into the
bloodstream, magnified at low and intermittent dietary exposure levels. Dose-dependent kinetics
could potentially explain the lower renal boron clearance, as the dietary exposure was about two
orders of magnitude lower than the intravenous dose. While this possibility cannot be
completely eliminated, it does not appear to be the most likely explanation. The individual data
on boron clearance and dose (based on urinary excretion of boron/day) does not show a
dose-dependent relationship. Overall, clearance appeared to be independent of dose within the
range studied.
The urinary elimination of boron administered to male rats has been investigated
following the oral administration of sodium tetraborate (at 11 different doses ranging from 0-4
mg B/kg) by Usuda et al. (1998). The recovery of boron in 24-hour urine accounted for 99.6 ±
9.7% of the administered dose, demonstrating essentially total bioavailability of an
orally-administered boron dose in rats. In a study conducted in rats with stable-labeled boron,
Vanderpool et al. (1994) reported that 95% of the administered (20 |ig/kg) dose was eliminated
in the urine and 4% in the feces over the initial 3 days post-dosing.
Urinary elimination of boric acid in Sprague-Dawley female rats (nonpregnant and
pregnant) was examined in a pharmacokinetic study (U.S. Borax, 2000; Vaziri et al., 2001).
Proposal Draft
Boron — September, 2006 6-9
-------
Three groups of 10 nonpregnant and 10-11 pregnant rats were started on an initial 7-day
supplemented boron diet on gestation day 9, prior to gavage administration of boric acid.
According to the authors, the purpose of this initial 7-day diet was to achieve steady state
conditions for rats given a diet comparable to that ingested by humans in terms of boron. This
supplemented boron diet given during the initial 7 days was designed to deliver a dose of
approximately 0.3 mg/kg-day of boric acid or 0.05 mg B/kg-day. On the morning of day 8, the
diet for all rats was switched to the low boron casein diet containing 0.2 mg B/kg diet for a total
of 24 hours. The low boron casein-based diet was used in this study to minimize cross
contamination of the urine with boron in the diet and to minimize the dietary contribution of
boron on the day of gavage. After the initial 24 hours on the low casein diet, groups of pregnant
and nonpregnant rats were given a single oral dose of 0.3, 3.0, or 30 mg/kg of boric acid (0.052,
0.52, and 5.2 mg B/kg, respectively) by gavage in deionized water (ultrapure). According to the
authors, the low dose was chosen as an estimate of the high end human dietary dose level, and
the highest dose tested was approximately half of the no observed adverse effect level (NOAEL)
from the rat developmental toxicity study (Price et al., 1996a).
To determine the renal clearance of boron, two blood samples were drawn from each rat.
The first sample was taken 3 hours after gavage dosing on the assumption that the peak boron
concentration in the blood had been achieved (based on data from Usuda et al., 1998). The
second blood sample was taken 12 hours after the initial sample. Rats were placed in metabolic
cages after the first blood sample was taken, and urine was collected during the 12 hours
between the first and second blood sampling.
The urinary concentration of boron at the high dose was significantly higher in pregnant
rats compared with nonpregnant rats but not at the low and mid dose (Table 6-3). The urine
volume was not significantly different in pregnant and nonpregnant rats. The amount of boron
(|-ig/12 hours) excreted in the urine increased proportionately with increasing dose and during the
12-hour collection period was higher (32-73%) in pregnant rats compared to the nonpregnant
rats in the high dose level. This was attributed by the authors to the higher dose of boron
administered to pregnant rats due to their larger body weight and to the higher fractional
excretion of boron (boric acid clearance/creatinine clearance) in the pregnant rats which was
statistically significant at the high dose level. The percentage of administered dose of boric acid
recovered in the urine was significantly higher in the low-dose group compared to the mid- and
high-dose groups for both the nonpregnant and pregnant animals and higher in the pregnant
compared to the nonpregnant rats across dose groups, which was statistically significant at the
high dose only (Table 6-3). Although the diet used for this study was low in boron, it
contributed to the overall dose of boric acid, and these amounts were not included in the nominal
dose levels. When dietary contribution from the low boron diet was included in the dose, the
actual dose levels were approximately 0.4, 3.1, and 30.1 mg/kg boric acid. At the low dose, the
diet contributed another 27% and 33% to the overall dose given to nonpregnant and pregnant
rats, respectively, whereas at the mid and high doses, the diet contributed 3% and 0.3%,
respectively, to the total dose. The authors suggest that the incremental increase at the low dose
may explain the greater recovery of administered dose in the low-dose group.
Proposal Draft
Boron — September, 2006 6-10
-------
Table 6-3 Urinary Boron Concentration, Volume, Mean Excretion, and Percent Recovered in 12 Hours in Nonpregnant
and Pregnant Rats Given Boric Acid by Gavagea'b
Dose (mg
BAVkg-day)
0.3
3.0
30.0
Urinary B (^ig/mL)
Nonpregnantd
1.7±0.6f
(9)
10.1±8.2
(10)
66.8±47.0
(10)
Pregnant11
1.6±0.5
(9)
12.3±5.1
(9)
121.4±47.1g
(11)
Urine Volume (mL)
Nonpregnant
4.3±1.4
(9)
5.2±3.4
(10)
6.8±3.9
(10)
Pregnant
6.1±3.2
(9)
5.3±2.4
(9)
5.4±2.5
(11)
12-hour Urinary B Excretion
(Hg/12hr)
Nonpregnantd
6±1
(9)
32±7
(10)
324±61
(10)
Pregnant11
8±3
(9)
56±16
(9)
561±114g
(11)
Percent of Dose in 12-Hr Urine
(3-15 Hr)
Nonpregnantd'e
50.4±10.6%
(9)
24.6±4.5%
(10)
24.6±4.3%
(10)
Pregnant11'6
55.6±21.4%
(9)
35.6±9.4%
(9)
34.7±6.4%g
(11)
a Sources: U.S. Borax (2000); Vaziri et al. (2001)
b Numbers in parentheses represent number of animals
c Boric Acid (BA)
d Statistically significant difference in urinary boron concentration across dose levels based on two-way analysis of variance (ANOVA), p<0.05
e Statistically significant difference across groups (nonpregnant vs. pregnant) based on two-way ANOVA, p<0.05
f Mean + standard deviation (number of rats)
g Statistically significant difference between nonpregnant and pregnant rats based on multiple range test, p<0.05
Proposal Draft
Boron — September, 2006
6-11
-------
Table 6-4 shows the clearance rates of boron (boric acid), creatinine, and urea expressed
in three different ways: mL/min, mL/min/kg of body weight, and mL/min/cm2 of body surface
area. Boron clearance appeared to be independent of dose within the range of dose levels tested.
The average absolute clearance value for pregnant rats (mL/min) was 1.01 mL/min. The
measurements showed low to moderate variability with a standard deviation of 0.2 mL/min
(CV=0.2). Boron clearance was slightly higher in pregnant rats compared to nonpregnant rats,
but the difference was not statistically significant. The rate of creatinine clearance did not vary
significantly with the different doses of boric acid in either nonpregnant or pregnant rats.
Creatinine clearance, normalized against body weight, however, was significantly greater in
nonpregnant rats compared to pregnant rats. Urea clearance was not significantly different
between nonpregnant and pregnant rats. There were no consistent differences in the rate of urea
clearance with the different doses of boric acid.
Fractional excretion of boron (the ratio of boron clearance/creatinine clearance) was 65%
and 80% in nonpregnant and pregnant rats, respectively. Fractional excretion of urea was lower
in nonpregnant rats than in pregnant rats. The authors indicated that increased fractional
excretion of boron in pregnant rats may be related to physical factors associated with normal
pregnancy due to extracellular volume expansion and renal vasodilation.
Proposal Draft
Boron — September, 2006 6-12
-------
Table 6-4 Clearance of Boron (Boric Acid), Creatinine and Urea in Nonpregnant and Pregnant Rats Given Boric Acid by
Gavage Expressed as mL/min, niL/min/cm2, and mL/min/kga'b
Dose (mg BA/kg)
0.3
3.0
30.0
Boron Clearance (mL/min)
Nonpregnant0
0.77±0.2 (9)d
0.76±0.2 (10)
0.81±0.1 (10)
Pregnant0
1.01±0.2 (9)
0.95±0.2 (9)
1.07±0.2(ll)e
Creatine Clearance (mL/min)
Nonpregnant
1.3±0.4 (9)
1.2±0.4 (10)
1.3±0.4 (10)
Pregnant
1.3±0.5 (9)
1.3±0.4 (9)
1.3±0.3 (11)
Urea Clearance (mL/min)
Nonpregnant
0.85±0.2 (9)
0.84±0.3 (10)
0.96±0.3 (10)
Pregnant
0.89±0.3 (9)
1.14±0.4(9)
1.10±0.3 (11)
expressed as mL/min/cm2
0.3
3.0
30.0
0.0017±0.0004 (9)
0.0017±0.0003(10)
0.0018±0.0003 (10)
0.0020±0.0004 (9)
0.0019±0.0003 (9)
0.0020±0.0003(11)
0.0029±0.0007 (9)
0.0027±0.0008 (10)
0.0029±0.0008 (10)
0.0025±0.0009 (9)
0.0025±0.0006 (9)
0.0025±0.0006(11)
0.0019±0.0005 (9)
0.0018±0.0006 (10)
0.0021±0.0006 (10)
0.0017±0.0005 (9)
0.0022±0.0008 (9)
0.0021±0.0004(11)
expressed as mL/min/kg
0.3
3.0
30.0
3.1±0.8(9)
3.0±0.6 (10)
3.2±0.5 (10)
3.3±0.6 (9)
3.2±0.5 (9)
3.4±0.5(11)
5.2±1.1 (9)c
4.8±1.3 (10)c
5.3±1.6(10)c
4.3±1.5 (9)c
4.2±1.1 (9)c
4.3±1.0(ll)c
3.4±0.9 (9)
3.3±1.1(10)
3.8±1.0(10)
2.9±0.9 (9)
3.8±1.3 (9)
3.5±0.7(11)
a Sources: U.S. Borax (2000); Vaziri et al. (2001)
b Numbers in parentheses represent number of animals
c Statistically significant difference across groups (nonpregnant vs. pregnant) based on tow-way ANOVA, p<0.05
d Mean = standard deviation (number of rats)
e Statistically significant difference between nonpregnant and pregnant rats based on multiple range test, p<0.05
6-13
-------
A human study to measure renal clearance of boron normally consumed in the daily diet
in nonpregnant and pregnant women was conducted (U.S. Borax, 2000; Pahl et al., 2001) in 32
women in good health between the ages of 18 and 40 years, including 16 women in their second
trimester (14-28 weeks) and 16 age-matched nonpregnant women. At the beginning of the
study, all subjects were asked to empty their bladders, and a baseline blood sample was taken.
At the end of this 2 hours another blood sample was taken. The subjects were asked to collect all
urine for the next 22 hours (24 hours from the baseline). A 24-hour blood sample also was
collected.
Urine for each subject was pooled over the initial 2-hour period and over the subsequent
22-hour period. Boron content of blood and pooled urine was analyzed via inductively coupled
plasma-mass spectrometry (ICPMS) following laboratory analytical standards and practices, and
employing adequate quality control measures. Urinary clearance was measured by quantifying
the amount of boron (mg) in the urine and blood. Because the 22-hour clearance samples were
not collected onsite, the 2-hour clearance values were considered to be more accurate due to the
women's compliance with the collection procedures while at the clinic. The urinary clearance of
boron in humans was determined in all individuals and presented as mL blood cleared of boron
per minute per kg body mass. The average clearance rate for boron in pregnant women was 1.02
± 0.55 (mean ± standard deviation; range 0.252-2.028) and the average clearance rate for boron
in nonpregnant women was 0.80 ± 0.31 (mean ± standard deviation; range 0.229-1.358 )
mL/min-kg body mass. These results showed that pregnant women clear boron more effectively
than nonpregnant women, which is consistent with the normal increase in renal blood flow and
glomerular filtration rate during pregnancy.
For the purpose of toxicokinetic modeling, the individual body weights and clearance
values from U.S. Borax (2000) were used to calculate boron clearance in units of mL/min. Table
6-5 shows the clearances in mL/min-kg and body weights in kg for the pregnant women in the
U.S. Borax report. The absolute boron clearances are shown in the last column. The average
boron clearance for these subjects was 66.1 mL/min, with a standard deviation of 32.4 mL/min.
The clearance values, however, were characterized by high variability, with a CV of 0.49.
One factor that may contribute to a higher than expected variability in these clearance
estimates - relative to similar biological values estimated in the Jansen et al. (1984a) and Vaziri
et al. (2001) results - was the indirect estimation of boron intake. Although all subjects were
asked to record their 24-hour dietary intake, the subjects in the study provided incomplete
dietary information. The authors stated that estimates of dietary intake provided from food
frequency questionnaires are of limited accuracy. Boron intake estimated from the renal
excretion of boron in 24 hours was 1.3 mg B/day, from which an average consumption was
estimated at 0.02 mg B/kg-day.
In addition, these boron clearances probably underestimate the true clearance that would
be obtained with higher doses, as in Jansen et al. (1984a). The Pahl et al. (2001) study did not
have the detection limit problem of Jansen et al. (1984a), and only a single 2-hour urine sample
was collected. As complete bladder voiding is problematic in such a short time, underestimation
of total boron excreted is likely. The result would be lower estimated boron clearance values.
Pahl et al. (2001) reported evidence of under-collection of urine in some subjects, but
quantification of underestimate was not possible. In addition, the variance of boron clearance
Proposal Draft
Boron — September, 2006 6-14
-------
reported in the study is very likely an overestimate of the true variability of clearance in the
population. As study subjects could not be kept in the clinic for prolonged periods, multiple
urinary and plasma boron measurements over a longer time interval could not be made.
Therefore, the average of only two plasma samples over 2 hours had to suffice a surrogate for
AUC in the calculation of clearance. The average plasma boron concentration over 2 hours, with
no controls on exposure timing or magnitude, inherently will be more variable than plasma
concentrations obtained from a carefully controlled and monitored study, as in the infusion phase
of the Jansen study. The excess variance would reflect experimental error rather than true
interindividual variability. In the Jansen study, the CV for boron clearance was reduced by a
factor of 13 with larger doses and controlled conditions compared to uncontrolled dietary
exposure.
Proposal Draft
Boron — September, 2006 6-15
-------
Table 6-5 Urinary Clearance of Boron in Pregnant Women51
Subject
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Average
BW (kg)
91.10
53.22
59.08
63.59
69.45
55.92
47.36
59.53
73.96
55.92
76.22
84.34
76.67
64.49
82.53
67.60
2-Hour Boron Clearance Values
mL/min-kg
0.40
0.25
1.43
0.33
2.03
1.76
1.36
1.25
0.54
1.46
0.71
0.81
0.83
1.42
0.71
1.02
mL/min
36.35
13.41
84.43
21.11
140.85
98.37
64.50
74.18
39.72
81.82
54.34
68.23
63.87
91.58
58.27
66.10
a Sources: U.S. Borax (2000)
Proposal Draft
Boron — September, 2006
6-16
-------
Creatinine clearance was normal in all subjects and comparable in pregnant and
nonpregnant women. Comparison of the clearance of boron with creatinine gives insight into
renal tubular handling of boron. Tubular secretion (i.e., into the urine) is indicated if fractional
excretion - the ratio of clearance to glomerular filtration rate (GFR) - is greater than 1. Tubular
reabsorption (i.e., into the blood stream) is indicated if fractional excretion is less than 1. Pahl et
al. (2001) used creatinine clearance as a surrogate for GFR. On this basis, fractional excretion
was 0.57 (+0.32) and 0.47 (+0.14) in pregnant and nonpregnant women, respectively. There was
a trend toward increased fractional excretion or reduced tubular reabsorption in pregnant women,
but the difference was not statistically significant. Creatinine clearance, however, overestimates
GFR, as creatinine is actively secreted from the bloodstream into the kidney tubules. The
magnitude of the overestimation is about 20-30% (Shemesh et al., 1985), which would increase
the nominal fractional excretion of boron to about 70%. Furthermore, the probable
underestimation of boron clearance in the Pahl et al. (2001) study would result in higher actual
fractional excretion, such that boron clearance would approach GFR.
Several studies have addressed the application of hemodialysis in decreased renal
function as an effective method to remove boron from human blood. Although these studies
uniformly demonstrate the effective movement of boron across a non-biological dialysis
membrane from blood into dialysate, the study of Usuda et al. (1997) is perhaps the most
well-reported. In a study to ascertain whether plasma protein binding altered the effectiveness of
hemodialysis of boron, 17 human subjects in long-term hemodialysis were monitored before and
during dialysis employing a polyvinyl membrane. Clearances of boron, blood urea nitrogen,
phosphorus, and creatinine followed. Results indicated that boron clearance was equal to that of
blood urea nitrogen and slightly, but significantly, exceeded that of phosphorus and creatinine.
The fraction of serum boron available for dialysis was nearly 80%, indicating that approximately
20% of boron was not available for dialysis, potentially for the reason of association with plasma
constituents. However, the study did not derive the on- and off-rates of binding, so that even if
this approximately 20% of plasma boron was associated with proteins, the measure would only
represent the fraction of boron associated with plasma proteins at steady state. That is, at any
one time, 20% of boron would be associated with proteins. For this to have an impact on renal
filtration, the duration of association would have to exceed the time for a given unit of blood
containing boron to traverse the glomerulus. It also is possible that boron associates with and
dissociates from proteins multiple times during passage through the glomerulus. If this were the
case, the impact of association of boron with plasma protein on renal filtration would be
negligible, and would explain why boron clearance would not be impacted by association with
plasma proteins. In light of the similarity among the renal (filtration) clearance of these four
compounds, the authors concluded that there seems to be relatively little relation of boron to
serum constituents of macromolecules which might influence diffusion across membranes.
Several lines of evidence lead to the conclusion that the filtration mechanism, a passive
mechanism, is responsible for the urinary elimination of boron from mammals. This information
comprises chemical and biochemical data, as well as information from pharmacokinetic studies
in rats and humans. Renal filtration, or glomerular filtration, is routinely investigated in humans
in a clinical setting, and is monitored as part of prenatal care in this country. Glomerular
filtration rate is expressed in units of volume/time and indicates the volume of blood filtered
(cleared of substances) by the kidney per unit time, usually corrected for body mass
(mL/minute/kg). The characteristics of filtered contaminants include low molecular weight and
Proposal Draft
Boron — September, 2006 6-17
-------
diameter, neutrally charged molecule, lack of significant protein binding, and lack of interaction
with the active renal mechanisms of tubular secretion and/or tubular reabsorption.
Boron is always found in nature covalently bound to oxygen as some form of borate (e.g.,
boric acid, tetraborate, etc.). The boron-oxygen bonds are very strong and will not be broken
except under extreme laboratory conditions. Boron (borates) exists in the blood as neutral low
molecular weight and molecular diameter unbound molecules. The ionic form is controlled by
the pKa of the molecule and the pH of the aqueous medium. Uncharged monomeric boric acid is
B(OH)3, with a molecular weight of 58.8; in the negatively charged form, boric acid exists as
B(OH)4, with a molecular weight of 75.8. At the pH of the human blood (i.e., pH = 7.4), the
expected low concentrations of borate (10"6 to 10"5 M) will be present as 98.4% B(OH)3 and
1.6% B(OH)4 - ion because of the weak acidity (pKa = 9.2) of boric acid (Woods, 1994, 1996).
This has been confirmed analytically by nuclear magnetic resonance spectroscopy (Woods,
1994) and Raman spectrometry (De Vette et al., 2001). Thus, at concentrations below 0.025M,
essentially all borates dissociate to form low molecular weight, uncharged molecules. The
observed boron concentrations in pregnant rats were approximately 2.5 xlO"6 M (Vaziri et al.,
2001), and in humans were much lower (Pahl et al., 2001). Thus, 98.4% of the boron in blood
and biological fluids of rats and humans exists in the form of a small, uncharged molecule which
should pass through biological membranes, including those of the glomerulus. Any ionic or
covalent binding to plasma proteins would be negligible. These properties predispose boric acid
to urinary elimination through renal filtration mechanisms.
The effect of plasma protein binding is a decrease in the movement of the substance from
blood into extravascular tissues and fluids, including urine. The rapid absorption and urinary
elimination of near-complete administered doses of boron across multiple studies are
inconsistent with the concept of plasma protein binding for boron. Magour et al. (1982) and Ku
et al. (1991) separately demonstrated that concentrations of boron in plasma and soft tissues
reached equilibrium at dramatically similar concentrations within hours of administration.
Subsequently, elimination profiles from plasma and soft tissues were similar. Usuda et al.
(1997) demonstrated that if boron is associated with plasma macromolecular constituents, the
"relatively little" relation to these components does not result in a decrease in boron filtration as
compared to three plasma constituents whose renal filtration were concomitantly measured.
These and other findings indicate that binding is unlikely in either plasma or soft tissues, and that
administered boron readily passes from blood across biological membranes. In both rats and
humans, boron concentration data have been evaluated to reveal a volume of distribution
consistent with distribution of boron into total body water. This finding is consistent with lower
concentrations being attained in adipose tissue, given its low content of water compared with
other soft tissues. Human studies conducted by Usuda et al. (1997) and others investigated the
removal of boron from human subjects undergoing routine hemodialysis therapy for renal
dysfunction. Those data demonstrated an effective removal of boron from human blood across a
non-biological membrane (devoid of active transport or reabsorption mechanisms) consistent
with ready movement of boron across permeable membranes. Although the plasma protein
binding of boron has not been specifically investigated in either rats or humans, these lines of
evidence lead to the conclusion that plasma protein binding, if it occurs, does not inhibit the
movement of boron across biological membranes and, thus, would not impede effective filtration
of boron in either rats or humans.
Proposal Draft
Boron — September, 2006 6-18
-------
Tubular reabsorption, if it is a factor, will be an issue at dietary levels, and its impact will
diminish with increasing dose. The magnitude of the contribution to boron clearance variability,
however, is much less than would be suggested by the fractional clearance data from both the
human (Pahl et al., 2001) and rat (Vaziri et al., 2001) studies. An average fractional excretion of
0.57 was reported for pregnant women in the Pahl et al. (2001) study (similar results for rats),
suggesting that 43% of boron filtered through the glomerulus was reabsorbed into the
bloodstream. Boron fractional excretion in the Pahl study, however, was calculated relative to
creatinine clearance, which overestimates GFR by about 20% (Shemesh et al., 1985). Correcting
for that overestimate yields a fractional clearance of about 0.7, indicating a lesser influence of
reabsorption on boron clearance than reported. The variability in reabsorption is probably small
by comparison to the variability in GFR. Furthermore, the high boron clearance variability for
uncontrolled low-dose dietary exposure decreases dramatically under more controlled,
higher-dose conditions (Jansen et al., 1984a). In the Jansen et al. (1984a) study, the CV of 0.09
for boron clearance at a dose of 105 mg (see Table 6-2), or 1.5 mg/kg (assuming an average
body weight of 70 kg), is less than that for GFR in females, which ranges from 0.11 to 0.21 for
pregnant or nonpregnant women (Dunlop, 1981; Sturgiss et al., 1996; Krutzen et al., 1992).
Thus, the variability in GFR may actually slightly overestimate variability of boron clearance in
exposed humans. GFR is slightly higher in men than women (Ventura et al., 1999), but
increases by over 50% in pregnancy (Dunlop, 1981; Sturgiss et al., 1996; Krutzen et al., 1992).
GFR variability appears to be similar in pregnant and nonpregnant women (Dunlop, 1981;
Sturgiss et al., 1996; Krutzen et al., 1992). Assuming that GFR variability in men and women is
the same, by analogy, boron clearance variability should be similar. In addition, the variance of
boron clearance is less than the variance of creatinine clearance (a measure of GFR) when
assessed in the same subjects (Jansen, 1984a). Therefore, it is unlikely that GFR variance
underestimates boron clearance variance, and would not need further quantitative adjustment.
The contribution of tubular reabsorption is unlikely to affect the variability of renal elimination
of boron at the higher doses (compared to dietary levels) of concern in deriving an RfD.
Plasma Clearance and Half-Life
In a study conducted with human volunteers and carefully administered doses of 570-620
mg boric acid (91-108.5 mg B), plasma concentration-time curves were followed over 3 days
and were markedly biphasic. Terminal elimination half-lives were calculated for individuals
(n=6) and demonstrated a range of 12.5-26.6 hours and a mean value of 21.0 ± 4.9 hours when
calculated from the data collected over the initial 72 hours post-dose (Jansen et al., 1984a).
From this study, a total mean volume of distribution of 104.7 rnL/100 g body weight can be
calculated. A second study reported by Litovitz et al. (1988) investigated incidences of boron
poisoning. Although this study did not document many important data (dose, time post-dose that
examination began, number of concentrations used to estimate half-lives, etc.), the range of
half-lives compares favorably with the well-controlled study presented by Jansen et al. (1984a).
When linear regression analysis was used to fit the plasma concentration data, estimates of
half-lives ranged from 4.0-27.8 hours, with an overall mean value of 13.4 ± 7.1 hours. Astier
(1988) reported a plasma half-life of 28.7 hours after acute ingestion of 45 g boric acid (7.9 g B)
in two doses over a 20-hour period.
A pharmacokinetic study (Usuda et al., 1998) in 10 rats, following an oral administration
of sodium tetraborate containing 0.4 mg B/100 g body weight where 0.5-1 mL samples were
drawn at nine different times during a 24-hour time period, reported a monophasic elimination of
Proposal Draft
Boron — September, 2006 6-19
-------
boron from plasma, demonstrating a plasma half-life mean of 4.64 ± 1.19. This study also cited
a high volume of distribution of 142.0 ± 30.2 rnL/100 g body weight. One of the limitations of
this study was that the large amount of blood drawn from the rats in the 24-hour period may have
physiologically compromised the rats.
A human study (U.S. Borax, 2000; Pahl et al., 2001) was conducted to measure renal
clearance of boron normally consumed in the daily diet in nonpregnant and pregnant women. At
the beginning of the study, a baseline blood sample was taken. During the 2 hours following the
baseline blood sample, all urine samples were collected. Blood samples were drawn at 2 hours
and 24 hours after the baseline blood samples. Plasma boron levels were measured at these three
time periods. Mean plasma boron levels obtained at baseline and 2 hours after the beginning of
the study were similar between the pregnant and nonpregnant subjects. After 24 hours, plasma
boron levels were lower in the pregnant women when compared with nonpregnant women, but
there was a significant variability in the plasma values in both groups.
In a plasma clearance study of boron sponsored by U. S. Borax (Vaziri et al., 2001) in
pregnant and nonpregnant rats given boric acid at dose levels of 0.3, 3.0, and 30 mg boric acid,
plasma concentrations of boron were markedly lower 15 hours after dosing than at 3 hours after
dosing. Mean plasma levels of boron were slightly higher in pregnant rats than in nonpregnant
rats (statistically significant in only the high dose) given the same dose of boric acid.
In a study (U.S. Borax, 2000; Vaziri et al., 2001) conducted to estimate the plasma
half-life of boric acid in the Sprague-Dawley rat, six nonpregnant and six pregnant rats were
given low B in the diet for 7 days. On day 8 of the study, all rats received a single oral dose of
30 mg/kg of boric acid at approximately 9:00 a.m. This dose was the high dose used in the renal
clearance study and was selected as the best to examine the linearity of the boron plasma curve
at the highest concentration. Six 0.25 mL blood samples were drawn from each animal during a
12-hour period starting at noon on day 8 of the study. The blood samples were taken at 2- to
3-hour intervals. Gavage administration of 30 mg/BA/kg-day resulted in plasma levels of 1.82 ±
0.32 and 1.78 ± 0.32 |_ig/mL among pregnant and nonpregnant rats in the first blood sample
taken 3 hours after dosing. This was followed by a monophasic decline in plasma boron
concentration in both the pregnant and nonpregnant rats. The plasma concentration curves were
consistent with a one-compartment model. Based on the shape of the plasma concentration
curve, there was no evidence of saturation kinetics in either the nonpregnant or pregnant rats.
The estimated half-life of boric acid in nonpregnant and pregnant rats was 2.9 and 3.2 hours,
respectively, which was not statistically different.
Proposal Draft
Boron — September, 2006 6-20
-------
7.0 HAZARD IDENTIFICATION
7.1 Studies in Humans - Epidemiology and Case Reports
7.1.1 Oral Exposure
Sayli et al. (1998) reported on a study of the relationship between exposure to boron in
the drinking water and fertility in two geographic regions of Turkey. Drinking water boron
concentrations were markedly higher in one region (2.05-29 mg/L) than in the other (0.03-0.4
mg/L). The study population comprised residents (primarily males who had ever been married)
from these regions who could provide reproductive histories for three generations of family
members (n=159 in one region and 154 in the other, 6.7% of the population in both). There was
no difference between the regions regarding percentage of married couples with live births in
any generation. Secondary sex ratios appeared to differ, with an excess of female births in the
high-boron region (M/F = 0.89) and a slight excess of male births in the low-boron region (M/F
= 1.04), but no statistical analysis was performed, and other factors reported to affect sex ratio
(parental age, rate of elective abortion, multiple births) were not taken into account.
A large number of accidental poisoning cases are reported in the literature; however,
quantitative estimates of absorbed dose are limited. Baker and Bogema (1986) reported
quantitative estimates of two sibling infants who ingested formulas accidentally prepared from a
boric acid eyewash solution. These infant doses ranged from 30.4-94.7 mg B/kg-day. The
sibling who ingested 30.4 mg B/kg-day had a serum level of 9.79 mg B/mL and displayed a rash
on his face and neck but later remained asymptomatic. The sibling who ingested 94.7 mg
B/kg-day had serum boron values of 25.7 mg B/mL and experienced diarrhea, erythema of the
diaper area, and vomiting a small amount of formula. Case reports and surveys of poisoning
episodes were recently reviewed by Craan et al. (1997), WHO (1998a), Culver and Hubbard
(1996), and Ischii et al. (1993). The most frequent symptoms of boron poisoning are vomiting,
abdominal pain, and diarrhea. Other common symptoms include lethargy, headache,
lightheadedness, and rash. For boric acid, the minimum lethal dose by oral exposure is
approximately 15-20 g in adults, 5-6 g in children, and 2-3 g in infants.
Acute adult quantitative dose response data range from 1.4 mg B/kg to a high of 70 mg
B/kg (Culver and Hubbard, 1996). In cases where ingestion was less than 3.68 mg B/kg,
subjects were asymptomatic. Data in the 25-35 mg B/kg range were from patients undergoing
boron neutron capture therapy for brain tumors. They displayed nausea and vomiting at 25 mg
B/kg, and at 35 mg B/kg additional symptoms included skin flush. A patient recuperating from
surgery had boric acid solution (70 mg B/kg) injected into the subcutaneous fluid infusion,
which resulted in severe cutaneous and gastrointestinal symptoms. The patient recovered after
hydration and diuresis.
Because boron compounds were used for various medical conditions including epilepsy,
malaria, urinary tract infections, and exudative pleuritis from the mid 1800s until around 1900,
some data are available on longer term exposure. Culver and Hubbard (1996) report on early
cases of boron treatment for epilepsy from 2.5 to 24.8 mg B/kg-day for many years. Signs and
symptoms reported in patients receiving 5 mg B/kg-day and above were indigestion, dermatitis,
alopecia, and anorexia. One epilepsy patient who received 5.0 mg B/kg-day for 15 days
Proposal Draft
Boron — September, 2006 7-1
-------
displayed indigestion, anorexia, and dermatitis, but the signs and symptoms disappeared when
the dose was reduced to 2.5 mg B/kg-day.
O'Sullivan and Taylor (1983) reported seizures and other milder effects in seven infants
who consumed boron in a honey-borax mixture applied to pacifiers. Five of the infants had a
history suggestive of a familial-reduced convulsive threshold. The seizures ceased when the
honey-borax treatment was stopped. The infants, who ranged in age from 6-16 weeks (at the end
of the exposure period), were exposed to the honey-borax mixture over a period of 4-10 weeks.
Original estimates of exposure were based on an error by the author (Taylor, 1997) concerning
intake in jars versus grams of boron per week. The doses were recalculated from the information
given by the author, based an estimated daily ingestion of honey-borax mixture and the analysis
of the borax content in the mixture. Details of the analytic methods were not provided. Average
estimated daily intakes of borax ranging from 429-1287 mg can be calculated directly from data
provided by the authors. Average body weights over the exposure period for the infants in this
study ranging from 4.3-5.3 kg based on estimates from the Exposure Factors Handbook (U.S.
EPA, 1997a). Using the estimated body weights and a factor of 0.113 to estimate the boron
content in borax, the equivalent boron exposure levels would have been about 9.6-33 mg/kg-day.
The lowest exposure level of 3.2 mg/kg-day would be considered a lowest observed adverse
effect level (LOAEL) for a fairly severe effect. Concentrations of boron in blood of 2.6, 8.4, and
8.5 |ig/mL were reported for three of the subjects. Blood boron concentrations did not correlate
well with estimated ingestion levels; the lowest blood boron concentration was measured for the
infant with the highest estimated boron intake. Blood boron levels also were reported for a
control group of 15 children aged 2-21 months, who had received no boron supplement and,
presumably, had suffered no seizures. The control group blood boron values ranged from 0-0.63
|ig/mL and averaged 0.21 jig/mL, with a standard deviation of 0.17 jig/mL. The lowest boron
blood level associated with seizures, 2.6 |ig/mL, was about 4 times the highest control level and
12 times the average control level, suggesting that the standard 10-fold uncertainty factor may be
adequate for estimating an NOAEL. However, there was no indication whether any infants
predisposed to seizures were in the control population. The presumptive boron NOAEL would
be 0.32 mg/kg-day for a sensitive human subpopulation. Given the relatively uncomplicated
boron toxicokinetics, the lack of correlation of blood boron and estimated ingestion rates suggest
that the data may not be completely reliable. Based on the latter consideration, the indirect
exposure estimation, and the lack of detail in the publication, this study should not be considered
as the critical factor for derivation of the RfD, but the potential for seizures in infants should be
considered in establishing the RfD.
Case reports and surveys of poisoning episodes were recently reviewed by Craan et al.
(1997), WHO (1998a), Culver and Hubbard (1996), and Ischii et al. (1993). The most frequent
symptoms of boron poisoning are vomiting, abdominal pain, and diarrhea. Other common
symptoms include lethargy, headache, lightheadedness, and rash. For boric acid, the minimum
lethal dose by oral exposure is approximately 15-20 g in adults, 5-6 g in children, and 2-3 g in
infants.
Wegman et al. (1994) conducted a longitudinal study of respiratory function in workers
with chronic exposure to sodium borate dusts. Participants in the Garabrant et al. (1985) study
were re-tested for pulmonary function 7 years after the original survey. Of the 629 participants
in the original study in 1981, 371 were available for re-testing in 1988. Of these, 336 performed
Proposal Draft
Boron — September, 2006 7-2
-------
pulmonary function tests (303 produced acceptable tests in both years). Cumulative exposure
was estimated for each participant for the years 1981-1988 as a time-weighted sum of the
exposure in each job held during that time. Exposure prior to 1981 was not included due to the
scarcity of monitoring data for those years. Pulmonary function FEVj (forced expiratory volume
in 1 sec) and FVC (forced vital capacity) in study subjects declined over the 7-year period at a
rate very close to that expected based on standard population studies. Cumulative borate
exposure over the years 1981-1988 was not related to the change in pulmonary function. Acute
studies showed statistically significant, positive dose-related increases in eye, nasal, and throat
irritation; cough; and breathlessness with borate exposure (6-hr TWA or 15-min TWA). The
same relationships were present when effects were limited to moderate severity or higher. There
was no evidence for an effect of borate type (decahydrate, pentahydrate, anhydrous) on response
rate.
7.1.2 Inhalation Exposure
Tarasenko et al. (1972) reported low sperm count, reduced sperm motility, and elevated
fructose content of seminal fluid in semen analysis of 6 workers who were part of a group of 28
male Russian workers exposed for 10 or more years to high levels of vapors and aerosols of
boron salts (22-80 mg/m3) during the production of boric acid. The men in this report were
studied using an Sexual Function of Man questionnaire. The results indicated that the group of
28 male exposed workers had decreased sexual function compared with 10 workers who had no
contact with boric acid. However, the analysis of data from wives of the men from the exposed
and control groups showed no differences. This study is of limited value for risk determinations
due to the small sample size; sparse details on subjects regarding smoking habits, diet, other
chemical exposures; and lack of methodology information on semen analysis. In response to this
report and reports of reproductive effects in animal studies, a controlled epidemiology study of
reproductive effects was initiated in U.S. workers exposed to sodium borates.
Whorton et al. (1994a,b, 1992) examined the reproductive effects of sodium borates on
male employees at a borax mining and production facility in the United States. A total of 542
subjects participated in the study (72% of the 753 eligible male employees) by answering a
questionnaire prepared by the investigators. The median exposure concentration was
approximately 2.23 mg/m3 sodium borate (roughly 0.31 mg B/m3). Average duration of
employment in participants was 15.8 years. Reproductive function was assessed in two ways.
First, the number of live births to the wives of workers during the period from 9 months after
occupational exposure began through 9 months after it ended was determined, and this number
was compared to a number obtained from the national fertility tables for U.S. women (an
unexposed control population). Wives of workers and controls were matched for maternal age,
parity, race, and calendar year. This comparison produced the standardized birth ratio (SBR),
defined as the observed number of births divided by the expected number. The investigators
then examined possible deviations of the ratio of male to female offspring relative to the U.S.
ratio.
There was a significant excess in the SBR among participants as a whole (Whorton et al.,
1994a,b, 1992). Study participants fathered 529 births versus 466.6 expected (SBR=113,
p<0.01). This excess occurred even though the percentage of participants who had vasectomies
(36%) was 5 times higher than the national average of 7% implicit in the expected number of
Proposal Draft
Boron — September, 2006 7-3
-------
births. Participants were divided into five equal-size groups (n = 108/109) based on average
workday exposure to sodium borates (<0.82, 0.82-1.77, 1.78-2.97, 2.98-5.04, and >5.05 mg/m3).
There was no trend in SBR with exposure concentration; the SBR was significantly elevated for
both the low- and high-dose groups, and close to expected for the three mid-dose groups. There
were 42 participants who worked high-exposure jobs for 2 or more consecutive years. Mean
sodium borate exposure in this group was 23.2 mg/m3 (17.6-44.8 mg/m3), and mean duration of
employment in a high-exposure job was 4.9 years (range: 2.1-20.4 years). The SBR for the 42
workers was close to expected (102) despite a 48% vasectomy rate. These workers also had
elevated SBR during the actual period of high exposure. An examination of SBR for all
participants by 5-year increments from 1950 to 1990 revealed no significant trend in either
direction over time.
Analyses of the percentage of female offspring showed an excess of females that
approached statistical significance (52.7% vs. 48.8% in controls) (Whorton et al., 1994a,b,
1992). This excess was not related to exposure, however, as the percentage of female offspring
decreased with increasing sodium borate exposure concentration (from 55.3% in the low-dose
group to 49.2% in the high-dose group). Moreover, individuals with 2 or more consecutive years
in high borate exposure jobs had more boys than girls. The investigators concluded that
exposure to inorganic borates did not appear to adversely affect fertility in the population
studied. This study, while adequately conducted, has several inherent limitations (SBR is less
sensitive than direct measures of testicular effects, exposure information was limited,
applicability of total U.S. fertility rates as control is questionable). Thus, the human data are
insufficient to determine if boron may cause male reproductive toxicity (IEHR, 1997).
Whorton et al. (1992) also studied the effects of borates on reproductive function of
exposed female employees. Reproductive function was assessed in the same way as it was for
wives of male employees. A total of 81 employees were eligible, 68 of whom participated in the
study. No information was provided regarding matching of the exposed and control groups. The
SBR was 90 (32 offspring observed, 35.4 expected), indicating a deficiency, although not
statistically significant, in live births among exposed females. When the data were analyzed per
exposure category, the 76 employees (some nonparticipants apparently were included) in the
low- and medium-exposure category showed a nonstatistically significant deficit of births (37
compared to 43.5 expected, SBR=85). No statistical differences were observed between exposed
and controls when the results were analyzed by exposure categories. The authors concluded that
the exposure to inorganic borates did not appear to affect fertility in the population studied.
However, the small sample size may have precluded a meaningful statistical analysis of the
results.
Swan et al. (1995) investigated the relationship between spontaneous abortion in women
employed in the semiconductor manufacturing industry and various chemical and physical
agents used in the industry, including boron. The study population consisted of 904 current and
former female employees who became pregnant while working at 1 of 14 U.S. semiconductor
companies between 1986 and 1989. Approximately one-half of those included were fabrication
workers with some chemical exposure. Exposure classifications were based on jobs held at
conception and level of exposure to specific agents during the first trimester. The risk of
spontaneous abortion was increased in fabrication workers compared with other workers, and
particularly within the subgroup of workers who performed masking (a group with relatively low
Proposal Draft
Boron — September, 2006 7-4
-------
boron exposure). No significant association was found between exposure to boron and
spontaneous abortion risk.
The respiratory and irritant effects of industrial exposure to boron compounds have also
been studied. The studies were conducted at the same borax mining and production facility as
the reproduction study of Whorton et al. (1994a,b, 1992). A health survey of workers at the
plant found complaints of dermatitis, cough, nasal irritation, nose bleeds, and shortness of breath
(Birmingham and Key, 1963). Air concentrations of borate dust were not reported, but were
high enough to interfere with normal visibility. In response to this report, a cross-sectional study
of respiratory effects (questionnaire, spirometric testing, roentgenograms) was performed on 629
male workers at the plant (Ury, 1966). The study was inconclusive, but did find suggestive
evidence for an association between respiratory ill health and inhalation exposure to dehydrated
sodium borate dust based on analysis of forced expiratory volume and respiratory illness data in
the subgroup of 82 men who had worked for at least 1 year at the calcining and fusing processes
compared with 547 others who had never worked at these processes.
Additional studies were performed by Garabrant et al. (1984, 1985). Garabrant et al.
(1985) studied a group of 629 workers (93% of those eligible) employed for 5 or more years at
the plant and employed in an area with heavy borax exposure at the time of the study. Workers
were categorized into four groups according to borax exposure (1.1, 4.0, 8.4, and 14.6 mg/m3
borax), and frequency of acute and chronic respiratory symptoms was determined. Statistically
significant, positive dose-related trends were found (in order of decreasing frequency) for
dryness of mouth, nose, or throat; eye irritation; dry cough; nose bleeds; sore throat; productive
cough; shortness of breath; and chest tightness. Frequency of these symptoms in the high-dose
group ranged from 33% down to 5%. Pulmonary function tests and chest x-rays were not
affected by borax exposure. The researchers concluded that borax appears to cause simple
respiratory irritation that leads to chronic bronchitis, with no impairment of respiratory function
at the exposure levels in this study. Irritation occurred primarily at concentrations of 4.4 mg/m3
or more. Garabrant et al. (1984) studied a subgroup of the 629 workers who were exposed to
boric oxide and boric acid. Workers who had held at least one job in an area with boron oxide or
boric acid exposure (n=l 13) were compared with workers who had never held a job in an area
with boron oxide or boric acid, but who had held at least one job in an area with low- or
minimal- exposure to borax (n=214). The boron oxide/boric acid workers had significantly
higher rates of eye irritation; dryness of mouth, nose, or throat; sore throat; and productive
cough. Mean exposure was 4.1 mg/m3, with a range of 1.2 to 8.5 mg/m3. The researchers
concluded that boron oxide and boric acid produce upper respiratory and eye irritation at less
than 10 mg/m3.
7.2 Prechronic and Chronic Studies and Cancer Bioassays in Animals - Oral and
Inhalation
7.2.1 Oral Exposure
In the following studies, doses not reported by the investigators were approximated from
dietary or drinking water concentrations of boron using food factors (rat: 0.05; dog: 0.025;
mouse: 0.1) (1 ppm = 0.025 mg/kg-day assumed dog food consumption) and body-weight and
water consumption values (mouse: 0.03 kg and 0.0057 L/day; rat: 0.35 kg and 0.049 L/day)
Proposal Draft
Boron — September, 2006 7-5
-------
specified by the U.S. EPA (1980, 1988). Doses in mg boric acid were converted to mg boron by
multiplying by the ratio of the formula weight of boron to the molecular weight of boric acid
(10.81/61.84 = 0.1748). Similarly, doses in mg borax were converted to mg boron by
multiplying by the ratio of the formula weight of boron to the molecular weight of borax (4 x
10.81/381.3=0.1134).
The subchronic and chronic toxicity of borax and boric acid has been studied in dogs
administered these compounds in the diet (Weir and Fisher, 1972; U.S. Borax Research Corp.,
1963, 1966, 1967). In the subchronic study, groups of beagle dogs (5/sex/dose/compound) were
administered borax (sodium tetraborate decahydrate) or boric acid for 90 days at dietary levels of
17.5, 175, and 1750 ppm boron (male: 0.33, 3.9 and 30.4 mg B/kg-day; female: 0.24, 2.5 and
21.8 mg B/kg-day) and compared with an untreated control group of 5 dogs/sex (Weir and
Fisher, 1972; U.S. Borax Research Corp., 1963). On day 68 of the study, a high-dose male dog
died as a result of complications of diarrhea with severe congestion of the mucosa of the small
and large intestines and congestion of the kidneys. No clinical signs of toxicity were evident in
the other dogs. The testes were the primary target of boron toxicity. At the high dose, mean
testes weight was decreased 44% (9.6 g) in males fed borax and 39% (10.5 g) in males fed boric
acid compared with controls (17.2 g). Also at this dose, mean testes:body weight ratio (control:
0.2%; borax: 0.1%; boric acid: 0.12%) and mean testes:brain weight ratio (control: 22%; borax:
12%) were significantly reduced. Decreased testes:body weight ratio also was observed in one
dog from the mid-dose (175 ppm) boric acid group. Microscopic pathology revealed severe
testicular atrophy in all high-dose male dogs, with complete degeneration of the spermatogenic
epithelium in 4/5 cases. No testicular lesions were found in the lower-dose groups.
Hematological effects were also observed in high-dose dogs. Decreases were found for both
hematocrit (15 and 6% for males and females, respectively) and hemoglobin (11% for both
males and females) at study termination in borax-treated dogs. Pathological examination
revealed accumulation of hemosiderin pigment in the liver, spleen, and kidney, indicating
breakdown of red blood cells, in males and females treated with borax or boric acid. Other
effects in high-dose dogs were decreased thyroid:body weight ratio (control: 0.009%; borax:
0.006%; boric acid: 0.006%) and thyroid:brain weight ratio (control: 0.95%; borax: 0.73%) in
males; increased brain:body weight ratio (borax) and liverbody weight ratio (boric acid) in
females; a somewhat increased proportion of solid epithelial nests and minute follicles in the
thyroid gland of borax-treated males; lymphoid infiltration and atrophy of the thyroid in
boric-acid treated females; increased width of the zona reticularis (borax males and females,
boric acid females); and zona glomerulosa (boric acid females) in the adrenal gland. This study
identified an LOAEL of 1750 ppm boron (male: 30.4 mg B/kg-day; female: 21.8 mg B/kg-day)
and an NOAEL of 175 ppm boron (male: 3.9 mg B/kg-day; female: 2.5 mg B/kg-day) based on
systemic toxicity in dogs following subchronic exposure.
In the chronic toxicity study, groups of beagle dogs (4/sex/dose/compound) were
administered borax or boric acid by dietary admix at concentrations of 0, 58, 117, and 350 ppm
boron (0, 1.4, 2.9, and 8.8 mg B/kg-day) for 104 weeks (Weir and Fisher, 1972; U.S. Borax
Research Corp., 1966). There was a 52-week interim sacrifice and a 13-week "recovery" period
after 104 weeks on test article for some dogs. Four male dogs served as controls for the borax
and boric acid dosed animals. One male control dog was sacrificed after 52 weeks, two male
control dogs were sacrificed after 104 weeks, and one was sacrificed after the 13-week recovery
period with 104 weeks of treatment. The one male control dog sacrificed after the 13-week
Proposal Draft
Boron — September, 2006 7-6
-------
recovery period demonstrated testicular atrophy. Sperm samples used for counts and motility
testing were taken only on the control and high-dose male dogs prior to the 2-year sacrifice. At a
dose level of 8.8 mg B/kg-day in the form of boric acid, one dog sacrificed at 104 weeks had
testicular atrophy. Two semen evaluations (taken after 24 months treatment) were preformed on
dogs treated at the highest dose (8.8 mg B/kg-day). Two of two borax-treated animals had
samples that were azoospermic and had no motility, while one of two boric acid treated animals
had samples that were azoospermic. The authors reported that there did not appear to be any
definitive test article effect on any parameter examined. The study pathologist considered the
histopathological findings to be "not compound-induced." Tumors were not reported.
In a follow-up to this study, groups of beagle dogs (4/sex/dose/compound) were given
borax or boric acid in the diet at concentrations of 0 and 1170 ppm boron (0 and 29.2 mg
B/kg-day) for up to 38 weeks (Weir and Fisher, 1972; U.S. Borax Research Corp., 1967). New
control dogs (4 males) were used for this follow up study. Two were sacrificed at 26 weeks and
two at 38 weeks. At the 26-week sacrifice, one of two had spermatogenesis and (5%) atrophy.
One was reported normal. At 38 weeks, one had decreased spermatogenesis, and the other had
testicular atrophy. The test animals had about an 11% decrease in the rate of weight gain when
compared with control animals, throughout the study. Interim sacrifice of two animals from
each group at 26 weeks revealed severe testicular atrophy and spermatogenic arrest in male dogs
treated with either boron compound. Testes weight, testes:body weight ratio, and testes:brain
weight ratios were all decreased. Effects on other organs were not observed. Exposure was
stopped at 38 weeks; at this time, one animal from each group was sacrificed and the remaining
animal from each group was placed on the control diet for a 25-day recovery period prior to
sacrifice. After the 25-day recovery period, testes weight and testes weightbody weight ratio
were similar to controls in both boron-treated males, and microscopic examination revealed the
presence of moderately active spermatogenic epithelium in one of the dogs. The researchers
suggested that this finding, although based on a single animal, indicates that boron-induced
testicular degeneration in dogs may be reversible upon cessation of exposure. When the 2-year
and 38-week dog studies are considered together, an overall NOAEL and LOAEL for systemic
toxicity can be established at 8.8 and 29.2 mg B/kg-day, respectively, based on testicular atrophy
and spermatogenic arrest.
Weir and Fisher (1972) fed Sprague-Dawley rats a diet containing 0, 117, 350, or 1170
ppm boron as borax or boric acid for 2 years (approximately 0, 5.9, 17.5, or 58.5 mg B/kg-day).
There were 70 rats/sex in the control groups and 35/sex in the groups fed boron compounds. At
1170 ppm, rats receiving both boron compounds had decreased food consumption during the
first 13 weeks of study and suppressed growth throughout the study. Signs of toxicity at this
exposure level included swelling and desquamation of the paws, scaly tails, inflammation of the
eyelids, and bloody discharge from the eyes. Testicular atrophy was observed in all high-dose
males at 6, 12, and 24 months. The seminiferous epithelium was atrophied, and the tubular size
in the testes was decreased. Testes weights and testes:body weight ratios were significantly
(p<0.05) decreased. Brain and thyroid:body weight ratios were significantly (p<0.05) increased.
No treatment-related effects were observed in rats receiving 350 or 117 ppm boron as borax or
boric acid. This study identified an LOAEL of 1170 ppm (58.5 mg B/kg-day) and an NOAEL of
350 ppm (17.5 mg B/kg-day) for testicular effects. Based on effects observed in the high-dose
group, it appears that a maximum tolerated dose (MTD) was achieved in this study. The study
was designed to assess systemic toxicity; only tissues from the brain, pituitary, thyroid, lung,
Proposal Draft
Boron — September, 2006 7-7
-------
heart, liver, spleen, kidney, adrenal, pancreas, small and large intestine, urinary bladder, testes,
ovary, bone, and bone marrow were examined histopathologically. Tumors were not mentioned
in the report. Nevertheless, NTP (1987) concluded that this study provided adequate data on the
lack of carcinogenic effects of boric acid in rats, and accordingly, conducted its carcinogenicity
study only in mice.
Weir and Fisher (1972) also conducted studies of boron toxicity in rats. Sprague-Dawley
rats (10/sex/dose) were fed borax or boric acid in the diet for 90 days at levels of 0, 52.5, 175,
525, 1750, and 5250 ppm boron (approximately 0, 2.6, 8.8, 26.3, 87.5, and 262.5 mg B/kg-day,
respectively) calculated by assuming reference values of 0.35 kg bw and a food factor of 0.05 for
rats. Both borax and boric acid produced 100% mortality at the highest dose and complete
atrophy of the testes in all males fed diets containing 1750 ppm boron. Overt signs of toxicity at
these two dose levels included rapid respiration, eye inflammation, swelling of the paws, and
desquamation of the skin on paws and tails. At 1750 ppm boron, both compounds produced
significant (p<0.05) decreases in body weight and in the mean weights of the liver, kidneys,
spleen, and testes. At lower doses, changes in organ weights were inconsistent. At 52.5 ppm
boron, increases in the mean weights of the brain, spleen, kidneys, and ovaries were seen in
females fed borax, and an increase in mean liver weight was seen in females fed boric acid; no
organ weight changes were seen in the males. At 175 ppm boron, the only change in organ
weight reported by the investigators was increased kidney weights in males fed borax. These
changes, however, were not observed at 525 ppm boron for either compound. Microscopic
examination revealed complete testicular atrophy at 1750 ppm in all males fed borax or boric
acid, and partial testicular atrophy at 525 ppm boron in four males fed borax and in one male fed
boric acid. Changes in organ weights that were reported at 52.5 ppm were not dose related and
were not confirmed in follow-up chronic studies by the same investigators. This study identified
an NOAEL of 175 ppm boron (8.8 mg B/kg-day) and an LOAEL of 525 ppm boron (26.3 mg
B/kg-day) boron for systemic toxicity in rats following subchronic dietary exposure.
A subchronic study in rats using drinking water exposure is also available. Borax was
administered in the drinking water to male Long Evans rats (15/group) at levels of 0, 150, and
300 mg B/L for 70 days; the basal diet contained approximately 54 g B/g of feed (Seal and
Weeth, 1980). The approximate intake of boron for the treated rats was 23.7 and 44.7 mg
B/kg-day, respectively, using reference values for body weight, food, and water consumption.
Treatment with borax at both doses produced significant (p<0.05) decreases in body weight;
testis, seminal vesicle, spleen, and right femur weight; and plasma triglyceride levels. At the
highest dose level, spermatogenesis was impaired and hematocrit was decreased slightly. From
this study, an LOAEL of 23.7 mg B/kg-day can be identified. An NOAEL was not identified.
The subchronic and chronic toxicity of boron (boric acid) in mice was studied by NTP
(1987) and Dieter (1994). In the subchronic study, groups of 10 male and 10 female B6C3F1
mice were fed diets containing 0, 1200, 2500, 5000, 10,000, or 20,000 ppm boric acid (0, 210,
437, 874, 1748, or 3496 ppm boron) for 13 weeks (NTP, 1987; Dieter, 1994). These dietary
levels correspond to approximately 0, 34, 70, 141, 281, and 563 mg B/kg-day for males and 0,
47, 97, 194, 388, and 776 mg B/kg-day for females, respectively, based on reported average
values for feed consumption (161 g/kg bw/day for males, 222 g/kg bw/day for females) by
controls in week 4 of the experiment. At the highest dose level, hyperkeratosis and acanthosis of
the stomach and >60% mortality were observed. At 10,000 ppm boric acid, 10% mortality was
Proposal Draft
Boron — September, 2006 7-8
-------
observed among the males. At 5000 ppm and higher, degeneration or atrophy of the
seminiferous tubules was observed in males, and weight gain was suppressed in animals of both
sexes. Minimal to mild extramedullary hematopoiesis of the spleen was observed in all dosed
groups. The lowest dose tested, 1200 ppm (34 mg B/kg-day for male mice), appears to be the
LOAEL for this study. The NOAEL (no toxicity in absence of body weight loss) was at or
below 1200 ppm (34 mg/kg-day for males and 47 mg/kg-day for females). From this study,
dietary doses of 2500 ppm (70 mg B/kg-day for males and 97 mg B/kg-day for females) and
5000 ppm (141 mg B/kg-day for males and 194 mg B/kg-day for females) were selected to be
tested in both sexes in the chronic 2-year study based on body weight depression and mortality in
the two highest doses tested in the subchronic study.
In the chronic study, male and female (50/sex/group) B6C3F1 mice were fed a diet
containing 0, 2500, or 5000 ppm boric acid for 103 weeks (NTP, 1987; Dieter, 1994). The low-
and high-dose diets provided approximate doses of 275 and 550 mg/kg-day (48 and 96 mg B/kg-
day), respectively. Mean body weights of high-dose mice were 10-17% lower than those of
controls after 32 (males) or 52 (females) weeks. No treatment-related clinical signs were
observed throughout the study. Survival of the male mice was significantly lower than that of
the control group after week 63 in the low-dose group and after week 84 in the high-dose group.
Survival was not affected in females. At termination, the survival rates were 82, 60, and 44% in
the control, low-, and high-dose males, respectively, and 66, 66, and 74% in the control, low-,
and high-dose females, respectively. The low number of surviving males may have reduced the
sensitivity of the study for evaluation of carcinogenicity (NTP, 1987). Administration of boric
acid to male mice induced testicular atrophy and interstitial cell hyperplasia in the high-dose
group. There also were dose-related increased incidences of splenic lymphoid depletion in male
mice. According to NTP (1987), this lesion is associated with stress and debilitation and is
reflected in the increased mortality in these groups of male mice. Increased incidences of other
nonneoplastic lesions were not believed to have been caused by the administration of boric acid,
because they either were not consistently dose-related or did not occur in both sexes.
Low-dose male mice demonstrated increased incidences of hepatocellular carcinoma
(5/50, 12/50, 8/49) and combined adenoma or carcinoma (14/50, 19/50, 15/49), relative to
control and the high-dose male mice (NTP, 1987; Dieter, 1994). The increases were statistically
significant by life table tests, but not by incidental tumor tests. The incidental tumor tests were
considered to be the more appropriate form of statistical analysis in this case, because the
hepatocellular carcinomas did not appear to be the cause of death for males in this study; the
incidence of these tumor types in animals that died prior to study completion (7/30 or 23%) was
similar to the incidence at terminal sacrifice (5/20 or 25%) (NTP, 1987; Elwell, 1993). The
hepatocellular carcinoma incidence in this study was within the range of male mice historical
controls both at the study lab (131/697 or 19 ± 6%) and for NTP (424/2084 or 20 ± 7%) (NTP,
1987; Elwell, 1993). Also, the hepatocellular carcinoma incidence in the male control group of
this study (10%) was lower than the historical controls. NTP concluded that the increase in
hepatocellular tumors in low-dose male mice was not due to administration of boric acid.
There was also a significant increase in the incidence of combined subcutaneous tissue
fibromas, sarcomas, fibrosarcomas, and neurofibrosarcomas in low-dose male mice (2/50, 10/50,
2/50) by both incidental and life table pair-wise tests (NTP, 1987; Dieter, 1994). This higher
incidence of subcutaneous tissue tumors is within the historical range (as high as 15/50 or 30%)
Proposal Draft
Boron — September, 2006 7-9
-------
for these tumors in control groups of group-housed male mice from other dosed feed studies
(Elwell, 1993). The historical incidence at the study laboratory was 39/697 (6 ± 4%) and in NTP
studies was 156/2091 (7 ± 8%) (NTP, 1987). Based on the comparison to historical controls and
lack of any increase in the high-dose group, NTP concluded that the increase in subcutaneous
tumors in low-dose male mice was not compound-related. Overall, NTP concluded that this
study produced no evidence of carcinogenicity of boric acid in male or female mice, although the
low number of surviving males may have reduced the sensitivity of the study.
Schroeder and Mitchener (1975) conducted a study in which 0 or 5 ppm of boron as
sodium metaborate was administered in the drinking water to groups of 54 male and 54 female
Charles River Swiss mice (approximately 0.95 mg B/kg-day) for their life span; controls
received deionized water. In adult animals, there generally were no effects observed on
longevity body weights (at 30 days, treated animals were lighter than controls, and at 90 days,
treated males were significantly heavier than controls). The life spans of the dosed group did not
differ from controls. Gross and histopathologic examinations were performed to detect tumors.
Limited tumor incidence data were reported for other metals tested in this study, but not for
boron. Investigators reported that at this dose, boron was not tumorigenic for mice; however,
only one dose of boron (lower than other studies) was tested, and an MTD was not reached.
7.2.2 Inhalation Exposure
There are few data available regarding the toxicity of boron compounds by inhalation in
laboratory animals. Wilding et al. (1959) investigated the toxicity of boron oxide aerosols by
inhalation exposure in rats and dogs. Three dogs were exposed to 57 mg/ms (18 mg B/m3) for 23
weeks. A group of 70 albino rats, including both males and females, was exposed to an average
concentration of 77 mg/m3 of boron oxide aerosols (24 mg B/ms) for 24 weeks (6 hours/day, 5
days/week). Additional groups of rats were exposed to 175 mg/ms (54 mg B/m3) for 12 weeks
(n=4) or 470 mg/ms (146 mg B/m3) for 10 weeks (n=20) using the same exposure regimen. At
the latter concentration, the aerosol formed a dense cloud of fine particles, and the animals were
covered with dust. No clinical signs were noted, except a slight reddish exudate from the nose of
rats exposed to 470 mg/m3, which the researchers attributed to local irritation. Growth was
reduced roughly 9% in rats exposed to 470 mg/m3 compared to controls. Growth in the lower
dose rat groups and in dogs was not affected. There was a significant drop in pH and increase in
urine volume in rats exposed to 77 mg/m3. The researchers hypothesized that this was due to
formation of boric acid from boron oxide by hydration in the body and the diuretic properties of
boron oxide. There was also a significant increase in urinary creatinine at this dose. No effect
on serum chemistry, hematology, organ weights, histopathology, bone strength, or liver function
was found in either rats or dogs (not all endpoints were studied in all exposure groups).
Proposal Draft
Boron — September, 2006 7-10
-------
7.3 Developmental/Reproductive Toxicity
7.3.1 Developmental Studies
Heindel et al. (1994, 1992) and Price et al. (1990) treated timed-mated Sprague-Dawley
rats (29/group) with a diet containing 0, 0.1, 0.2, or 0.4% boric acid from gestation day (gd)
0-20. The investigators estimated that the diet provided 0, 78, 163, or 330 mg boric acid/kg-day
(0, 13.6, 28.5, or 57.7 mg B/kg-day). Additional groups of 14 rats each received boric acid at 0
or 0.8% in the diet (539 mg/kg-day or 94.2 mg B/kg-day) on gd 6 through 15 only. Exposure to
0.8% was limited to the period of major organogenesis in order to reduce the preimplantation
loss and early embryolethality indicated by the range-finding study and, hence, provide more
opportunity for teratogenesis. (The range-finding study found that exposure to 0.8% on gd 0-20
resulted in a decreased pregnancy rate [75% as compared with 87.5% in controls] and in greatly
increased resorption rate per litter [76% as compared with 7% in the control group].) Food and
water intake and body weights, as well as clinical signs of toxicity, were monitored throughout
pregnancy. On gd 20, the animals were sacrificed; the liver, kidneys, and intact uteri were
weighed; and corpora lutea were counted. Maternal kidneys, selected randomly (10
dams/group), were processed for microscopic evaluation. Live fetuses were dissected from the
uterus, weighed and examined for external, visceral, and skeletal malformations. Statistical
significance was established at p<0.05. There was no maternal mortality during treatment. Food
intake increased 5-7% relative to that of controls on gd!2-20 at 0.2 and 0.4%; water intake was
not significantly altered by administration of boric acid (data not shown). At 0.8%, water and
food intake decreased on gd 6-9 and increased on gdl5-18, relative to controls. Pregnancy rates
ranged between 90 and 100% for all groups of rats and appeared unrelated to treatment.
Maternal effects attributed to treatment included a significant and dose-related increase in
relative liver and kidney weights at 0.2% or more, a significant increase in absolute kidney
weight at 0.8%, and a significant decrease in body-weight gain during treatment at 0.4% or
more. Corrected body weight gain (gestational weight gain minus gravid uterine weight) was
unaffected except for a significant increase at 0.4%. Examination of maternal kidney sections
revealed minimal nephropathy in a few rats (unspecified number), but neither the incidence nor
the severity of the changes was dose related.
Treatment with 0.8% boric acid (gd 6-15) significantly increased prenatal mortality, as
seen in increases in the percentage of both resorptions and late fetal deaths per litter. The
number of live fetuses per litter was also significantly decreased at 0.8%. Average fetal body
weight (all fetuses or male or female fetuses) per litter was significantly reduced in all treated
groups versus controls in a dose-related manner. Mean fetal weights were 94, 87, 63, and 46%
of the corresponding control means for the 0.1, 0.2, 0.4, and 0.8%, respectively. The percentage
of malformed fetuses per litter and the percentage of litters with at least one malformed fetus
were significantly increased at 0.2% or more. Treatment with 0.2% or more boric acid also
increased the incidence of litters with one or more fetuses with a skeletal malformation. The
incidence of litters with one or more pups with a visceral or gross malformation was increased at
0.4 and 0.8%, respectively. The malformations consisted primarily of anomalies of the eyes, the
central nervous system (CNS), the cardiovascular system, and the axial skeleton. In the 0.4 and
0.8% groups, the most common malformations were enlarged lateral ventricles of the brain and
agenesis or shortening of rib XIII. The percentage of fetuses with variations per litter was
reduced relative to controls in the 0.1 and 0.2% dosage groups (due primarily to a reduction in
Proposal Draft
Boron — September, 2006 7-11
-------
the incidence of rudimentary or full ribs at lumbar I), but was significantly increased in the 0.8%
group. The variation with the highest incidence among fetuses was wavy ribs. Based on the
changes in organ weights, a maternal LOAEL of 0.2% boric acid in the feed (28.5 mg B/kg-day)
can be established; the maternal NOAEL is 0.1% or 13.6 mg B/kg-day. Based on the decrease in
fetal body weight per litter, the level of 0.1% boric acid in the feed (13.6 mg B/kg-day) is an
LOAEL; an NOAEL was not defined.
In a follow-up study, Price et al. (1996a, 1994) administered boric acid in the diet (at 0,
0.025, 0.050, 0.075, 0.100, or 0.200%) to timed-mated CD rats, 60 per group, from gd 0-20.
Throughout gestation, rats were monitored for body weight, clinical condition, and food and
water intake. This experiment was conducted in two phases, and in both phases offspring were
evaluated for post-implantation mortality, body weight, and morphology (external, visceral, and
skeletal). Phase I of this experiment was considered the teratology evaluation and was
terminated on gd 20, when uterine contents were evaluated. The calculated average dose of
boric acid consumed for Phase 1 dams was 19, 36, 55, 76, and 143 mg/kg-day (3.3, 6.3, 9.6, 13.3,
and 25 mg B/kg-day). During Phase I, no maternal deaths occurred, and no clinical symptoms
were associated with boric acid exposure. Maternal body weights did not differ among groups
during gestation, but statistically-significant trend tests associated with decreased maternal body
weight (gd 19 and 20 at sacrifice) and decreased maternal body weight gain (gd 15-18 and gd
0-20) were indicated. In the high-dose group, there was a 10% reduction (statistically significant
in the trend test p<0.05) in gravid uterine weight when compared with controls. The authors
indicated that the decreasing trend of maternal body weight and weight gain during late gestation
reflected reduced gravid uterine weight. Corrected maternal weight gain (maternal gestational
weight gain minus gravid uterine weight) was not affected. Maternal food intake was only
minimally affected at the highest dose and only during the first 3 days of dosing. Water intake
was higher in the exposed groups after gd 15. The number of ovarian corpora lutea and uterine
implantation sites, and the percentage of preimplantation loss were not affected by boric acid
exposure.
Offspring body weights were significantly decreased in the 13.3 and 25 mg B/kg-day
dose groups on gd 20. The body weights of the low- to high-dose groups, respectively, were 99,
98, 97, 94, and 88% of control weight. There was no evidence of a treatment-related increase in
the incidence of external or visceral malformations or variations when considered collectively or
individually. On gd 20, skeletal malformations or variations considered collectively showed a
significant increased percentage of fetuses with skeletal malformations per litter. Taken
individually, dose-related response increases were observed for short rib XIII, considered a
malformation in this study, and wavy rib or wavy rib cartilage, considered a variation. Statistical
analyses indicated that the incidence of short rib XIII and wavy rib were both increased in the
13.3 and 25 mg B/kg-day dose groups relative to controls. A statistically significant trend
(p<0.05) was found for decreases in rudimentary extra rib on lumbar I, classified as a variation.
Only the high-dose group had a biologically relevant, but not statistically significant, decrease in
this variation. The LOAEL for Phase I of this study was considered to be 0.1% boric acid (13.3
mg B/kg-day), based on decreased fetal body weight. The NOAEL for Phase I of this study was
considered to be 0.075% boric acid (9.6 mg B/kg-day).
In Phase II, dams were allowed to deliver and rear their litters until postnatal day (pnd)
21. The calculated average doses of boric acid consumed for Phase II dams were 19, 37, 56, 74,
Proposal Draft
Boron — September, 2006 7-12
-------
and 145 mg/kg-day (3.2, 6.5, 9.7, 12.9, and 25.3 mg B/kg-day). This phase allowed a follow-up
period to determine whether the incidence of skeletal defects in control and exposed pups
changed during the first 21 postnatal days. Among live born pups, there was a significant trend
test for increased number and percentage of dead pups between pnd 0 and 4, but not between pnd
4 and 21; this appeared to be due to an increase in early postnatal mortality in the high dose,
which did not differ significantly from controls and was within the range of control values for
other studies in this laboratory. On pnd 0, the start of Phase II, there were no effects of boric
acid on the body weight of offspring (102, 101, 99, 101, and 100% of controls, respectively).
There were also no differences through termination on pnd 21; therefore, fetal body weight
deficits did not continue into this postnatal period (Phase II). The percentage of pups per litter
with short rib XIII was still elevated on pnd 21 in the 0.200% boric acid dose group (25.3 mg
B/kg-day), but there was no incidence of wavy rib, and none of the treated or control pups on
pnd 21 had an extra rib on lumbar 1. The NOAEL and LOAEL for phase II of this study were
12.9 and 25.3 mg B/kg-day, respectively.
Price et al. (1997) provides an analysis of maternal whole blood taken on gd 20 from the
previously described study (Price et al., 1996a, 1994) in which dietary concentration of added
boric acid yielded average daily intakes equivalent to 0, 3, 6, 10, 13, or 25 mg B/kg body weight.
Blood samples were analyzed using inductively coupled plasma optical emission spectrometry.
Increasing dietary concentrations of boric acid were positively associated with whole blood
concentration in pregnant rats. Whole blood concentrations in confirmed pregnant rats were
0.229 ± 0.143, 0.564 ± 0.211, 0.975 ± 0.261, 1.27 ± 0.298, 1.53 ± 0.546, 2.82 ± 0.987 ug boron/g
whole blood (mean ± SD) for the control through the high-dose groups. Positive correlations
between maternal blood boron concentrations and indices of maternal dietary intake of boron
with embryo/fetal toxicity (Price et al., 1996a, 1994) were observed at average daily
concentration of 13 and 25 mg B/kg. Blood boron concentrations of 1.27 ± 0.298 and 1.53 ±
0.546 ug boron/g were associated with the NOAEL (10 mg B/kg-day) and the LOAEL (13 mg
B/kg-day) for the developmental toxicity reported in Price et al. (1996a, 1994).
The developmental effects of boric acid also have been studied in mice and rabbits.
Heindel et al. (1994, 1992) and Field et al. (1989) examined the developmental effects of boric
acid in pregnant CD-I mice using the same experimental design as in the initial study with rats
(Price et al., 1990), except that a 0.8% dietary level was not used in the mouse study. The diets
containing 0, 0.1, 0.2, or 0.4% boric acid were estimated by the investigators to provide 0, 248,
452, or 1003 mg boric acid/kg-day (0, 43.4, 79.0, or 175.3 mg B/kg-day); the mice were treated
during gd 0-17. Neither survival rates nor pregnancy rates were affected by treatment with boric
acid. Pale kidneys were noted in several treated dams, particularly in the high-dose group, and
one dam in this group had fluid accumulation in the kidney. Maternal body weight was
significantly reduced by 10-15% during gd 12-17 in the high-dose group. Maternal weight gain
was significantly reduced during treatment in the high-dose group, but was not affected when
corrected for gravid uterine weight. At the 0.4% dietary level, food intake was increased
between days 12 and 15 and water intake was increased on days 15-17 (statistical significance
not provided for either effect). Organ weight changes were limited to significant increases in
relative kidney weight and absolute liver weight in the 0.4% groups. A dose-related increase in
maternal renal tubular dilation and/or regeneration was observed; the incidence was 0/10, 2/10,
8/10, and 10/10 in the 0, 0.1, 0.2, and 0.4% dosage groups, respectively. Treatment with boric
acid did not affect preimplantation loss or the number of implantation sites per litter, but
Proposal Draft
Boron — September, 2006 7-13
-------
significantly increased the percentage of resorptions per litter and the percentage of litters with
one or more resorptions at the 0.4% level. There was a significant dose-related decrease in
average fetal body weight (all fetuses or male or female fetuses) per litter at 0.2% or more. The
percentage of malformed fetuses per litter increased significantly at 0.4%, whereas the
percentage of fetuses with variations per litter was decreased at 0.1 and 0.2% and was not
affected at 0.4%. The most frequent malformation observed among fetuses of the 0.4% group
was a short rib XIII. In contrast, full or rudimentary lumbar I rib (a variation) was less frequent
in fetuses of treated mice. Although the level of 0.1% boric acid in the diet induced an increase
in renal lesions in mice, the increased incidence did not achieve statistical significance (Fisher
Exact Test). The 0.1% level (43.4 mg B/kg-day) is a maternal NOAEL and the 0.2% level (79
mg B/kg-day) is a maternal LOAEL. For developmental effects, the 0.2% dietary level of boric
acid is an LOAEL based on decreased fetal body weight per litter, and the 0.1% level is an
NOAEL.
Artificially inseminated New Zealand White rabbits (30/group) were administered 0,
62.5, 125, or 250 mg boric acid/kg-day (0, 10.9, 21.9, and 43.7 mg B/kg-day) in aqueous
solution by gavage on gd 6-19 (Price et al., 1996b, 1991; Heindel et al., 1994). Food
consumption, body weight and clinical signs were monitored throughout the study. At gd 30, the
animals were sacrificed and the following endpoints were examined: pregnancy status; number
of resorptions; fetal body weight; viability; and external, visceral, and skeletal malformations.
No treatment-related clinical signs of toxicity were observed during the study, except for vaginal
bleeding noted in 2-11 does/day on gd 19-30 at the high dose; these does had no live fetuses on
day 30. Vaginal bleeding was also observed in one female in the low-dose group and in one in
the mid-dose group. Two maternal deaths occurred (one each at the low- and mid-dose), but
were not treatment-related. Food intake was decreased relative to that of controls on treatment
days 6-15 at the high dose, and was increased after treatment ceased on days 25-30 at the mid
and high doses. Body weight on gd 9-30, weight gain on gd 6-19, gravid uterine weight, and
number of corpora lutea per dam were each decreased in the high-dose group. After correction
for gravid uterine weights, however, maternal body-weight gain was increased at both the mid
and high doses. Treatment with boric acid did not affect absolute or relative liver weight.
Relative, but not absolute kidney weight increased at the high dose; kidney histopathology was
unremarkable. Boric acid caused frank developmental effects at the high dose. These effects
consisted of a high rate of prenatal mortality (90% of implants/litter were reabsorbed compared
with 6% in the control group). Also, the percentage of pregnant females with no live fetuses was
greatly increased (73% compared with 0% in controls), whereas the number of live fetuses per
litter on day 30 was significantly reduced (2.3/litter compared with 8.8/litter in the control
group). Malformed live fetuses per litter increased significantly at the high dose, primarily due
to the incidence of fetuses with cardiovascular defects, the most prevalent of which was
interventricular septal defect (8/14 at the high dose compared with 1/159 in the control group).
The incidence of skeletal malformations was comparable among groups. Relative to controls,
the percentage of fetuses with variations (all types combined) was not significantly increased in
any treated group, but the percentage with cardiovascular variations was significantly increased
from 11% in controls to 64% in the high-dose group. Fetal body weights per litter at the high
dose were depressed relative to the control, but the difference was not statistically significant;
however, this could have been due to the small sample size in the high-dose group. No
developmental effects were found in the low- and mid-dose groups. In this study, the mid dose
of 125 mg boric acid/kg-day (21.9 mg B/kg-day) represents the NOAEL based on maternal and
Proposal Draft
Boron — September, 2006 7-14
-------
developmental effects. The high dose of 250 mg boric acid/kg-day (43.7 mg B/kg-day) is the
LOAEL.
Narotsky et al. (2003) dosed rat dams (number not specified) with 500 mg/kg boric acid
twice daily on single days during development (gd 6, 7, 8, 9, 10, or 11) and examined fetal body
weight and skeletal malformations. These were compared to the effect of boric acid on the hox
gene family, genes clustered among four loci and thought to confer positioning and development
of vertebrae. Their expression in the paraxial mesoderm begins during gastrulation. Boric acid
(0 or 500 mg/kg) was administered via gavage to pregnant Sprague-Dawley rats twice daily
(totaling 1000 mg/kg-day) on gd 6, 7, 8, 9, 10, or 11, and examinations were performed on gd
21. Skeletal malformations were evaluated following alizarin red and alcian blue staining. Boric
acid was administered on gd 9, and hox gene expression was determined by in situ hybridization
in fixed sections at gd 13.5. Fetal weights were significantly decreased in animals treated on gd
7, 9, 10, or 11. Fetuses exposed on gd 8 or 9 demonstrated a "low but significant" elevation of
the frequency of rudimentary cervical ribs. The authors indicate that fetuses exposed on gd 6, 7,
8, or 11 generally demonstrated "no such effect" of boric acid on ribs, vertebrae, and sternebrae
compared with the striking alterations observed following treatment on gd 9. The
cephalo-caudal expression pattern of the hoxc6 and hoxa6 genes in pre-vertebral tissues was
altered by boric acid treatment on gd 9. These authors demonstrated that exposure on gd 6 "had
no developmental effects, and treatment on gd 7 and 11 caused only relatively mild
developmental toxicity (reduced fetal weights but did not alter the frequency or type of skeletal
malformations); treatment on gd 8, 9, and 10 disrupted axial development. Gestational day 8
exposure induced cervical ribs and rib or vertebral malformations, but only treatment on gd 9 or
10 dramatically altered numbers of vertebrae, ribs or sternebrae."
Cherrington and Chernoff (2002) evaluated the developmental toxicity of boric acid in
pregnant CD-I mice in three separate experimental designs. In the first design, mice were dosed
daily from gd 6-10 by gavage with either 0, 500, or 750 mg/kg. The control group had 6
animals, while the 500 and 750 mg/kg boric acid-dosed groups contained 10 animals each. The
second exposure scenario consisted of 160 timed pregnant animals weighed on gd 6 and assigned
to 1 of 10 groups: controls treated on each of gd 6-8 (one group); controls treated only on a
single gd 6, 7, or 8 (three groups); and groups of dams treated with a gavage dose of 400 mg/kg
twice daily (total dose 800 mg/kg-day) on each of gd 6-8 (one group) or only on a single gd 6, 7,
8, 9, 10 (four groups). The third exposure regimen consisted of either a single or two gavage
doses of 750 mg/kg each on gd 8. In the group with a single gavage dose on gd 8, 52 pups from
four control litters and 33 pups from three boric acid-dosed litters were examined. For the group
with two gavage doses of 750 mg/kg each on gd 8, 103 controls and 94 boric acid-treated fetuses
were examined, weighed, and stained with alizarin red and alcian blue for skeletal evaluation on
gd!7.
Results from the first experiment indicated that 400 mg/kg daily doses resulted in
decreased rib length, and daily doses of 750 mg/kg resulted in decreased rib length and femur
length. Fetal body weight was not significantly decreased at either dose. In the second study,
the results for the gd 9 and 10 daily exposures were not presented due to the lack of a concurrent
control. Fetal body weight was reduced in all boric acid treatment groups (single days gd 6, 7, 8,
9, or 10 and consecutive days from gd 6 to 8). Femur length was decreased on gd 7 and in
fetuses exposed for the gd 6-8 period. Cervical ribs were observed in fetuses exposed on gd 6, 7,
Proposal Draft
Boron — September, 2006 7-15
-------
or 8. Results from the third experiment indicated that the two doses of 750 mg/kg each on gd 8
significantly increased frequency of 11 separate malformations over background incidence. The
most prevalent malformations were those associated with rib development. In contrast, the
single dose on this day produced only increased incidences of unilateral thoracic vertebrae and
cervical rib formation/ossification differences. These results demonstrate a separation of the
effects of boric acid on fetal body weight and rib malformation with respect to the timing of the
dose. The authors concluded that the accumulation of the effect, rather than the accumulation of
boric acid, was responsible for the temporal dependence of boric acid-induced fetotoxicity, citing
a rapid clearance of borates from the blood. They specifically indicated that, "because of boric
acid's short half-life, these data suggest that these earlier processes, gastrulation and presomitic
mesoderm formation and patterning, are the processes boric acid is affecting."
To examine the molecular basis for boric acid's effect on axial skeletal development,
Wery et al. (2003) dosed pregnant Sprague-Dawley rats (animal number not given) with two
separate gavage doses of 500 mg/kg each on gd 9 and sacrificed the dams on gd 11 or gd 13.5.
Embryos were removed and fixed for in situ hybridization to ascertain the distribution of several
hox genes. These genes show a distinct pattern of expression among the Semites responsible for
the cranial-caudal development of the axial skeleton (vertebrae, ribs). Following boric acid
administration on gd 9, the anatomic boundary for expression ofhoxd4, hoxa4, hoxc5, and hoxc6
were altered when assessed on gd 11. When assessed on gd 13.5, the boundary for expression of
hoxd4, hoxa4, hoxa5, and hoxcS was not altered, while the boundary for hoxa6 was altered. The
authors concluded that the nature and exposure timing-dependency of the skeletal malformations
support a role for hox gene alteration in the mechanism of boric acid-induced axial skeletal
malformations.
7.3.2 Reproductive Studies
7.3.2.1 Male-Only Exposure
Studies of subchronic and chronic toxicity of boron compounds in dogs, rats, and mice
have identified the testes as a primary target organ in males of these species (e.g., Weir and
Fisher, 1972; NTP, 1987). These studies were described in Section 4.2.1. Several other studies
have been conducted to investigate the effects of boron compounds on male reproductive
performance and testicular morphology in more detail.
Dixon et al. (1976) studied the effects of borax on reproduction in male rats following
acute and subchronic exposure. In the acute study, groups of 10 adult male Sprague-Dawley rats
were given single oral doses of borax at 0, 45, 150, and 450 mg B/kg. Fertility was assessed by
serial mating trials in which each male was mated with a series of untreated virgin females in
sequential 7-day periods (for up to 70 days). The females were sacrificed 9 days after the end of
their breeding periods (when they would be 9-16 days pregnant), and uteri and fetuses were
examined. Male rats were sacrificed on days 1 and 7, and at subsequent 7-day intervals for
histopathological examination of the testes. No effect on male fertility was found at any dose in
this study. Testicular lesions were not reported. This study found an NOAEL of 450 mg B/kg
for reproductive effects in male rats following single-dose oral exposure.
Proposal Draft
Boron — September, 2006 7-16
-------
In the subchronic study, male Sprague-Dawley rats (10/group) were given 0, 0.3, 1.0, or
6.0 mg B/L, as borax, in the drinking water for 30, 60, or 90 days (Dixon et al., 1976). The
investigators estimated the highest exposure level provided 0.84 mg B/kg-day. Based on this
estimate, the lower two levels provided 0.042 and 0.14 mg B/kg-day. There were no noticeable
reproductive effects or changes in serum chemistry; plasma levels of follicle stimulating
hormone (FSH) and luteinizing hormone (LH); or weight of the body, testes, prostate or seminal
vesicles. Fructose, zinc and acid phosphatase levels in the prostate were unchanged. Breeding
studies revealed no effects on male fertility. Therefore, the dose of 0.84 mg B/kg-day, the
highest dose tested, represents an NOAEL for this study.
In a follow-up study reported by Dixon et al. (1979) and Lee et al. (1978), diets
containing 0, 500, 1000, or 2000 ppm boron, as borax, were administered to male
Sprague-Dawley rats (18/group) for 30 or 60 days (approximately 0, 25, 50, or 100 mg
B/kg-day). Significant (p<0.05) decreases in the weight of liver, testes, and epididymis were
observed at the 1000 and 2000 ppm dietary levels. Seminiferous tubule diameter was
significantly (p<0.05) decreased in a dose-dependent manner in all treatment groups; however,
significant loss of germinal cell elements was observed only at the 1000 and 2000 ppm dietary
levels. Aplasia was complete at the highest dose. Plasma levels of the hormone FSH were
significantly (p<0.05) elevated in a dose- and duration-related manner at all dose levels, while
plasma LH and testosterone levels were not affected significantly. Serial mating studies revealed
reduced fertility without change in copulatory behavior at the two higher dose levels. Based on
dose-related tubular germinal aplasia, which is reversible at low doses, this study defines an
LOAEL of 50 mg B/kg-day and an NOAEL of 25 mg B/kg-day.
Linder et al. (1990) examined the time- and dose-response of male rat reproductive
endpoints after acute administration of boric acid. In the time-response experiment,
Sprague-Dawley rats (6/group) were given 0 or 2000 mg boric acid/kg bw (0 or 350 mg B/kg,
respectively) by gavage and were sacrificed at 2, 14, 28, and 57 days after dosing. In the
dose-response experiment, groups of eight male rats were administered 0, 250, 500, 1000, or
2000 mg boric acid/kg (0, 44, 87, 175, or 350 mg B/kg) by gavage and were sacrificed 14 days
later. In both the time-response and the dose-response studies, the above doses are the totals of
two doses administered at 9:00 a.m. and 4:00 p.m. on the same day. No significant clinical signs
of toxicity were observed during the study. Histopathologic examinations of the testes and
epididymis revealed adverse effects on spermiation, epididymal sperm morphology, and caput
sperm reserves. The testicular effects, apparent at 14 days, included enlarged irregular
cytoplasmic lobes of Step 19 spermatids in stage VIII seminiferous tubules and retention of Step
19 spermatids in stage IX-XIII tubules at the 175 and 350 mg B/kg dose levels. There was also a
substantial increase (p<0.05) in the testicular sperm head count per testis and per g testis in the
350 mg/kg time-response group. Epididymal effects, also apparent at 14 days, included an
increase in abnormal caput epididymal sperm morphology (percentage with head or tail defects,
p<0.05) and reduced caput epididymal sperm reserves (p<0.05). In the day 28 time-response
group (350 mg B/kg), significant effects (p<0.05) included an increase in abnormal caput and
cauda epididymal sperm morphology and a decreased percentage of motile cauda spermatozoa
with reduced straight-line swimming velocities. Substantial recovery occurred by day 57. This
study described an LOAEL for male reproductive effects of 175 mg B/kg bw and an NOAEL of
87 mg B/kg bw following acute oral exposure in rats.
Proposal Draft
Boron — September, 2006 7-17
-------
Treinen and Chapin (1991) examined the development and progression of reproductive
lesions in 36 mature male F344 rats treated with boric acid in the diet for 4-28 days. Thirty
animals served as controls. Boric acid was added to the feed at a level of 9000 ppm. Based on
food consumption and body weight data, the investigators estimated that over the 28-day period
the mean intake of boric acid was 348.3 mg/kg-day, or 60.9 mg B/kg-day. Sacrifices were
conducted at 4, 7, 10, 14, 21, and 28 days on six treated and four control animals per time point.
Liver, kidney, and testicular histology; serum testosterone; androgen binding protein (ABP)
levels; and tissue boron levels were assessed. In half of the treated rats, there was inhibition of
spermiation in 10-30% of stage-IX tubules at 7 days. Inhibited spermiation was observed in all
stage-IX and stage-X tubules of exposed rats at 10 days. Advanced epithelial disorganization,
cell exfoliation, luminal occlusion, and cell death were observed after 28 days, causing
significant loss of spermatocytes and spermatids from all tubules in exposed rats. Throughout
the study, specific lesions became more severe with increasing duration of exposure. Treatment
with boric acid had no effect on kidney and liver histology. In treated rats, basal serum
testosterone levels were significantly decreased (p<0.05) from 4 days on, but serum testosterone
levels stimulated by human chorionic gonadotropin or luteinizing hormone releasing factor were
not affected. Steady-state levels of boron were reached in tissues by 4 days of treatment, and
there was no selective accumulation of boron in blood, epididymis, liver, or kidney. After 4 days
of treatment with boric acid, serum ABP levels were significantly reduced relative to controls;
however, this difference disappeared by day 7.
Ku et al. (1993a) and Chapin and Ku (1994) compared testis boron dosimetry to lesion
development. Rats were fed 0, 3000, 4500, 6000, or 9000 ppm boric acid (0, 545, 788, 1050, or
1575 ppm boron) for up to 9 weeks and examined. Based on food intake and body weight data,
the researchers estimated the daily intake of boron as <0.2, 26, 38, 52, or 68 mg B/kg-day. At 32
weeks post-treatment, recovery was assessed. Inhibited spermiation occurred at 3000 and 4500
ppm, and atrophy occurred at 6000 and 9000 ppm. A mean testis boron level of 5.6 g B/g of
tissue was associated with inhibited spermiation, whereas 11.9 g B/g was associated with
atrophy, with no boron accumulation during the 9-week exposure. This suggests that separate
mechanisms may be operating for these effects based on testis boron concentration. Severely
inhibited spermiation at 4500 ppm was resolved by 16 weeks post-treatment, but some areas of
focal atrophy in the 6000 and 9000 ppm dose groups did not change post-treatment. The low
dose of 26 mg B/kg-day was an LOAEL in this study.
Following in vitro boric acid exposure, Ku et al. (1993b) evaluated endpoints in the cell
culture system that suggest that boric acid has an effect on DNA synthesis that occurred at
concentrations associated with atrophy in vivo, and suggests that boric acid interferes with the
production and maturation of early germ cells.
Ku and Chapin (1994) showed that testicular atrophy and CNS hormonal effects were not
due to selective accumulation in testis or brain/hypothalamus with boron testis concentrations of
1-2 mM. In vitro studies addressed boric acid testicular toxicity: mild hormone effect, the initial
inhibited spermiation, and atrophy. No effect of boric acid on the steroidogenic function of
isolated Leydig cells was observed, supporting the suggestion of a CNS mediated hormonal
effect. The authors found that inhibited spermiation was not due to increased testicular cyclic
adenosine monophosphate (cAMP) or reduced serine proteases plasminogen activators (PA).
Boric acid effects were evaluated in Sertoli-germ cell co-cultures on Sertoli cell energy
Proposal Draft
Boron — September, 2006 7-18
-------
metabolism (lactate secreted by Sertoli cells is a preferred energy source for germ cells) and
DNA/RNA syntheses (germ cells synthesize DNA/RNA and boric acid impairs this nucleic acid
in the liver). The most sensitive in vitro endpoint was DNA synthesis of mitotic/meiotic germ
cells (with energy metabolism in germ cells affected to a lesser extent), which was manifested in
vivo as a decrease in early germ cell/Sertoli cell ratio prior to atrophy in the testes.
Naghii and Samman (1996) administered boric acid in deionized drinking water to adult
male Sprague Dawley rats (10 per group) at 2, 12.5, and 25 mg B/day for up to 6 weeks. Plasma
testosterone levels increased in rats fed 2 mg B/day, but increasing boron dose from 2 mg to 12.5
and 25 mg resulted in lower plasma testosterone concentrations which tended to rebound at 6
weeks of treatment. The response tended to be greater after 6 weeks compared to 3 weeks.
Similarly testicular testosterone concentrations also decreased with increasing boron dose, but
the difference between weeks 3 and 6 was more marked. The authors suggested that Leydig
cells, which are responsible for production of testosterone, are intact in rats fed 25 mg B in spite
of testicular atrophy. The authors also stated that these results are consistent with Weir and
Fisher (1972) who found testicular histopathology in rats fed 23-30 mg B/day for 90 days and
atrophy when boron concentration in the testes was greater than 20 ppm.
Naghii and Samman (1997) studied the specificity of the effect of boron on steroid
hormones and the impact of plasma lipids in eight male volunteers whose diets were
supplemented withlO mg B per day for 4 weeks. Plasma total cholesterol, triglyceride
concentrations, or distribution among LDL and HDL fractions were not altered. The mean total
plasma testosterone concentration increased after 4 weeks of supplementation, but this increase
was not statistically significant. The mean plasma 17Bp-estradiol concentration increased
significantly, and the ratio of 17p-estradiol to testosterone increased significantly after
supplementation.
7.3.2.2 Male and Female Exposure
In a multigeneration study, Weir and Fisher (1972) administered 0, 117, 350, or 1170
ppm boron (approximately 0, 5.9, 17.5, or 58.5 mg B/kg-day) as borax or boric acid in the diet to
groups of 8 male and 16 female Sprague-Dawley rats. No adverse effects on reproduction or
gross pathology were observed in the rats dosed with 5.9 or 17.5 mg B/kg-day that were
examined to the F3 generation. Litter size, weights of progeny, and appearance were normal
when compared with controls. The test groups receiving 58.5 mg B/kg-day boron from either
compound were found to be sterile. In these groups, males showed lack of spermatozoa in
atrophied testes, and females showed decreased ovulation in the majority of the ovaries
examined. An attempt to obtain litters by mating the treated females with the males fed only the
control diet was not successful. An LOAEL of 58.5 mg B/kg-day and an NOAEL of 17.5 mg
B/kg-day were identified from this study.
Fail et al. (1990, 1991) examined the effects of boric acid in Swiss CD-I mice in a
reproductive study using a continuous breeding protocol. Male and female F0 mice (11 weeks
old) were fed a diet containing 0, 1000, 4500, or 9000 ppm boric acid for up to 27 weeks. There
were 40 pairs in the control group and 20 pairs per treatment group. Based on an average food
consumption of 5 g/mouse and on body weights, the authors predicted the diet would provide
boric acid at 152 mg/kg-day (26.6 mg B/kg-day) to males and 182 mg/kg-day (31.8 mg
Proposal Draft
Boron — September, 2006 7-19
-------
B/kg-day) to females in the 1000 ppm group; 636 mg/kg-day (111 mg B/kg-day) to males and
868 mg/kg-day (152 mg B/kg-day) to females in the 4500 ppm group; and 1260 mg/kg-day (220
mg B/kg-day) to males and 1470 mg/kg-day (257 mg B/kg-day) to females in the 9000 ppm
group. According to the authors, actual boric acid consumption during the study did not differ
from the predicted consumption by more than 12%. Following 1 week of treatment, the F0 mice
were caged as breeding pairs for 14 weeks. During weeks 2-18, the average body weight gain of
high-dose males and females was significantly reduced relative to controls. Mortality rates in
the treated groups over the 27 weeks were not significantly different from controls. Treatment
with boric acid significantly impaired fertility. None of the 9000 ppm pairs were fertile. The
number of litters per pair, number of live pups per litter, proportion of pups born alive, live pup
weight, and adjusted pup weight (adjusted for litter size) were significantly (p<0.05) decreased at
the 4500 ppm level. The initial fertility index (percentage of cohabited pairs having at least one
litter) was not significantly altered in the 1000 and 4500 ppm groups, but the progressive fertility
index (percentage of fertile pairs that produced four litters) was decreased relative to controls in
the 4500 ppm group. The trend toward a lower fertility index at 4500 ppm started with the first
mating and progressed in severity with subsequent matings.
To determine the affected sex, the control and 4500 ppm F0 mice were then assigned to
three crossover mating groups: control male x control female, 4500 ppm male x control female,
and control male x 4500 ppm female. Each group was composed of 19-20 pairs that were mated
for 7 days or until a copulatory plug was detected, whichever occurred first; control feed was
provided for all mice during this week, followed by a resumption of the same diets they had
received previously. Mating and fertility indices were significantly depressed in the 4500 ppm
male x control female group, and only one pair in that group produced a live litter; these indices
were not affected in the control male x 4500 ppm female group. Dosed females mated to control
males had a lower body weight on pnd 0, had a longer gestational period than control groups and
gave birth to pups with decreased litter-adjusted weight. After completion of the crossover
mating trial (total of 27 weeks on test), a necropsy was performed on control and 4500 ppm F0
males and females and on 1000 and 9000 ppm F0 males that had been maintained on their
respective diets to allow a comparison of semen parameters and testicular histology among all
four treatment groups. Males treated with 9000 ppm boric acid had significantly reduced body,
testis and epididymal weights. In the 4500 ppm males, body weight was not affected, but testis,
epididymal, and prostate weights were reduced; these parameters were not altered in the 1000
ppm males. Significant reductions in sperm motility were observed in the 1000 and 4500 ppm
groups and in sperm concentration in the 4500 and 9000 ppm groups. The percentage of
abnormal sperm was significantly increased in the 4500 ppm group. Sperm motility and
morphology could not be fully evaluated in the 9000 ppm group due to absence of sperm (in 12
of 15 observed males) or severe reduction in sperm counts (in the other 3 males) of this group.
Seminiferous tubular atrophy occurred in mid-and high-dose males; the severity was
dose-related. Tissues of low-dose males exhibited no significant changes. Other indices of
testicular morphology (spermatogenic index, seminiferous tubule diameter, spermatids per testis)
also were altered at 4500 ppm or more. Effects observed at necropsy in 4500 ppm females (1000
and 9000 ppm females were not examined) were limited to a reduction in both relative and
absolute liver weights and absolute kidney plus adrenal weights in comparison with controls.
The final Fx litters (exposed during gestation and lactation) from the continuous breeding
experiment were fed the same dosage of boric acid in the diet as their parents had received.
Proposal Draft
Boron — September, 2006 7-20
-------
Because there were no litters at 9000 ppm and few of the mice born alive in the final litters at
4500 ppm survived through weaning, only the 0 and 1000 ppm Fx mice were included in a
fertility trial. The Fx mice were cohabited in nonsibling pairs (40 pairs of 0 ppm and 20 pairs of
1000 ppm mice) for 7 days or until a copulatory plug was observed, whichever occurred first.
They were maintained on their respective diets during mating and until the F2 litters were
delivered, and then were necropsied. The fertility of the 1000 ppm Fx mice was not affected, but
the litter-adjusted body weights of the F2 pups (females and combined males and females) were
significantly decreased relative to controls. Effects in 1000 ppm Fx females were significant
increases in uterine and kidney plus adrenal weights, significantly shorter estrous cycles, and
fewer ambiguous vaginal smears. A reduction in epididymal sperm concentration in the 1000
ppm Fj males approached significance (p=0.053); sperm motility and morphology were not
affected. Histopathologic examination was unremarkable. The lowest dose tested, 1000 ppm,
decreased sperm motility in the F0 males, marginally decreased epididymal sperm concentration
in Fj males, increased uterine and kidney/adrenal weights and shortened estrus cycles in Fx
females, and reduced litter-adjusted birth weights in the F2 pups. Hence, the LOAEL for this
study is 1000 ppm boric acid (26.6 and 31.8 mg B/kg-day for males and females, respectively).
An NOAEL was not identified.
7.4 Other Studies
7.4.1 Genotoxicity Studies
Results of most short-term mutagenicity studies indicate that boron is not genotoxic. In
the streptomycin-dependent Escherichia coli Sd-4 assay, boric acid was either not mutagenic
(Iyer and Szybalski, 1958; Szybalski, 1958) or produced equivocal results (Demerec et al.,
1951). In Salmonella typhimurium strains TA1535, TA1537, TA98, and TA100, boric acid was
not mutagenic in the presence or absence of either a rat or hamster liver S-9 activating system
(Benson et al., 1984; Haworth et al., 1983; NTP, 1987). Boric acid (concentration, stability, and
purity not tested by investigators) was also negative for mutagenicity in the Salmonella
microsome assay using strains TA1535, TA1537, TA1538, TA98, and TA100 in both the
presence and absence of rat liver metabolic activation (Stewart, 1991). Although a positive
result was reported both with and without metabolic activation for induction of p-galactosidase
synthesis (a response to DNA lesions) in E. coli PQ37 (SOS chromotest) (Odunola, 1997), this is
an isolated finding at present.
Results in mammalian mutagenicity test systems were all negative. Boric acid
(concentration, stability, and purity not tested by investigators) was negative in inducing
unscheduled DNA synthesis in primary cultures of male F344 rat hepatocytes (Bakke, 1991).
Boric acid did not induce forward mutations in L5178Y mouse lymphoma cells with or without
S-9 (NTP, 1987). Boric acid did not induce mutations at the thymidine kinase locus in the
L5178Y mouse lymphoma cells in either the presence or absence of a rat liver activation system
(Rudd, 1991). Crude borax ore and refined borax were both negative in assays for mutagenicity
in V79 Chinese hamster cells, C3H/1OT1/2 mouse embryo fibroblasts, and diploid human
foreskin fibroblasts (Landolph, 1985). Similarly, boric acid did not induce chromosome
aberrations or increase the frequency of sister chromatid exchanges in Chinese hamster ovary
cells with or without rat liver metabolic activating systems (NTP, 1987).
Proposal Draft
Boron — September, 2006 7-21
-------
O'Loughlin (1991) performed a micronucleus assay on Swiss-Webster mice (10
animals/sex/dose). Boric acid was administered in deionized water orally (no verification of
stability, concentration, or homogeneity was made of the boric acid by the investigators) for 2
consecutive days at 900, 1800 or 3500 mg/kg. Five mice/sex/dose were sacrificed 24 hours after
the final dose, and 5/sex/dose were sacrificed 48 hours after the final dose. A deionized water
vehicle control (10/sex) and a urethane positive control (10 males) were also tested. Boric acid
did not induce chromosomal or mitotic spindle abnormalities in bone marrow erythrocytes in the
micronucleus assay in Swiss-Webster mice.
7.4.2 Neurological Studies
Sodium tetraborate was administered in the drinking water to 2-month-old Wistar rats for
up to 14 weeks. Exposure to approximately 20.8 mg B/kg-day caused an increase in cerebral
succinate dehydrogenase activity after 10-14 weeks of exposure (Settimi et al., 1982). Increased
acid proteinase activity and increased RNA were also noted at the end of the 14-week
experiment.
ATSDR (1992) and Wong et al. (1964) reported on case reports of neurological effects
after accidental ingestion of high levels of boron as boric acid. Newborn infants (number not
given) who ingested 4.5-14 g boric acid showed these CNS symptoms. Doses of about 500 mg
B/kg-day showed CNS involvement with headaches, tremors, restlessness and convulsions
followed by weakness, coma, and death. Histological examination of 2/11 infants revealed
degenerative changes in brain neurons, congestion, and edema of brain and meninges with
perivascular hemorrhage and intravascular thrombosis.
O'Sullivan and Taylor (1983) reported convulsions and seizures in seven infants exposed
to a honey-borax mixture for 4-10 weeks, in which the estimated ingestion was 9.6-33 mg
B/kg-day.
Litovitz et al. (1988) conducted a retrospective review of 784 cases of boric acid
ingestion. An estimate of the amount of boric acid ingested was obtained historically in 659
cases. The average amount ingested was 1.4 g. The average dose was estimated to be 0.5 g for
children under 6 years of age, compared to 4.1 g for individuals 6 years of age and above.
Symptoms most frequently reported were vomiting, abdominal pain, diarrhea, and nausea. Other
symptoms, including CNS and cutaneous effects, occurred in six or fewer cases and included
rash, lethargy, headache, lightheadedness, fever, irritability, and muscle cramps. The average
dose for asymptomatic cases was 0.9 g compared with 3.2 g for symptomatic cases.
Neurological effects were noted in human case reports after ingestion of high levels of
boron. Animal data are limited to increased brain enzyme levels after 10-14 weeks of exposure
(Settemi et al., 1982). There is an uncertainty about neurological effects at lower doses and
other than acute duration because no data are available. This is identified as an area where
further research may be beneficial.
7.4.3 Mechanistic Studies - Testicular Effects
Proposal Draft
Boron — September, 2006 7-22
-------
The occurrence of testicular effects in the absence of overt systemic toxicity suggests a
testicular-specific mechanism of action for boron. Many studies have been conducted to
elucidate the mechanism by which boron produces testicular effects (see Section 7.2.5 for
descriptions of some of these studies). Recent reviews of this work have been published by Fail
et al. (1998) and ECETOC (1994). Despite the number of studies that have been done, the
mechanism of boron testicular toxicity remains unknown. The available data suggest an effect
on the Sertoli cell, resulting in altered physiological control of sperm maturation and release
(Failetal., 1998).
7.4.4 Mechanistic Studies - Developmental Effects
Studies regarding the mechanism of developmental toxicity produced by boron were
reviewed by Fail et al. (1998). The two most sensitive effects of boron on developing rodents
are decreased fetal body weight and malformations and variations of the ribs. Fail et al. (1998)
concluded that reduced fetal growth probably results from a general inhibition of mitosis
produced by boric acid, as documented in studies on the mammalian testis, insects, yeast, fungi,
bacteria, and viruses (Beyer et al., 1983; Ku et al., 1993b), while the rib malformations probably
result from direct binding of boron to the bone tissue. More recent investigations of the
developmental effects of boric acid (Narotsky et al., 2003; Wery et al., 2003) have produced
evidence supporting a role of altered gene expression in boron's developmental effects. These
data indicate that boric acid administration during the normal period of expansion of hox gene
expression results in rib and vertebrae alterations, coincident with altered hox gene expression.
7.4.5 Nutrition Studies
Since the 1920s, boron has been known to be an essential micronutrient for the growth of
all plants. In humans, boron is a trace element for which essentiality is suspected but has not
been directly proven (Nielsen, 1991, 1992, 1994; NRC, 1989; Hunt, 1994; Mertz, 1993).
Because deficiency in humans has not been established, there are no adequate data from which to
estimate a human requirement, and no provisional allowance has been established (NRC, 1989).
However, boron deprivation experiments with animals and three human clinical studies have
yielded some persuasive findings for the hypothesis that boron is nutritionally essential as
evidenced by the demonstration that it affects macromineral and cellular metabolism at the
membrane level (Nielsen, 1994). Experimental boron nutrition research data indicate that boron
can affect the metabolism or utilization of a number of substances involved in life processes,
including calcium, copper, magnesium, nitrogen, glucose, triglyceride, reactive oxygen, and
estrogen. These effects can affect the composition of several body systems including blood,
brain, and skeleton (Nielsen, 1996). Boron may prevent inflammatory disease because several
key regulatory enzymes in the inflammatory response are inhibited by physiological amounts of
supplemental dietary boron (Hunt, 1996). New boron nutrition research should better
characterize the mechanisms through which boron modulates immune function, insulin release,
and vitamin D metabolism (Hunt, 1996). A close interaction between boron and calcium has
been suggested. This interaction appears to affect similar systems that indirectly affect many
variables, including modification of hormone action and alteration of cell membrane
characteristics (Nielsen et al., 1987; Nielsen, 1991, 1992, 1994; Penland, 1994). Data from three
human studies of potential boron essentiality demonstrate that dietary boron can affect bone,
brain, and kidney variables. The subjects in most of these studies, however, were under some
Proposal Draft
Boron — September, 2006 7-23
-------
form of nutritional or metabolic stress affecting calcium metabolism, including reduced intake of
magnesium or physiologic states associated with increased loss of calcium from bone or the
body (e.g., postmenopausal women).
Based on these studies, in which most subjects who consumed 0.25 mg B/day responded
to additional boron supplementation, Nielsen (1991) concluded that the basal requirement for
boron is likely to be greater than 0.25 mg/day. Limited survey data indicate that the average
dietary intake of boron by humans is 0.5-3.1 mg-day (7-44 |ig/kg-day) (Nielsen, 1991). The
average U.S. adult male dietary intake of 1.52 ± 0.38 mg B/day (mean ± standard deviation)
(lyengar et al., 1988) was determined by U.S. Food and Drug Administration (FDA) Total Diet
Study methods. In a more recent study, Anderson et al. (1994) reported an intake of 1.21 ± 0.07
mg B/day for an average diet for 25- to 30-year-old males, as determined by FDA Total Diet
Study analyses. Similarly, the average dietary boron intake in Canada is reported to be 1.33 ±
0.13 mg B/day for women (Clarke and Gibson, 1988). Dietary boron consumption in Europe
could be higher than in the United States and Canada due to wine consumption (ECETOC,
1994). These and other investigators (Nielsen, 1992) also recognized that greater consumption
of fruits, vegetables, nuts, and legumes (e.g., vegetarian diets) could raise dietary boron intake.
The Institute of Medicine (IOM, 2002) developed a tolerable upper intake level (UL), the
highest daily nutrient intake that is likely to pose no risk of adverse health effects for most
individuals, for various life stages of humans. A UL for infants was judged not determinable.
The UL for adults was 20 mg B/day. The UL was set at 17 mg B/day for pregnant women 14-18
years of age, while the UL for pregnant women 19-50 years of age was set at 20 mg B/day.
Section 5.1.3. describes how these ULs were determined.
7.5 Synthesis and Evaluation of Major Noncancer Effects and Mode of Action - Oral
and Inhalation
7.5.1 Oral Exposure
Studies in laboratory animals conducted by oral exposure have identified the developing
fetus and the testes as the two most sensitive targets of boron toxicity in multiple species (Weir
and Fisher, 1972; Seal and Weeth, 1980; NTP, 1987; Fail et al., 1991; Price et al., 1996a,b; Field
et al., 1989). The testicular effects that have been reported include reduced organ weight and
organ:body weight ratio, atrophy, degeneration of the spermatogenic epithelium, impaired
spermatogenesis, reduced fertility, and sterility (Weir and Fisher, 1972; Seal and Weeth, 1980;
NTP, 1987; Fail et al., 1991; Dixon et al., 1979; Linder et al., 1990; Treinen and Chapin, 1991;
Ku et al., 1993a). The mechanism for boron's effect on the testes is not known, but the available
data suggest an effect on the Sertoli cell, resulting in altered physiological control of sperm
maturation and release (Fail et al., 1998). Developmental effects have been reported in mice,
rabbits, and rats (Heindel et al., 1992, 1994; Field et al., 1989; Price et al., 1991, 1996a,b). The
developmental effects that have been reported following boron exposure include high prenatal
mortality; reduced fetal body weight; and malformations and variations of the eyes, CNS,
cardiovascular system, and axial skeleton (Price et al., 1996a,b; Field et al., 1989). Increased
incidences of short rib XIII (a malformation) and wavy rib (a variation), and decreased incidence
of rudimentary extra rib on lumbar I (a variation), were the most common anomalies in both rats
and mice. Cardiovascular malformations, especially interventricular septal defect, and variations
Proposal Draft
Boron — September, 2006 7-24
-------
were the frequent anomalies in rabbits. Fail et al. (1998) attributed reduced fetal growth, the
most sensitive developmental endpoint, to a general inhibition of mitosis by boric acid, as
documented in studies on the mammalian testis, insects, yeast, fungi, bacteria, and viruses
(Beyer et al., 1983; Ku et al., 1993b).
7.5.2 Inhalation Exposure
Studies in humans and animals have shown that borates are absorbed following
inhalation exposure (Culver et al.,1994; Wilding et al., 1959). It is not clear what percentage of
the absorbed material in these studies was absorbed via the respiratory tract directly; transport of
deposited material from the upper respiratory tract to the gastrointestinal tract may have played
an important role (Culver et al.,1994). However, because borates in the body exist as boric acid,
are distributed evenly throughout the soft tissues in the body water, and are not metabolized (Ku
et al., 1991; Naghii and Samman, 1996; WHO, 1998a), there is no reason to expect
route-specific differences in systemic targets. Therefore, systemic target tissues identified in oral
studies comprise the potential systemic targets following inhalation exposure. There may be
route-specific differences in ability to deliver toxic doses to the targets, in that very high
exposure concentrations may be required to produce effects by inhalation exposure.
Portal-of-entry effects may also differ with exposure route.
The literature regarding the toxicity of boron by inhalation exposure is sparse. There is a
report from the Russian literature of reduced sperm analysis of 6 workers who were part of a
group of 28 male workers exposed to high concentrations of boron (boric acid) aerosols (22-80
mg/m3) for more than 10 years (Tarasenko et al., 1972). These effects are consistent with the
testicular effects reported in oral studies, but have not been confirmed by other inhalation
studies. However, data from Tarasenko et al. (1972) are of limited value for risk determination
due to sparse details and small sample size. No effect on fertility was found in a far larger study
of U.S. borate production workers (Whorton et al., 1992; 1994a,b), but exposure concentrations
were much lower (about 2.23 mg/m3 sodium borate or 0.31 mg B/m3) in this study. No target
organ effects were found in the lone animal study in which rats were exposed to 77 mg/m3 of
boron oxide aerosols (24 mg B/m3) for 24 weeks, but testicular effects were examined only by
limited histopathology (Wilding et al., 1959). This study also included a high-dose group
exposed to 470 mg/m3 boron oxide (146 mg B/m3) for 10 weeks, a concentration at which the
aerosol formed a dense cloud of fine particles that covered the animals with dust. Systemic
endpoints were not examined, but growth was reduced by 9% in the high-dose group, and there
was evidence of nasal irritation. Acute irritant effects are well documented in human workers
exposed to borates, primarily at concentrations greater than 4.4 mg/m3 (Wegman et al., 1994;
Garabrant et al., 1984, 1985). However, there is no evidence for reduced pulmonary function in
workers with chronic exposure (Wegman et al., 1994). These data are inadequate to support
derivation of an RfC for boron compounds.
7.6 Weight of Evidence Evaluation and Cancer Characterization - Synthesis of Human,
Animal, and Other Supporting Evidence, Conclusions About Human
Carcinogenicity, and Likely Mode of Action
Under the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005), the data are
considered to be inadequate for an assessment of the human carcinogenic potential of boron. No
Proposal Draft
Boron — September, 2006 7-25
-------
data were located regarding the existence of an association between cancer and boron exposure
in humans. Studies available in animals were inadequate to ascertain whether boron causes
cancer. The chronic rat feeding study conducted by Weir and Fisher (1972) was not designed as
a cancer bioassay. Only a limited number of tissues were examined histopathologically, and the
report failed to mention any tumor findings. The chronic mouse study conducted by NTP (1987)
was adequately designed, but the results are difficult to interpret. There was an increase in
hepatocellular carcinomas in the low dose, but not the high dose, in which male mice were
within the range of historical controls. The increase was statistically significant using the life
table test, but not the incidental tumor test. The latter test is more appropriate when the tumor in
question is not the cause of death, as appeared to be the case for this study. There also was a
significant increase in the incidence of subcutaneous tumors in low-dose male mice. However,
once again the increase was within the range of historical controls and was not seen in the
high-dose group. Low survival in both the low- and high-dose male groups (60 and 40%,
respectively) may have reduced the sensitivity of this study for evaluation of carcinogenicity.
The chronic mouse study conducted by Schroeder and Mitchener (1975) was inadequate to
detect carcinogenicity because only one, very low dose level was used (0.95 mg B/kg-day), and
the MTD was not reached. No inhalation cancer data were located. Studies of boron compounds
for genotoxicity were overwhelmingly negative, including studies in bacteria, mammalian cells,
and mice in vivo.
7.7 Susceptible Populations
7.7.1 Possible Childhood Susceptibility
One of the most sensitive targets of boron that has been identified is the developing fetus
(rats, mice and rabbits) carried by the pregnant female. A set of well-designed developmental
studies in rats provided an LOAEL of 13.3 mg B/kg-day and an NOAEL of 9.6 mg B/kg-day in
the developing fetus, based on decreased fetal body weight (Price et al., 1996a).
7.7.2 Possible Gender Differences
Another sensitive target of boron that has been identified is the testis of the male. A
study in dogs provided an LOAEL of 29 mg B/kg-day and an NOAEL of 8.8 mg B/kg-day,
based on histopathological effects (Weir and Fisher, 1972). Sensitivity to boron exposure does
not appear to differ markedly for these two targets, although there is some uncertainty in this
determination due to the less comprehensive design of the dog study.
Effects on the pregnant females themselves are seen only at considerably higher doses
(no clearly adverse maternal effects even at 94.2 mg B/kg-day in the same study used to derive
the NOAEL and LOAEL values for the developing fetus reported above). A specific target of
boron toxicity has not been identified in nonpregnant females, who are markedly less susceptible
to boron than males. Data are inadequate to assess differences in gender susceptibility with
regard to non-reproductive, non-developmental effects.
7.7.3 Physiological and Disease Anomalies
Proposal Draft
Boron — September, 2006 7-26
-------
Because the removal of boron (boric acid) from mammals occurs via renal elimination of
the unchanged molecule, alterations of renal function result in increased residence time.
Decrements of renal function, therefore, will increase internal exposure, and may predispose
affected individuals to greater risk from compounds for which renal elimination is important.
The observed developmental toxicity of boron indicates that fetuses of pregnant women may be
the susceptible group; those fetuses of women who are experiencing renal insufficiency may
represent a sensitive sub-population. Preeclampsia is a health condition of pregnancy in which
renal function, including glomerular filtration, is reduced.
Proposal Draft
Boron — September, 2006 7-27
-------
8.0 DOSE-RESPONSE ASSESSMENT
8.1 Oral Reference Dose (RfD)
8.1.1 Choice of Principal Study and Critical Effect — with Rationale and
Justification
Developmental effects (decreased fetal weights) are considered the critical effect. The
studies by Price et al. (1990, 1994, 1996a) and Heindel et al. (1992) in rats were chosen as
critical developmental studies because they were well-conducted studies of a sensitive endpoint
that identified both an NOAEL and LOAEL. Rats were more sensitive than mice and rabbits,
which were also studied for developmental toxicity (Price et al., 1996b; Heindel et al., 1994).
There was a consistent correlation between boric acid exposure and different effects on
ribs and vertebral development in rats, mice and rabbits for which the rat was the most sensitive
to low-dose effects. Because decreased fetal body weight in rats occurred at the same dose or at
lower doses than those at which skeletal changes were observed, the decreased fetal body weight
data set was chosen for developing a reference dose. IEHR (1997) agreed with the correlation
between boric acid exposure and the different effects on rib and vertebral development in rats,
mice, and rabbits and the causal association between exposure to boric acid and the short rib XIII
(when fetuses were examined at late gestation or when pups were examined at pnd 21) and that
decreased fetal body weight should be used for deriving quantitative estimates.
The dog study by Weir and Fisher (1972) identified an NOAEL of 8.8 mg/kg-day and
LOAEL of 29 mg/kg-day for testicular effects. Testicular effects were found at higher doses in
rats and mice in this and other studies (Weir and Fisher, 1972; Seal and Weeth, 1980; NTP,
1987; Fail et al., 1991; Dixon et al., 1979; Linder et al., 1990; Treinen and Chapin, 1991; Ku et
al., 1993a). These effects include testicular atrophy, inhibition of spermiation, degeneration of
seminiferous tubules with germ cell loss, and loss of fertility. In a rat multigeneration study by
Weir and Fisher (1972) an NOAEL of 17.5 mg/kg-day and an LOAEL of 58.5 mg/kg-day for
testicular atrophy was reported in male Sprague Dawley rats. Ku et al. (1993a) reported an
NOAEL of 26 mg/kg-day for inhibited spermiation in male Sprague Dawley rats. Fail et al.
(1991) reported an LOAEL of 26.8 mg/kg-day in male Swiss CD mice for decreased sperm
motility. Because the LOAELs for testicular effects were more than 2-fold greater than the
LOAEL for developmental effects, the Weir and Fisher dog study was not considered as the
critical study. However, as no exposure level was tested in the dog study between 8.8 and 29
mg/kg-day, uncertainty remains as to whether testicular effects would have occurred near the
same exposure leading to developmental effects.
The Weir and Fisher (1972) study in dogs had other limitations for RfD derivation,
including small number of test animals per dose group (n=4), the use of shared control animals
in the borax and boric acid studies so that at most two control animals were sacrificed at any
time period, the observation of testicular damage in three of four control animals, and the
NOAEL and LOAEL taken from two different studies of different duration. Also, the study
pathologist considered the histopathological findings to be "not compound-induced." Based on
the small number of animals and the wide range of background variability among the controls,
these studies do not appear to be adequate for establishment of a defensible RfD.
Proposal Draft
Boron — September, 2006 8-1
-------
8.1.2 Methods of Analysis — Including Models
The RfD was derived by the benchmark dose (BMD) approach. Several BMD analyses
were conducted by Allen et al. (1996) using all relevant endpoints in the Heindel et al. (1992)
and Price et al. (1994, 1996a) developmental studies in rats. Allen et al. (1996) concluded that
decreased fetal body weight was the most suitable endpoint for developing a point of departure,
because the benchmark doses calculated for the other endpoints (incidence of total
malformations, enlarged lateral ventricles in the brain, shortening of rib XIII, and variations of
the first lumbar rib) were higher.
Changes in fetal weight were analyzed by taking the average fetal weight for each litter
with live fetuses. Those averages were considered to represent variations in a continuous
variable, and a continuous power model was used. A BMD was defined in terms of a
pre-specified level of response, referred to as the benchmark response (BMR) level (Kavlock et
al., 1995). For mean fetal weight analysis, the 95% lower confidence limit on the benchmark
dose (BMDL) was defined as the 95% lower bound on dose corresponding to a 5% decrease in
the mean (that is, the BMR in this case is a 5% decrease in mean fetal weight per litter). This
BMR is approximately equivalent to a 0.5 standard deviation decrease in the control mean, or an
extra risk of about 5% of an exposed population having litters with mean fetal body weights less
than those of 98% of the control population. Goodness of fit was evaluated using F-tests that
compared the lack of model fit to an estimate of pure error.
The earlier study by Heindel et al. (1992) did not define an NOAEL, while the later study
by Price et al. (1996a) was designed as a follow up study to the Heindel study to examine fetal
body weight at lower doses to define an NOAEL. Allen et al. (1996) examined the
dose-response patterns for the two studies to determine if a single function could adequately
describe the responses in both studies. This determination was based on a likelihood ratio test.
The maximum log-likelihoods from the models fit to the two studies considered separately were
added together; the maximum log-likelihood for the model fit to the combined results was then
subtracted from this sum. Twice that difference is distributed approximately as a chi-square
random variable (Cox and Lindley, 1974). The degrees of freedom for that chi-square random
variable are equal to the number of parameters in the model plus 1. The additional degree of
freedom was available because the two control groups were treated as one group in the combined
results, which eliminates the need to estimate one of the intra-litter correlation coefficients (for
beta-binomial random variables) or variances (for normal random variables) that was estimated
when the studies were treated separately. The critical values from the appropriate chi-square
distributions (associated with a p-value of 0.01) were compared to the calculated values. When
the calculated value was less than the corresponding critical value, the combined results were
used to estimate BMDLs. The data and details of the modeling are provided in Appendix B.
The results of the Allen et al. (1996) BMD analysis for decreased fetal body weight for
the Price study alone gave a BMDL of 47 mg boric acid/kg-day (8.2 mg B/kg-day), and for the
Heindel study alone, the BMDL reported by Allen et al. (1996) was 56 mg boric acid/kg-day
(9.8 mg B/kg-day). The statistical analysis described above demonstrated that the data were
consistent, and could be combined to estimate a single dose-response function. The combined
data from Heindel et al. (1992) and Price et al. (1994, 1996a) gave a BMDL05 of 59 mg boric
acid/kg-day (10.3 mg B/kg-day). The BMDL based on the combined results of the two studies
Proposal Draft
Boron — September, 2006 8-2
-------
was very close to the NOAEL of 9.6 mg B/kg-day from the Price et al. (1994, 1996a) study. The
BMDL05 from the combined studies was chosen to derive the RfD because they were similarly
designed studies conducted in the same laboratory, and all the dose response data were
consistent enough to be used in the BMDL estimation, thereby increasing the confidence that the
dose response pattern has been estimated satisfactorily.
Allen et al. (1996) noted that merely increasing sample size does not always increase the
precision of the estimates of the BMD. For these datasets, however, the BMDLs estimated for
the combined mean fetal weight data were closer to the corresponding BMDs than for either of
the studies alone. That is, the confidence intervals around the best estimates of dose
corresponding to the selected response level were narrower in the combined analysis.
8.1.3 Derivation of the RfD
Uncertainty factors (UFs) are applied in the RfD methodology to account for recognized
uncertainties in extrapolation from experimental conditions to lifetime exposure for humans.
These UFs cover somewhat broad areas of uncertainty, such as "animal-to-human" (interspecies;
UFA) and "sensitive human" (interindividual; UFH) extrapolations. Both UFA and UFH,
however, can be addressed as a combination of two sub factors, one each for toxicokinetics (TK)
and toxicodynamics (TD).1 The TK/TD "paradigm" formally allows for the quantitative
incorporation of additional data previously used in only a qualitative fashion. The concept is
applied in the Methods for Derivation of Inhalation Reference Concentrations and Application of
Inhalation Dosimetry (U.S. EPA, 1994b), in which the kinetic component deals primarily with
airway anatomy and physiology, but does not address systemic kinetics and dynamics.
Otherwise, the U.S. EPA has not established guidance in this area. The International Programme
on Chemical Safety (IPCS) has drafted guidance in the selection of chemical-specific adjustment
factors (CSAF), which does cover systemic kinetics and dynamics (IPCS, 2001). The IPCS
document has not been formally reviewed by the U.S. EPA. Much of the toxicokinetic factor
development in the boron RfD derivation, however, is consistent with IPCS (2001).
Additionally, IPCS previously applied the TK/TD subfactor approach in their assessment of
boron (WHO, 1998a). The values for the TK component of UFA and UFH have been adjusted
based on relevant data, but no such data exist to support an adjustment of the TD components.
For boron, the animal-to-human and sensitive-human uncertainty factors (UFA and UFH)
are each split into toxicokinetic (TK) and toxicodynamic (TD) components to apply existing rat
and human toxicokinetic data to reduce the uncertainty in the boron RfD. The product of AF^
and AF^ replaces the animal-to-human (interspecies) uncertainty factor (UFA) in the standard
RfD methodology. Similarly, the product of AF^ (the interspecies toxicokinetic adjustment
factor) and AF^, (the interspecies toxicodynamic adjustment factor) replaces the sensitive
human (interindividual variability) uncertainty factor (UFH). Each of the adjustment factors is
the product of data-derived scaling factors and residual uncertainty.
8.1.3.1 Derivation of Adjustment Factor Values
Commonly known as pharmacokinetics and pharmacodynamics in the medical literature.
Proposal Draft
Boron — September, 2006 8-3
-------
As presented below, the examination of species differences in boron distribution to
extravascular fluids and renal elimination served as the basis for the replacement of the default
value for UFA-TK, while critical evaluation of the human interindividual variation of underlying
renal clearance mechanism (GFR) served as the basis upon which to replace the default value for
the TK component of UFH. Because no data were available to inform a mode or mechanism of
action for boron, the default values for the TD component of both UFA and UFH remain; they are
10°-5, or 3.16 for each.
In the most simple terms, toxicokinetics deal with what the body does to the chemical,
while toxicodynamics deal with what the chemical does to the body. In essence, the
toxicokinetic factor addresses internal exposure, in that the objective is to determine the dose of
the ultimate toxic form of the compound at the target tissue. The toxicodynamic factor, then,
deals with the response of the target tissue given a specific dose. A "pure" toxicodynamic factor
must be independent of the toxicokinetics. As it is unlikely that in vivo responses will be free of
kinetic variability, toxicodynamic data will be obtained largely from in vitro (cellular level)
studies. In these cases, a connection to systemic dynamics must be established, as well. Given
enough data, the form of the resulting model could be manifested as a sophisticated
multi-compartment, highly non-linear, biologically-based toxicokinetic model linked to a
mathematical dose-response model relating cellular response to whole-organism response. Most
of the time, however, the model will be a simple multiplicative combination of two factors, one
for TK and one for TD. Even more often, data will only be available for determination of the TK
factor, requiring the use of a default value for TD. Lacking a sophisticated model, the usual
approach will be to find one or more kinetic variables (relating to internal dose) for which an
animal-to-human ratio can be estimated, using that ratio to scale the human exposure (external
dose) relative to the test animal. Whenever the kinetic factors are used in this manner, additional
factors must be considered to relate the internal kinetics back to the external dose. Simple
absorption and distribution constants usually suffice.
TK/TD Subfactor Default Values (Uncertainty)
WHO (1994) and IPCS (2001) have maintained a default value of 10 for both UFA
(interspecies uncertainty) and UFH (intraspecies uncertainty). For UFA, they have apportioned
the factor of 10 between the TD and TK components so that the default value for the TD
component is 2.5 (1004), and the default value for the TK component is 4.0 (1006) in the absence
of data describing toxicodynamic or toxicokinetic differences. Similarly, WHO (1994) and IPCS
(2001) divided UFH into TD and TK components with assigned default values of 3.16 (1005)
each. The U.S. EPA has assumed an equal contribution (10°5 each) of TK and TD for both UFA
and UFH when deriving the RfC, but has not explicitly addressed the issue for RfDs. As the
factors are now meant to include kinetic and dynamic dose adjustments based on data, as well as
uncertainty, they more appropriately are termed "adjustment factors." As standard notation in
this document, these factors henceforth will be designated as AFAK, AFAD, AF^, and AF^,,
respectively. Note that these factors serve as both variability factors when relevant data exist
and uncertainty factors when relevant data do not exist.
The default half-order of magnitude partition of uncertainty factors (i.e., UFA and UFH)
for toxicokinetics and toxicodynamics is primarily based on lack of knowledge; if there is no
evidence to the contrary, an equal contribution from each source of uncertainty is assumed.
Although there is empirical and conceptual support for a value other than 1005 for the TK default
Proposal Draft
Boron — September, 2006 8-4
-------
for UFA for compounds kinetically similar to boron2, there are no data addressing the TD
component. In addition, lacking a formal review, the IPCS uneven split is not adopted here.
Therefore, any uneven split of the 10-fold factor for UFA would be somewhat arbitrary, and the
half-order-of-magnitude TK/TD default partition is maintained for this analysis. The even split
is also adopted for UFH, as there are no strong arguments for different values for either the TK or
TD factors.
Revised RfD Calculation Formula
The revised formula for calculating the RfD with UFA and UFH split into TK and TD
subfactors is given in Equation 5.1.
RfD = D^AF^ AF^ AF^ AF^ UF) (5.1)
where:
DC is the "critical" dose (NOAEL, LOAEL, BMD) defined in the critical study,
is the interspecies toxicokinetic adjustment factor (default = 3.16)
is the interspecies toxicodynamic adjustment factor (default = 3.16)
, is the interindividual toxicokinetic adjustment factor (default = 3.16)
is the interindividual toxicodynamic adjustment factor (default = 3.16)
UF is the aggregate uncertainty factor
The product of AF^ and AFAD replaces the animal -to-human (interspecies) uncertainty
factor (UFA) in the standard RfD methodology. Similarly, the product of AFAK and AF^
replaces the sensitive human (interindividual variability) uncertainty factor (UFjj). Each of the
adjustment factors is the product of data-derived scaling factors and residual uncertainty. That
is, if there are significant issues concerning the data or modeling of the data, the adjustment
factor may be increased to reflect remaining uncertainty. If there are no applicable data, the
adjustment factors are equal to their default uncertainty factor values. The aggregate uncertainty
factor (UF) is equal to the product of all other uncertainty factors: subchronic-to-chronic (UFS),
LOAEL-to-NOAEL (UFS), and data base adequacy (UFD). For boron, a subchronic-to-chronic
uncertainty factor was not used to account for extrapolation from less than chronic results
because developmental toxicity (decreased fetal body weight) was used as the critical effect.
The developmental period is recognized as a susceptible lifestage where exposure during certain
time windows is more relevant to the induction of developmental effects than lifetime exposure.
An uncertainty factor for extrapolation from an LOAEL to an NOAEL was not necessary
because BMD modeling was used to determine the point of departure. The dose corresponding
to a 5% decrease in pup weight, relative to control, was selected as the point of departure.
Because decreased weights did not persist in the companion study (Phase II of Price et al.,
1996a, 1994), no further adjustment was considered for identifying a level of oral exposure to
This class of substances would include those that are water soluble and eliminated unchanged through the
kidneys. The difference in elimination would be primarily in the renal clearance rate. A fairly large body of
evidence suggests that many of the factors that determine kinetics generally scale to BW075 across species. In
particular, renal clearance values scale across species with an exponent ranging from 0.69-0.89 (Davidson et al.,
1986). For rats to humans, the allometric argument supports a value near 4.0 as the average, or expected, factor for
scaling test-animal kinetics to human kinetics. The default TK value would be somewhat larger to allow for
departures from the expected value. In addition, the default value would be species specific.
Proposal Draft
Boron — September, 2006 8-5
-------
boron associated with the minimal level of risk. A database uncertainty factor was not deemed
necessary due to boron's extensive data base. For convenience and sake of reference, the
product of all the terms in the denominator of Equation 5.1 is given the term "total adjustment
factor" and is designated as AFTOT.
8.1.3.2 Toxicokinetic Modeling Issues for Boron
While no data presently exist to address the toxicodynamic components of UFA or UFH,
existing data are adequate to establish non-default values for AFAK and AF^ and reduce
uncertainty in the toxicokinetic components of both uncertainty factors. The most relevant
internal dose metric for boron toxicity, which is most likely a result of continuous exposure over
an extended period, is the average fetal concentration for the entire gestational period. Although
there are no direct measurements of fetal boron concentrations, boron concentrations in the fetus
should be the same as in the mother because boron is freely diffusible across biological
membranes and will rapidly and evenly equilibrate in all body water compartments. As the
boron RfD is based on developmental effects observed in rats, the most relevant kinetic data are
those pertaining to pregnant rats and pregnant humans. There are insufficient data to compare
plasma boron in rats and humans at the same exposure levels. Therefore, boron clearance is used
as an estimator of internal dose. Assuming steady state conditions, clearance, expressed in units
of mL/min (volume of plasma cleared of the substance per unit time), is inversely related to
plasma concentration. Clearance is calculated by dividing the total mass of substance eliminated
in the urine in a specific time (i.e, mg/min) by the concentration of the substance in the plasma
(mg/mL). Therefore, the higher the clearance value, the lower the plasma concentration. Other
processes, such as fecal elimination, metabolism, and distribution to other compartments also
reduce the plasma concentration. However, as boron is not metabolized and almost entirely
eliminated in the urine, clearance of boron by the kidney can be used as the key toxicokinetic
factor, with a consideration of the relative volumes of distribution between rats and humans.
Although the toxic effects of boron are manifested in the offspring, pregnant females (for
both humans and test animals) are considered to be the "sensitive" population, with respect to
establishing an equivalent toxic dose across species. For the RfD, toxicity benchmarks are
expressed in terms of external (maternal) exposure, rather than internal (fetal) dose. In this
sense, the maternal boron concentration is treated as a surrogate for the fetal boron
concentration. A compartmentalized toxicokinetic model, with the fetus as one of the
compartments, would be needed to directly assess the dose to the fetus. Given the near first
order kinetics of boron, maternal toxicokinetic variability is an adequate surrogate for the fetal
dose variability.
Proposal Draft
Boron — September, 2006 8-6
-------
Interspecies Uncertainty
As the rathuman boron clearance ratio is being used essentially as an (inverse) estimator
of relative internal dose and, subsequently, as a scalar of "external dose" (ingested dose rate in
mg/kg-day), an additional factor must be considered that ties internal dose to external dose. As
there is an assumption of relatively constant intake of boron and the toxic outcome is most likely
related to a continuous exposure over an extended critical period (the period of organogenesis
during fetal development), the most appropriate estimator for internal dose is the average
(steady-state) circulating boron concentration.
Boron distributes primarily to total body water and bone, reaching a 4-fold higher
concentration in whole bone than in plasma (Chapin et al., 1997). Boron freely transfers from
bone to body water, as well. Therefore, a two-compartment steady-state model is assumed for
this analysis. The generalized two-compartment steady-state model is described in O'Flaherty
(1981). The steady-state circulating concentration (Css) of boron (or other compound) for a
two-compartment model, given a constant rate of administration (oral ingestion), simplifies to
Equation 5.2.
Css = (DefaBW)/Cl (5.2)
where:
De is the external ingested dose rate in mg (boron) per kg body mass per day
fa is the fraction of ingested boron absorbed into the body from the gut
BW is body weight (kg)
Cl is the renal clearance rate (mL/minute)
An assumption is made that all of the boron is eliminated in the urine. Small losses in
sweat, saliva, and the feces are ignored.
The interspecies toxicokinetic adjustment factor, AFAK, is used to adjust the test-animal
dose rate to obtain an equivalent human exposure. In this case, AFAK is equal to the ratio of
De-rat to De-human at a fixed target tissue dose. As Css is used as the estimator for target tissue
dose, the latter condition (fixed target tissue dose) is satisfied by setting the rathuman Css ratio
to 1. Therefore, solving Equation 5.2 for De, taking the ratio of rat and human De, and setting the
rathuman Css ratio to 1, yields Equation 5.3, where the trailing subscript designates the species
® = rat, h = human).
= >i
Clh x far x BWr (5.3)
The mean boron clearance for pregnant rats (Clr)is 1.00, determined from the kinetic
studies of U.S. Borax (2000) and Vaziri et al. (2001) (Table 6-4). The mean boron clearance for
pregnant women (C1A) was determined from the kinetic studies of U.S. Borax (2000) and Pahl et
al. (2001) to be 66. 1 mL/min (Table 6-5). The mean body weights for pregnant rats (BWr) and
pregnant women (BWA) from those studies are 0.303 and 67.6 kg, respectively. The average
clearance of 66 mL/min for pregnant women determined by Pahl et al. (2001) represents a
possible underestimation of the true boron clearance, particularly at the relatively higher doses
near the RfD. Boron clearance values obtained in adult men (Jansen et al., 1984a) given an
Proposal Draft
Boron — September, 2006 8-7
-------
intravenous infusion of boric acid, representing exposures 66 times dietary levels, were 1.5 times
greater than boron clearance measured at dietary levels. Taking into account the possibility of
dose-dependence, and that the RfD is somewhere between the dietary exposure and infusion
level in the Jansen study (but much closer to the latter), the factor could be less than 1.5 (1.3 by
linear interpolation). Therefore, C1A could actually be 30-50% higher (86-99 mL/min). An
independent estimate in the range of 86 to!07 mL/min boron clearance in pregnant women can
be obtained from the adult male boron clearance of 60.5 mL/min/1.73 m2 (Jansen et al., 1984a)
by assuming that boron clearance will scale the same as GFR from male to female to pregnant
female. GFR is about 8-12% higher in adult males than females (Smith, 1951; Wesson, 1969),
but increases by a factor of about 1.6 in pregnancy (Dunlop, 1981; Sturgiss et al., 1996; Krutzen
et al., 1992). Furthermore, GFR values normalized to a standardized unit surface area (1.73 m2)
for pregnant women may underestimate absolute GFR (mL/min) by an additional factor of 1.2
(Krutzen et al., 1992). Therefore, the adult male boron clearance of 60.5 mL/min/1.73 m2
represents a clearance of at least 86 mL/min and as much as 107 mL/min in pregnant women.
Although this evidence is suggestive that C\h may be higher, it is not strong enough for a
quantitative adjustment in the derivation of AFAK. Therefore, C\h is assigned the value of 66. 1
mL/min, C\h is 1.00 mL/min, BWr is 0.303 kg, and BWA is set to 67.6 kg.
Absorption across the gut is similar in rats and humans. Although there are no data
specifically for pregnant individuals, boron is 95% absorbed from the G.I. tract by adult rats
(Vanderpool et al., 1994) and about 92% by adult humans (Schou et al., 1984). Therefore, /ah
are set to 0.92 and 0.95, respectively.
Substituting the foregoing estimates for all the variables in Equation 5.3 yields a value of
3.3 for AF^ ([1.00/66.1] x [0.92/0.95] x [67.6/0.303]). Although there are a number of
uncertainties in the estimation of the variables in Equation 5.4, there is a likely net upward bias
in AFAK because of the potential underestimation of Clh. The value of 3.3 for AF^, therefore,
represents a somewhat health protective value, and an additional adjustment for residual
uncertainty is judged to be unnecessary. There are no data for estimating AFAD; it remains the
default value of 10°-5 (3. 16).
Intraspecies Uncertainty
Conceptually, the intraspecies toxicokinetic adjustment factor (AF^) accounts for the
range of human interindividual variability from where AFAK left off to where the sensitive
sub-population is adequately protected. For boron, the range is between the mean and a "lower
bound" boron clearance in the pregnant human population. AF^ needs to cover a sufficient
fraction of the population (on the toxicokinetic scale) so that the probability of having both a low
clearance and high sensitivity (on the toxicodynamic scale3) is low enough to preclude
appreciable risk of deleterious effects in the population (including sensitive individuals).
For the assessment of interindividual toxicokinetic variability, GFR is used as a surrogate
for boron clearance. Although the study of Pahl et al. (2001) provides an estimate of boron
clearance variability in pregnant women, the data are judged to be inadequate for this purpose.
The Pahl et al. (2001) study is considered to be a good study for estimating the mean boron
3Toxicodynamic sensitivity is represented by AFHD.
Proposal Draft
Boron — September, 2006
-------
clearance in pregnant women, but was not designed to assess interindividual variability, given its
fairly low number of subjects (16) and a lack of control of dietary intake of boron. The variance
of boron clearance in this study was somewhat high (CV = 0.49), such that estimation of an
adequate lower bound would be highly uncertain. In contrast, in the controlled infusion
exposure study of Jansen et al. (1984a), the boron clearance CV was 0.09 (Section 3.4.1). In that
same study, clearance determined for uncontrolled dietary exposure at much lower levels was
characterized by high variability (CV = 0.78). Lack of controls on exposure magnitude and
timing would be expected to contribute substantially to the variance of the measurements. The
high variability reported by Pahl et al. (2001), therefore, is attributed to experimental "noise" and
should not be included in the estimate of true population variability. As boron clearance is
largely a function of GFR, the larger more certain data base on GFR and its variability among
humans is used to estimate boron clearance variability. Because the measured boron clearances
in the rat and human kinetic studies were less than GFR, tubular reabsorption could be
contributing to the variability of boron clearance in the population. Variability in these factors,
however, is judged to be minor in comparison to the variability in GFR (Section 6).
GFR data have been used previously in the context of the boron RfD by Dourson et al.
(1998), who proposed the ratio of the mean GFR to the GFR value 2 standard deviations (SD)
below the general population mean (mean/[mean - 2 SD]) as the metric for the interindividual
toxicokinetic adjustment factor. This approach will be referred to as the sigma method, which is
a common term used for statistical methods using multiple standard deviations to establish
"acceptable" lower bounds. For the derivation of AFj^, the sigma method is modified by using 3
SD as the reduction factor for establishing the lower bound (i.e., mean GFR - 3 SD) (equation
5.4). The basic formula modified from Dourson et al. (1998) for AF^ is:
GFR
AVG_
GFRAVG- 3 SDGFR (5.3)
where GFRAVG and SDGFR are the mean and standard deviation of the GFR (mL/min) for the
general healthy population of pregnant women. The use of 3 standard deviations rather than 2
(as in Dourson et al., 1998) is based on obtaining adequate coverage of pregnant women with
very low GFR.
The selection of 3 SD is based on a statistical analysis of the published GFR data, with
more consideration being given to the full range of GFR values likely to be found in the
population of pregnant women. In the aggregate, the data suggest that a lower bound GFR 2 SD
below the mean does not provide adequate coverage of the susceptible sub-population. While no
conclusive information exists from controlled-dose studies in humans, it may be possible that the
variability in boron clearance might be greater than GFR variability. Therefore, AF^ must also
account for any residual uncertainty in using GFR as a surrogate.
GFR is measured most accurately using substrates that are not metabolized and not
actively secreted or reabsorbed from the kidney tubules, such as inulin and iohexol. Three such
studies were located in the published literature that address GFR variability in pregnant women
(Dunlop, 1981; Krutzen et al., 1992; Sturgiss et al., 1996). Because no data exist that identify a
specific developmental period, data from the entire pregnancy duration are used where possible.
Proposal Draft
Boron — September, 2006 8-9
-------
Dunlop (1981) assessed GFR for 25 women at three different time points during
pregnancy (16, 26, and 36 weeks) and again after delivery. GFR was measured as inulin
clearance. The mean values for GFR for these measurement periods were 148.6, 152.4, and
150.5 mL/min, respectively. The standard deviations were 17.2 and 17.6 mL/min for the first
two measurements, rising to 31.8 mL/min for the 36-week measurement. For the present
analysis (Table 8-1), the overall average and standard deviation (150.5 and 17.6 mL/min,
respectively) for the serially-averaged measurements for each individual across the three
pregnancy time points were used.
Sturgiss et al. (1996) performed a similar assessment of GFR (using inulin clearance) for
21 women in early (12-19 weeks) and late (30-35 weeks) pregnancy and again at 15-25 weeks
post partum. The primary purpose of the study was to determine whether the increase in GFR
normally occurring in pregnancy represents a maximal utilization of renal reserve (it did not in
this study). To evaluate that hypothesis, GFR for 14 of the 21 women (Index group) was
assessed following an infusion of an amino acid solution (known to increase GFR) in each of the
three measurement periods, subsequent to assessment of their basal GFR for each period. The
other seven women (control group) received an infusion of Hartman's solution instead of amino
acids, and basal GFR was assessed in the same manner as the Index group. Combining the basal
(unperturbed) measurements for all 21 subjects4, serially averaged for each individual for both
pregnancy time points, resulted in a mean GFR of 138.9 mL/min with a standard deviation of
26.1 mL/min.
Krutzen et al. (1992) evaluated GFR during pregnancy for four different groups of
women: 13 normal healthy women, 16 diabetic women, 8 hypertensive women, and 12 women
diagnosed with preeclampsia. GFR was determined by iohexol clearance in the second and third
trimester and again 6-12 months post partum. The authors reported absolute clearance values (in
mL/min) for only the third trimester. The third trimester mean GFR and standard deviation for
the healthy women were 195 and 32 mL/min, respectively. Mean GFR in the third trimester was
not reduced for the hypertensive women and was slightly reduced in the diabetic women, with a
mean of 169 mL/min (SD = 34.7). The third trimester mean GFR of 128 mL/min (SD = 33.9
mL/min) for the preeclamptic women, however, was more than two standard deviations below
the healthy mean GFR. In general, the GFR values reported in this study are much higher than
those reported by Dunlop (1981) and Sturgiss et al. (1996). The reason for this discrepancy is
not known.
By virtue of their lower GFR, pregnant women diagnosed with preeclampsia could be
considered to be a sensitive subpopulation, at least on the toxicokinetic scale. Toxicodynamic
sensitivity is presumably independent of toxicokinetic sensitivity. The onset of preeclampsia
generally occurs after the week 20 of pregnancy and is characterized by acute hypertension,
often accompanied by edema and proteinuria. Women with preeclampsia are at increased risk
for premature separation of the placenta from the uterus and acute renal failure, among other
adverse health effects. The fetus may become hypoxic and is at increased risk of low birth
weight or perinatal death. Preeclampsia has recently been estimated to affect 3-5% of pregnant
women (Skjaerven et al., 2002). With almost 4 million successful pregnancies per year in the
4That is, index plus control individuals in Table II, Sturgiss et al. (1996).
Proposal Draft
Boron — September, 2006 8-10
-------
United States (Ventura, 1999), or about 3 million at any one time, the size of the preeclamptic
population at any given time could be in the range of 150,000 to 200,000 women. Considering
the Krutzen et al. (1992) results in the context of the sigma method, a reduction of 2 SD from the
healthy population mean to establish the lower bound (which results in a GFR slightly higher
than the mean of the preeclamptic GFR), would appear to be insufficient for adequate coverage
of the susceptible population. The use of 3 SD below the healthy GFR mean gives coverage in
the sensitive subpopulation to about 1 SD below the mean preeclamptic GFR.
As no single study is considered to be definitive for assessment of population GFR
variability, AFj^ is determined from the average of the individual sigma-method values for each
of the three studies (Table 8-1). The mean GFR and standard deviation values in Table 8-1 are
based on average GFR across the entire gestational period, except for the Krutzen et al. (1992)
estimate, which was for the third trimester only. The average sigma-method value from the three
studies is 1.93. Considering a small residual uncertainty in the use of GFR as a surrogate for
boron clearance, the average sigma-method value of 1.93 is rounded upward to 2.0 and
established as the value for AF^. The data on preeclamptic women presented by Krutzen et al.
(1992) were considered insufficient to base the interindividual AF^ factor. Use of the mean
(128 mL/min) and standard deviation (33 mL/min) in this sensitive subgroup of preeclamptic
women likely overestimates the spread of GFR values below the mean, due to the likelihood of a
log normal distribution of GFR values, and the contribution of measurement variability (beyond
biological variability) to the statistical confidence limits. Given these considerations, the ~2-fold
interindividual variability factor derived from three standard deviations below the mean of three
studies for pregnancy GFR (mean = 161.5 mL/min; mean - 3 SD = 85.8) is considered preferable
for providing adequate coverage to women predisposed to adverse birth outcomes due to renal
complications.
The decrement of renal function can predispose individuals to both maternal and fetal
adverse effects. Thus, there are levels of renal function (GFR) which increase the risk of adverse
developmental effects that cannot be distinguished from the potential adverse effects of boron.
Thus, this level of renal function would serve as a physiological lower bound on the value for the
denominator of Equation 5.4. Establishing the level unequivocally is problematic, as the
incidence, severity, and relevance (to boron toxicity) of adverse pregnancy outcomes associated
with low GFR is difficult to establish. Further complicating the issue are the metrics reported in
the literature; pregnancy outcomes are commonly related to pre-pregnancy measures of renal
function, which are generally expressed as serum creatinine levels. There are no data directly
relating GFR or serum creatinine levels in pregnant women to adverse pregnancy outcomes. The
approach taken in the literature reflects the physician's need to advise kidney patients prior to
becoming pregnant. Also, at lower (normal) serum creatinine levels, serum creatinine is a
reliable measure of GFR. At higher serum creatinine levels (lower GFR), the relationship
apparently disappears (Levey et al., 1988). However, a linear regression analysis of the log-log
transformation of the published data (Shemesh et al., 1985, reproduced in Levey et al., 1988)
shows a significant relationship over a wide range of serum creatinine levels.
From the regression analysis shown in Appendix C of the IRIS Toxicological Review
(Regression Analysis of Serum Creatinine and Inulin Clearance, U.S. EPA, 2004a) and the
results of clinical studies, a ratio of average (nonpregnant) GFR to (nonpregnant) GFR levels
associated with significant adverse pregnancy outcomes can be calculated. This ratio would
Proposal Draft
Boron — September, 2006 8-11
-------
represent a "physiological" AF^ estimating the point at which low GFR would be a major factor
in adverse pregnancy outcomes. Several clinical investigations in humans have demonstrated a
clearly increased risk of adverse developmental and obstetrical complications (low birth weight,
intrauterine growth retardation, spontaneous abortion, placenta separation, fetal and neonatal
death, etc.) with serum creatinine levels of 1.4 mg/dl and above (Bear, 1976, 1978; Cunningham
et al., 1990; Abe, 1996; lungers et al., 1997). Applying the linear regression analysis shown in
Appendix C of the IRIS Toxicological Review (U.S. EPA, 2004a), a serum creatinine level of
1.4 mg/dl corresponds to a GFR of 37.2 mL/(min/1.73 m2).5 Similarly, the average serum
creatinine level of 0.8 mg/dl in the same population (nonpregnant women) corresponds to a GFR
of 79.4 mL/(min/1.73 m2). Dividing 79.4 by 39.8 yields a physiological AFj^ of 2.00, which is
identical to the sigma-method AF^ derived previously. This comparison is based on an
assumption that the ratio of normal nonpregnant GFR to adverse GFR holds for the increased
GFR values during pregnancy. There is considerable uncertainty in the regression model in the
estimate of the lower GFR values, which is not accounted for in the physiological estimate of
AFjjK, however. Also, the severity of the low-GFR effects and the proportion of the population
that would be affected is unclear. Overall, the clinical data supporting the physiological
approach are too far removed from the direct assessment needed to establish AF^ and serve
only as support for the assessment. Therefore, the selection of a lower bound 3 SD from the
mean GFR in healthy pregnant women in the statistical approach does not seem excessive and
would appear to be adequately protective. Thus, in Equation 5.1, AF^ is assigned a value of
2.0, and AF^, remains at its default value of 10
,0.5
5GFR values are corrected for body surface area in this study.
Proposal Draft
Boron — September, 2006 8-12
-------
Table 8-1 Sigma-method Value Calculation
Study
Dunlop(1981)
Krutzenetal. (1992)
Sturgissetal. (1996)
Averages
Mean GFR (SD)
mL/min
150.5 (17.6)b
195 (32)c
138.9 (26.1)d
161.5
Mean GFR-3SD
97.7
99
60.6
85.8
Sigma-Method Value3
1.54
1.97
2.29
1.93
a Mean GFR - (Mean GFR - 3 SD)
b Serially-averaged observations across three time periods (16, 26, and 36 weeks) for 25 pregnant women
c Third trimester values for 13 pregnant women
d Serially-averaged observations across two time periods (early and late pregnancy) for 21 pregnant women (basal
index plus basal control individuals)
8.1.3.3 Summary of Data-Derived Adjustment Factors and RfD Calculation
Table 8-2 demonstrates the division of UFA and UFHinto toxicokinetic and
toxicodynamic components and indicates the default values (in parentheses) and the data-derived
values used to replace default toxicokinetic values.
Table 8-2 Default and Data-derived Values for Components of UFA and UFH
Uncertainty Factor
UFA
UFH
Component
TD
(3.16)
not replaced
(3.16)
not replaced
TK*
(3.16)
3.3
(3.16)
2.0
Combined UFA and UFH
Combined Factor Values
10.5
6.3
66
* Valuation of the TK component of UFAwas based on species difference in the volume of total body water during
pregnancy and boron clearance rates; valuation of the TK component of UFR was based on differences in GFR
among pregnant women.
The RfD is calculated from Equation 5.1, where:
Dc= 10.3 mg/kg-day (Allen et al., 1996)
AFAK = 3.3 (data-derived)
AFAD = 3.16 (100.5, default)
= 2.0 (data-derived)
, = 3.16 (10as, default)
UF = 1 (UFS x UFD x UFL)
AFTOT = 3.3x2.0x3.16x3.16 = 66
RfD = 10.3/66 = 0.2 mg/kg-day
The RfD is consistent with a suggestion by Nielsen (1992) that an intake of 10 mg per
day is not too high, while 50 mg/day is probably toxic. If a representative body weight of 60 kg
Proposal Draft
Boron — September, 2006
8-13
-------
is assumed for a pregnant woman, the value of 10 mg/day translates to 0.17 mg/kg-day. As
boron appears to have some beneficial nutrient value, Nielsen (1992) also recommended a total
daily boron intake of 1 mg to avoid boron deficiency. The RfD would appear to give an
adequate margin of safety below, as well as above.
8.1.3.4 Other Uncertainty Factor Approaches
Other researchers and regulatory concerns have used different methods to derive
uncertainty factors. The U.S. EPA has not yet endorsed any of these approaches, as there are a
number of critical, unresolved scientific and methodological issues.
The International Program on Chemical Safety (IPCS) uses "data-derived" uncertainty
factors to estimate tolerable intake values (WHO, 1994; Renwick, 1993). This method allows
for subdivision of each of the interspecies and intraspecies default uncertainty factors to
incorporate data on toxicokinetics (pharmacokinetics) or toxicodynamics (pharmacodynamics).
For interspecies uncertainty, the 10-fold factor is divided into a default factor of 1006 (4.0) for
toxicokinetics and 1004 (2.5) for toxicodynamics in the absence of toxicokinetic and
toxicodynamic data. For intraspecies uncertainty, the 10-fold factor is subdivided into a default
of 1005 (3.2) each for toxicokinetics and toxicodynamics in the absence of toxicokinetic and
toxicodynamic data. Subsequently, the International Program for Chemical Safety (IPCS, 2001)
published a guidance document on the use of data to develop chemical specific adjustment
factors. This guidance calls for the use of a composite factor (CF), which is the composite of
specific adjustment factors (quantitative chemical specific data) for either toxicokinetics or
toxicodynamics and the remaining default uncertainty factors for which chemical specific data
were not available. The guidance document states that in some cases the split between
toxicokinetics and toxicodynamics in the framework may not be appropriate and some flexibility
in approach may need to be maintained; however, in the absence of data, the defaults for
interspecies toxicokinetics and toxicodynamics are 4.0 and 2.5, respectively. This subdivision,
according to the authors, was based on the approximate 4-fold difference between rats and
humans in basic physiological parameters that are major determinants of clearance and
elimination of chemicals, such as cardiac output and renal and liver blood flows. The defaults
for interindividual toxicokinetics and toxicodynamics are each 3.2. In addition to the IPCS
approach, a number of risk assessments have recently been completed for boron using an
uncertainty factor less than 100. A description of the critical effect chosen and the uncertainty
factors used follows. ECETOC (1994) developed a tolerable daily intake (TDI) for
developmental effects of boron. Decreased fetal body weight in rats was chosen as the critical
effect (Price et al., 1994) with an NOAEL of 9.6 mg B/kg-day. A factor of 10as was chosen for
interspecies uncertainty factor due to the similarity in pharmacokinetics (metabolism and
distribution were cited) between animals and humans. A default factor of 10 was chosen for the
intraspecies uncertainty factor. The composite uncertainty factor was 30.
Murray (1995, 1996) used the Price et al. (1994) study, choosing decreased fetal body
weight in rats as the critical effect with an NOAEL of 9.6 mg B/kg-day. The interspecies
uncertainty factor chosen was 4 (2 for pharmacokinetics and 2 for pharmacodynamics, 2x2=4).
Several reasons were cited for the reduced interspecies uncertainty factor for pharmacokinetics:
boron is not metabolized in animals or humans, eliminating a major potential source of
pharmacokinetic variation; it is rapidly distributed throughout body water and does not
Proposal Draft
Boron — September, 2006 8-14
-------
accumulate; the toxicity profile of boron is similar across species; and parameters of elimination
were considered by the author to be similar in humans and other animals. The authors cited the
following reasons for the reduced interspecies uncertainty factor for pharmacodynamics: the
sensitivity of the target tissue receptor appeared to be similar across species based on the
similarity of symptoms of acute toxicity in animals and humans, and developmental and
reproductive toxicity appear to be the most sensitive endpoints of toxicity in all animal species
tested. The intraspecies uncertainty factor chosen was 8 (2.5 for pharmacokinetics and 3.2 for
pharmacodynamics). The intraspecies pharmacokinetic factor was decreased because
metabolism is normally the major source of pharmacokinetic variance in humans, and borates are
not metabolized. The composite uncertainty factor chosen was 4x8=32.
IEHR (1997) determined an unlikely effect level for developmental toxicity for boron
based on the benchmark dose for decreased fetal body weight by Allen et al. (1996). The
interspecies uncertainty factor chosen for boron was 10as, which includes 10a25 each for
pharmacokinetics and pharmacodynamics. The justification for these other-than-default values
was stated as the variability in the intrinsic sensitivity of the target site (embryo, testis, ovary) to
the chemical's toxic effects in humans versus that in the experimental animal and metabolic and
pharmacokinetic differences among species. The intraspecies uncertainty factor chosen for
boron was a default value of 10. The composite human sensitivity factor was 30.
In Environmental Health Criteria, WHO (1998a) developed a TDI for boron, using
decreased fetal body weight in rats as the critical effect (Price et al., 1994), with an NOAEL of
9.6 mg B/kg-day. The interspecies uncertainty factor chosen was 10as (10ai x!0a4= 10as) which
used a 10°'l for pharmacokinetics due to the similarity of absorption, distribution, metabolism,
and elimination of boron in rats and humans and a 1004 (default) for pharmacodynamics. The
intraspecies uncertainty factor chosen was 10°'9 (10a4 x 10as =10a9), 100.4 for pharmacokinetics
due to lack of metabolism in humans and 1005 (default) for pharmacodynamics. The composite
uncertainty factor was 32.
In Guidelines for Drinking-Water Quality, WHO (1998b) developed a TDI for boron to
set a guidance value for drinking water. Decreased fetal body weight in rats was chosen as the
critical effect (Price et al., 1994) with an NOAEL of 9.6 mg B/kg-day. A default value of 10 was
chosen for the interspecies factor due to a reported lack of data to support reduction in the
pharmacokinetic and pharmacodynamic factors. For intraspecies extrapolation a default value of
3.2 for pharmacokinetic data was reduced to 1.8, and a default value of 3.2 was retained for
pharmacodynamic data. Thus, the uncertainty factor for intraspecies uncertainty was
1.8x3.2=5.7 rounded to 6. The composite uncertainty factor was considered to be 10x6=60.
Dourson et al. (1998), as part of the development of the WHO document (1998b),
developed a TDI for boron. Although the authors agreed to the lack of metabolism and the
similarity in absorption and elimination of boron in animals and humans, interspecies variation
in kinetics for boron was considered to relate to renal clearance rates. A 3-fold clearance rate
difference between rats and humans for boron was estimated, after eliminating studies with little
confidence from an earlier projected 4-fold difference. The calculated renal clearance rate
difference (3-fold) between rats and humans for boron was considered by the authors to be
similar to a 4-fold difference that would be expected of other chemicals (Renwick, 1993). Based
on this difference in clearance rates, the authors (Dourson et al., 1998) chose not to reduce the
Proposal Draft
Boron — September, 2006 8-15
-------
interspecies uncertainty factor for pharmacokinetics or pharmacodynamics. Therefore, a default
value of 10 was chosen for the interspecies factor. For intraspecies uncertainty, the
pharmacokinetic factor was reduced from a default of 3.2 to 1.8. The authors proposed that the
likely difference for humans in boron kinetics occurs during pregnancy and is based on an
increase in the GFR, a recognized physiological adaption during pregnancy. The estimation of
the 1.8 factor for intraspecies variation in pharmacokinetics was based on a ratio of the mean
GFR of 144 mL/min +/- 32(SD) from pooled data of healthy humans in late pregnancy (number
of subjects not mentioned) and this mean GFR minus two standard deviations from the mean to
account for variation in the average to the susceptible human 32(SD) x2=64;
144(GFR)-64(2SDs)=80; the ratio of 1.8 was calculated as 144 mL/min divided by 80=1.8. The
intraspecies pharmacodynamic factor used was a factor of 3.1, which the authors considered as a
default factor, although previous methodology considered it to be 3.2. The intraspecies
uncertainty factor was 1.8x3.1=5.58 rounded to 6. The composite uncertainty factor was
10x6=60.
Murray and Andersen (2001) detailed the use of reduced uncertainty factors for boron
risk assessments in recent years and noted the use of factors in the range of 25-60 using the
NOAEL from the Price et al. (1996a) rat developmental study. The authors recommended using
data derived uncertainty factors in a range of 22-44 using new rat and human clearance data
(Vaziri et al., 2001; Pahl et al., 2001). The authors detailed a method where they estimated the
human dose expected to provide the same boric acid area under the curve in target tissues as the
NOAEL in rats and then applying reduced uncertainty factors for pharmacokinetic and
pharmacodynamic uncertainty to this estimated human NOAEL. Interspecies pharmacokinetic
value was estimated at 3.1, while interspecies pharmacodynamic uncertainty was estimated at
1.25-2.5. Intraspecies factors for pharmacokinetics were 1.8-2.0 and intraspecies
pharmacodynamics were 3.2.
The IOM (2002) developed a tolerable upper intake level (UL) for various life stages of
humans. These ULs were based on the NOAEL (9.6 mg/kg-day) from Price et al. (1996a) and an
uncertainty factor of 30 (10 for interspecies uncertainty and 3 for intraspecies uncertainty based
on the similarity in pharmacokinetics among humans). The reference body weight for adult
women was 61 kg and was based on an average body weight from different female age groups.
The resulting UL for adults was rounded to 20 mg/day. The UL was set at 17 mg B/day for
pregnant women of 14-18 years of age, while theUL for pregnant women of 19-50 years of age
was set at 20 mg B/day.
8.1.4 Previous Oral Assessment
The previous RfD for boron on IRIS was 9E-2 mg/kg-day based on testicular atrophy and
spermatogenic arrest in a 2-year dog study from Weir and Fisher (1972). The NOAEL was 8.8
mg/kg-day, the LOAEL was 29 mg/kg-day and the uncertainty factor was 100. Newer studies
have identified developmental effects in three species. The newer RfD is based on the critical
effect of decreased fetal body weight in rats. The NOAEL of 9.6 mg/kg-day was identified from
Price et al. (1996a) and the LOAEL of 13.3 mg/kg-day was identified from Heindel et al. (1992).
Decreased fetal body weight was chosen from these studies because they are quality studies with
a sensitive endpoint that identified the lowest pair of NOAELs and LOAELs. Developmental
effects in mice and rabbits occurred at higher doses. The RfD uses data from these two studies
Proposal Draft
Boron — September, 2006 8-16
-------
performed in the same laboratory and is based on a BMDL05 from (Allen et al., 1996). With the
exception of the NOAEL from Weir and Fisher (1972) in dogs, reproductive effects occurred at
higher doses than the developmental NOAEL and LOAEL. The Weir and Fisher (1972) study in
dogs was not chosen due to the quality of the study (Section 7).
8.2 Inhalation Reference Concentration (RfC)
The minimal database needed for development of an RfC is considered to be a well-
conducted inhalation study that has adequately evaluated a comprehensive array of endpoints,
including the respiratory tract and established an NOAEL and an LOAEL (U.S. EPA, 1994b).
This criterion was not met for boron. No RfC could be derived, due to insufficiencies of the
database.
8.3 Cancer Assessment
The available data are inadequate for evaluation of the human carcinogenic potential of
boron. Derivation of slope factors and unit risks is, therefore, precluded.
8.4 CCL Health Reference Level
The EPA reference dose (RfD) for boron is 0.2 mg/kg/day (U.S. EPA, 2004d) based on
developmental effects in rats from two studies (Price et al., 1996a; Heindel et al., 1992). The
RfD was derived using the benchmark dose (BMD) method (Allen et al., 1996). As described in
Section 4.3.2, EPA established the Health Reference Level (HRL) for boron (1.4 mg/L or 1,400
|ig/L) using the RfD of 0.2 mg/kg-day and a 20 percent relative source contribution.
Proposal Draft
Boron — September, 2006 8-17
-------
9.0 REGULATORY DETERMINATION AND CHARACTERIZATION OF RISK
FROM DRINKING WATER
9.1 Regulatory Determination for Chemicals on the CCL
The Safe Drinking Water Act (SOWA), as amended in 1996, required the U.S. EPA to
establish a list of contaminants to aid the Agency in regulatory priority setting for the drinking
water program. The U.S. EPA published a draft of the first Contaminant Candidate List (CCL)
on October 6, 1997 (62 FR 52193, U.S. EPA, 1997b). After review of and response to
comments, the final CCL was published on March 2, 1998 (63 FR 10273, U.S. EPA, 1998c).
On July 18, 2003, the U.S. EPA announced final Regulatory Determinations for one
microbe and 8 chemicals (68 FR 42897, U.S. EPA, 2003) after proposing those determinations
on June 3, 2002 (67 FR 38222, U.S. EPA, 2002b). The remaining 40 chemicals and ten microbial
agents from the first CCL became was renamed CCL 2 and were published in the Federal
Register on April 2, 2004 (69 FR 17406, U.S. EPA 2004c).
The SDWA requires the U.S. EPA to make regulatory determinations for no fewer than
five contaminants from CCL 2 by August 2006. In cases where the Agency determines that a
regulation is necessary, the regulation should be proposed by August 2008 and promulgated by
February 2010. The Agency is given the freedom to determine that there is no need for a
regulation if a chemical on the CCL fails to meet one of three criteria established by the SDWA
and described in section 9.1.1.
9.1.1 Criteria for Regulatory Determination
These are the three criteria used to determine whether or not to regulate a chemical on
the CCL:
The contaminant may have an adverse effect on the health of persons.
• The contaminant is known to occur or there is a substantial likelihood that the
contaminant will occur in public water systems with a frequency and at levels of
public health concern.
• In the sole judgment of the Administrator, regulation of such contaminant presents a
meaningful opportunity for health risk reduction for persons served by public water
systems.
The findings for all criteria are used in making a determination to regulate a contaminant.
As required by the SDWA, a decision to regulate commits the U.S. EPA to publication of a
Maximum Contaminant Level Goal (MCLG) and promulgation of a National Primary Drinking
Water Regulation (NPDWR) for that contaminant. The agency may determine that there is no
need for a regulation when a contaminant fails to meet one of the criteria. A decision not to
regulate is considered a final Agency action and is subject to judicial review. The Agency can
choose to publish a Health Advisory (a nonregulatory action) or other guidance for any
contaminant on the CCL independent of the regulatory determination.
Proposal Draft
Boron — September, 2006 9-1
-------
9.1.2 National Drinking Water Advisory Council Recommendations
In March 2000, the U.S. EPA convened a Working Group under the National Drinking
Water Advisory Council (NDWAC) to help develop an approach for making regulatory
determinations. The Working Group developed a protocol for analyzing and presenting the
available scientific data and recommended methods to identify and document the rationale
supporting a regulatory determination decision. The NDWAC Working Group report was
presented to and accepted by the entire NDWAC in July 2000.
Because of the intrinsic difference between microbial and chemical contaminants, the
Working Group developed separate but similar protocols for microorganisms and chemicals.
The approach for chemicals was based on an assessment of the impact of acute, chronic, and
lifetime exposures, as well as a risk assessment that includes evaluation of occurrence, fate, and
dose-response. The NDWAC protocol for chemicals is a semi-quantitative tool for addressing
each of the three CCL criteria. The NDWAC requested that the Agency use good judgment in
balancing the many factors that need to be considered in making a regulatory determination.
The U.S. EPA modified the semi-quantitative NDWAC suggestions for evaluating
chemicals against criteria for the regulatory determination criteria and applied them in decision-
making. The quantitative and qualitative factors for boron that were considered for each of the
three criteria are presented in the sections that follow.
9.2 Health Effects
The first criterion asks if the contaminant may have an adverse effect on the health of
persons. Because all chemicals have adverse effects at some level of exposure, the challenge is
to define the dose at which adverse health effects are likely to occur, and estimate a dose at
which adverse health effects are either not likely to occur (threshold toxicant), or have a low
probability for occurrence (non-threshold toxicant). The key elements that must be considered
in evaluating the first criterion are the mode of action, the critical effect(s), the dose-response for
critical effect(s), the RfD for threshold effects, and the slope factor for nonthreshold effects.
A full description of the health effects associated with exposure to boron is presented in
Chapter 7 of this document and summarized below in Section 9.2.2. Chapter 8 and Section 9.2.3
present summarizes dose-response information.
9.2.1 Health Criterion Conclusion
The available toxicological data indicate that boron has the potential to cause adverse
health effects in humans and animals. However, data from human studies were inadequate to
determine if the major effects of boron toxicity seen in animal studies, in which the developing
fetus and the testes were the most sensitive targets, can be interpolated to humans exposed to
boron. The RfD was based on developmental studies in rats.
9.2.2 Hazard Characterization and Mode of Action Implications
Proposal Draft
Boron — September, 2006 9-2
-------
The National Academy of Science Institute of Medicine (IOM, 2001) categorizes boron
as a possible trace mineral nutrient for humans. Boron is essential for plant growth and
deficiency studies in animals and humans have provided some evidence that low intakes of
boron affects cellular function and the activity of other nutrients. It may interact with Vitamin D
and calcium homeostasis, influence estrogen metabolism, and play a role in cognitive function
(IOM, 2001). The average dietary intake for from the 1994-1996 USDA Continuing survey of
Food Intake by Individuals is 1.06 mg/day (IOM, 2001).
Some human oral data are available from cases where boron was ingested for medical
reasons. When the amount ingested was less than 3.68 mg/kg, subjects were asymptomatic,
while doses of 20 and 25 mg/kg resulted in nausea and vomiting. Case reports and surveys of
accidental poisonings indicate that the lethal doses of boron are range from 15 to 20 grams
(approximately 200 to 300 mg/kg) for adults, 5 to 6 grams (approximately 70 to 85 mg/kg) for
children, and 2 to 3 grams (approximately 30 to 45 mg/kg) for infants (U.S. EPA, 2004a).
There is a single occupational study of 6 workers from a group of 28 exposed to high
concentrations of boron (boric acid) aerosols (22-80 mg/m3) that reported testicular effects,
consistent with the testicular effects reported in oral animal studies. However, these data are
considered of limited value for risk determination, due to sparse details and small sample size. In
a far larger study, no effect on fertility was found in U.S. borate production workers; but
exposure concentrations were much lower in this study (about 2.23 mg/m3 sodium borate or 0.31
mg B/m3).
Acute irritant effects are well documented in human workers exposed to borates,
primarily at concentrations greater than 4.4 mg/m3. However, there is no evidence for reduced
pulmonary function in workers with chronic exposure. Boric acid and borates are distributed
evenly throughout the soft tissues in the body water, and are not metabolized. Accordingly,
there is no reason to expect route-specific differences in systemic targets. There may be
route-specific differences in ability to deliver toxic doses to the targets, in that very high
exposure concentrations may be required to produce effects by inhalation exposure.
Portal-of-entry effects may also differ with exposure route.
The primary effects seen in animals after chronic exposure to boron at low-effect doses
generally involve the testes and developing fetus. Chronic effects of dietary boron exposure in
two-year studies included testicular atrophy and spermatogenic arrest in dogs, decreased food
consumption, suppressed growth, and testicular atrophy in rats, and decreased survival, testicular
atrophy, and interstitial cell hyperplasia in mice. Although researchers observed some increases
in tumor incidences in the liver and in subcutaneous tissues in mice (NTP, 1987), based on
comparisons to historic controls, these tumors were determined not to be associated with
exposure to boron from boric acid. The chronic mouse study conducted by Schroeder and
Mitchener (1975) was inadequate to detect carcinogenicity because only one very low dose level
was used (0.95 mg B/kg-day), and the MTD was not reached. No inhalation cancer data were
located. Studies of boron compounds for genotoxicity were overwhelmingly negative, including
studies in bacteria, mammalian cells, and mice in vivo. Accordingly, the Agency determined that
there are inadequate data to assess the human carcinogenic potential for boron.
Proposal Draft
Boron — September, 2006 9-3
-------
In developmental studies with rats, mice, and rabbits, oral exposure to boric acid resulted
in decreased pregnancy rate, increased prenatal mortality, decreased fetal weights, and increased
malformations in fetuses and pups. However, these reproductive effects were associated with
maternal toxicity including changes in maternal organ weights, body weights, weight gain, and
increased renal tubular dilation and/or regeneration (Price et al., 1990, 1994, 1996a; Heindel et
al., 1992, 1994; Field et al., 1989). Reproductive effects in males were noted in the subchronic
and chronic studies described in the preceding paragraphs.
9.2.3 Dose-Response Characterization and Implications in Risk Assessment
The EPA RfD for boron is 0.2 mg/kg/day (U.S. EPA, 2004d) based on developmental
effects in rats from two studies (Price et al., 1996a; Heindel et al., 1992). The RfD was derived
using the benchmark dose (BMD) method (BMDL05 from Allen et al., 1996) using a data derived
uncertainty factor of 66. Allen et al. (1996) concluded that decreased fetal body weight was the
most suitable endpoint for developing a point of departure, because the benchmark doses
calculated for the other endpoints (incidence of total malformations, enlarged lateral ventricles in
the brain, shortening of rib XIII, and variations of the first lumbar rib) were higher. EPA
established the HRL for boron using the RfD of 0.2 mg/kg-day and a 20 percent relative source
contribution. The HRL is calculated to be 1.4 mg/L or 1,400 |ig/L.
9.3 Occurrence in Public Water Systems
The first criterion necessitates evaluation of the contaminant to determine if it may have an
adverse effect on the health of persons. The second criterion necessitates evaluation of the
contaminant to determine if there is a substantial likelihood that it will occur in public water
systems with a frequency and at levels of public health concern. In order to address this criterion
the following information was considered:
• Monitoring data from public water systems
• Ambient water concentrations and releases to the environment
• Environmental fate
Data on the occurrence of boron in public drinking water systems were used to evaluate
the second criterion. The U.S. EPA looked at the total number of systems that reported
detections of boron, as well those that reported concentrations of boron above an estimated
drinking-water health reference level (HRL). For noncarcinogens, the estimated HRL was
calculated from the RfD assuming that 20% of the total exposure would come from drinking
water. For carcinogens, the HRL was the 10"6 risk level (i.e, the probability of 1 excess tumor in
a population of a million people). The HRLs are benchmark values that were used in evaluating
the occurrence data while the risk assessments for the contaminants were being developed.
The available monitoring data on occurrence in drinking water, including indications of
whether or not the contaminant is a national or a regional problem, are included in Chapter 4 of
this document and summarized below. Additional information on production, use, and fate are
found in Chapters 2 and 3.
Proposal Draft
Boron — September, 2006 9-4
-------
9.3.1 Occurrence Criterion Conclusion
The available data for boron indicate its ubiquitous presence in the ambient environment.
Boron, as a naturally occurring element, was detected in many ambient waters, fish tissues, and
stream bed sediments. In addition, approximately 81.9% of groundwater PWSs had detections
of boron (minimum reporting level, MRL, of 0.005 mg/L). These detections affected about
88.1% of the population served by the PWSs, equivalent to approximately 75.5 million people
served by ground water nationally. Nevertheless, the frequency of boron occurrence at levels of
public health concern was relatively low. Concentrations in drinking water exceeded the HRL in
only approximately 1.7% of surveyed groundwater PWSs, affecting only about 0.4% of the
population served, equivalent to approximately 0.4 million people. Supplementary data from an
AWWARF-sponsored study indicate that boron contamination of surface water is less significant
than boron contamination of ground water. Of 228 ground water and 113 surface water samples
analyzed, boron was detected in 99.1% of the ground water samples and 97.3% of the surface
water samples. Boron was detected at concentrations greater than the HRL in only 3.1% of the
ground water samples and in none of the surface water samples. The data indicate that, although
boron is frequently found in the ambient environment and finished drinking water systems, little
to no boron at levels of public health concern is detected in most finished drinking water
systems.
9.3.2 Monitoring Data
Drinking Water
Approximately 81.9% of groundwater PWSs had detections of boron (>minimum
reporting level, >MRL, or >0.005 mg/L). These detections affected about 88.1% of the
population served by the PWSs, equivalent to approximately 75.5 million people served by
ground water nationally. Detections at a concentration greater then one-half the health reference
level (^HRL or >0.7 mg/L) occurred in 4.3% of surveyed PWSs, affecting 2.9% of the
population served, equivalent to approximately 2.5 million people nationally. Concentrations
greater than the HRL (>HRL or >1.4 mg/L) were found in approximately 1.7% of surveyed
PWSs, affecting 0.4% of the population served, equivalent to approximately 0.4 million people
nationally.
Supplementary data from an AWWARF-sponsored study indicate that boron
contamination of surface water is less significant than boron contamination of ground water. Of
228 ground water and 113 surface water samples analyzed, boron was detected in 99.1% of the
ground water samples and 97.3% of the surface water samples. Boron was detected at a
concentration greater then one-half the health reference level (>/^HRL or >0.7 mg/L) in 8.8% of
the ground water samples and none of the surface water samples. Boron was detected at
concentrations greater than the HRL (>HRL or >1.4 mg/L) in 3.1 % of the ground water samples
and in none of the surface water samples.
Ambient Water
Although boron is a naturally-occurring element that is widespread in nature, regional
ambient water data for boron were available from only two studies. In ground water from the
Sacramento Valley aquifer, boron was detected in all thirty-one samples at concentrations
ranging from 12 |ig/L to 1100 |ig/L. The median concentration was 42 |ig/L. Two of the thirty-
Proposal Draft
Boron — September, 2006 9-5
-------
one samples had concentrations in excess of 600 |ig/L (Dawson, 2001). In the lower Illinois
River Basin, 71% of ground water samples collected between 1984 and 1991 contained boron
concentrations higher than the minimum reporting level of 50 |ig/L. The highest detected
concentration was 2100 |ig/L. Higher boron concentrations were generally found in deeper and
more ancient aquifers (Warner, 1999).
9.3.3 Use and Fate Data
In 2003 the United States was the world's largest producer of refined boron compounds
with about one-half of the domestic production exported. Borax (hydrous or anhydrous) and
boric acid are widely used for a wide range of industrial applications. The principal uses for
boron compounds in the United States in 2001 were estimated as follows: 78% glass and
ceramics; 6% soaps and detergents; 4% agriculture; 3% flame retardants; and 9% as other boron-
containing products. The use pattern for borax in its decahydrate, pentahydrate, and anhydrous
forms was: 23% in insulation glass fibers; 20% in household cleaning products as germicide;
11% in borosilicate glasses; 11% as algicide in water treatment; 8% in enamel flux, frits, and
glazes; 8% as chemical intermediate for perborates; 7% in fertilizers; 5% as antifreeze corrosion
inhibitor; 4% as a chemical intermediate for other boron compounds; 3% in herbicides; 1% as
flame retardant and metallurgical flux; and 10% in other miscellaneous applications (HSDB,
2003a). Overall borate uses in 1985 were estimated as follows: 18% glass fiber insulation; 11%
textile glass fiber; 15% chemical fire retardants; 5% borosilicate glass; 4% soap and detergents;
13% miscellaneous; and 44% exports (HSDB, 2003a).
Boron enters the environment primarily through weathering of rocks containing boron
minerals, boric acid volatilization from seawater, and volcanic activity. Anthropogenic inputs
are lower than those from natural processes. Atmospheric boron usually exists as particulates;
therefore, particle size and weight determine the half-life of boron-containing particulates in
ambient air. Boron and boron-containing compounds in aqueous environments adsorb onto iron
and aluminum hydroxy compounds and clay minerals; this is a pH-dependent process with basic
conditions favoring the adsorption. Borate ion and boric acid establish an equilibrium in water
systems according to pH, with dissolved boric acid predominating at pHs below 9.3. In water
and soil, boron adsorbs to particulates high in amorphous aluminum oxide, iron oxide, clay, and
to a lesser extent, organic matter. Again, equilibria are pH-dependent and boron adsorption is
greatest under basic conditions (pH 7-9). Boron requires high pH and electron rich environments
associated with these particulates to form covalent bonds.
9.4 Risk Reduction
The third criterion used to determine if a contaminant requires regulation, states that "in
the sole judgment of the Administrator, regulation presents a meaningful opportunity for health
risk reduction for persons served by public water systems." In evaluating this criterion, the U.S.
EPA conducted an analysis of the total exposed human population, inclusive of sub-populations
exposed to levels above the estimated HRL. Estimates of the population exposure levels were
derived from monitoring data. These estimates are presented in Chapter 4 and summarized in
section 9.4.2.
Proposal Draft
Boron — September, 2006 9-6
-------
The U.S. EPA conducted an analysis which considered the exposure to boron from
drinking water relative to the total known environmental exposures from all media, to determine
if drinking water regulation could significantly reduce health risks. The findings are discussed in
Section 9.4.3 below.
In making its regulatory determination, the U.S. EPA also evaluated effects on
potentially sensitive populations including the fetus, infants and children; a brief description is
given in section 9.4.4.
9.4.1 Risk Criterion Conclusion
Nationally, approximately 2.5 million people consume water from groundwater PWS
where boron detections exceeded one-half the HRL of 0.7 mg/L and approximately 0.4 million
people consume water from groundwater PWS were detections exceeded the HRL. Mouse, rat,
and rabbit studies indicate that the developing fetus is sensitive to boron. Individuals with
severely impaired kidney function constitute a sensitive population since the kidney is the main
route of boron excretion. The U.S. EPA determined that health risk from boron exposure from
public water systems is small, even for sensitive populations, and therefore promulgation of a
boron regulation does not present a meaningful opportunity for health risk reduction.
9.4.2 Exposed Population Estimates
Nationally, approximately 81.9% of groundwater PWSs had detections of boron
(>minimum reporting level, >MRL, or >0.005 mg/L). Therefore, about 88.1% of the population
served by the surveyed groundwater PWSs is exposed to boron in drinking water; this
population is equivalent to approximately 75.5 million people. Detections at a concentration
greater than one-half the health reference level (>/^HRL or >0.7 mg/L) occurred in 4.3% of
surveyed groundwater PWSs, indicating that 2.9% of the population served, equivalent to
approximately 2.5 million people, are exposed to this level of boron Concentrations greater than
the HRL (>HRL or >1.4 mg/L) were found in approximately 1.7% of surveyed groundwater
PWSs, indicating that exposure at this level occurs in 0.4% of the population served, equivalent
to approximately 0.4 million people.
9.4.3 Relative Source Contribution
Relative source contribution analysis was conducted to compare the estimated magnitude
of exposure expected via in the general population from drinking water to the magnitude of
exposure from intake of boron in other media, such as magnitude of exposure from other media,
including food, air, and soil. The highest average boron exposure is from food and next from
water. Using the median concentration of boron in water from Table 4-1 of 0.047 mg/L, and an
daily water intake of 2 L/day average exposure from drinking water would be 0.094 mg/day. It
is reported that average daily boron intake in normal adult human diets ranges 0.87 to 1.35
mg/day (IOM, 2001). Thus, the average contribution of boron exposure from food is about
tenfold greater than that from water. A combination of the 99th percentile concentration in water
(2.44 mg/L x 2 L/day= 4.88 mg/day) with the CSFII 99th percentile value for foods (2.97
mg/day; IOM, 2001) is 56% of the 14 mg/day allowance for a 70 kg adult derived from the RfD.
Proposal Draft
Boron — September, 2006 9-7
-------
Children can potentially ingest significant amounts of boron via hand-to-mouth contact,
especially when concentrations in residential areas are naturally high in soil and where boron-
containing pesticides are applied in and around homes. Workers in boron-related industry are
subject to high boron exposure, but not from drinking water, as in the case of children.
9.4.4 Sensitive Populations
Studies in rats, mice, and rabbits identify the developing fetus as potentially sensitive to
boron. Price et al. (1996a) identified an LOAEL of 13.3 mg/kg-day and an NOAEL of 9.6
mg/kg-day in the developing fetus, based on decreased fetal body weight in rats. Accordingly,
boron at concentration greater than the HRL might have an effect on prenatal development.
Males may also be susceptible to testicular effects from boron exposures during development
(Weir and Fisher, 1972).
Individuals with impaired renal function may have an increased risk following exposure.
Preeclampsia can be a common complication of pregnancy in which renal function declines,
including glomerular filtration. This may increase boron retention, leading to elevated exposures
for the mother and fetuses.
9.5 Regulatory Determination Decision
As stated in Section 9.1.1, a positive finding for all three criteria is required in order to
make a determination to regulate a contaminant. In the case of boron, the only positive finding
is for the health effects criterion, and data are conclusive solely in animal studies. Ingestion of
boron may exert adverse effects on human health; however, based on monitoring conducted in
the 1980's, the frequency of occurrence and concentration levels of boron in drinking water are
believed insufficient to pose any appreciable public health concerns at the present time. Based
on low level of occurrence in regulated public water systems, coupled to its ubiquitous and on
its natural occurrence in the environment from natural sources and scarcity of any known
adverse public health effects, regulating boron in drinking water will not present a meaningful
opportunity for health risk reduction for persons served by public water systems.
Proposal Draft
Boron — September, 2006 9-8
-------
10.0 REFERENCES
Abe, S. 1996. Pregnancy in glomerulonephritic patients with reduced renal function. Hypertens.
Pregnancy 15:305-312 (as cited in U.S. EPA, 2004a).
Alexander, G.V., R.E. Nusbaum, andN.S. MacDonald. 1951. The boron and lithium content of
human bones. J. Biol. Chem. 192:489-496 (as cited in U.S. EPA, 2004a).
Allen, B.C., P.L. Strong, CJ. Price, et al. 1996. Benchmark dose analysis of developmental
toxicity in rats exposed to boric acid. Fund. Appl. Toxicol. 32:194-204 (as cited in U.S. EPA,
2004a).
Anderson, D.L., W.C. Cunningham, and T.R. Lindstrom. 1994. Concentrations and intakes of H,
B, S, K, Na, Cl, and NaCl in foods. J. Food. Comp. Anal. 7:59-82 (as cited in U.S. EPA, 2004a).
Astier, A., F. Baud, and A. Fournier. 1988. Toxicokinetics of boron after a massive accidental
ingestion of boric acid. J. Pharm. Clin. 7:57-62 (as cited in U.S. EPA, 2004a).
ATSDR (Agency for Toxic Substances and Disease Registry). 1992. Toxicological Profile for
Boron and Compounds; TP-91/05. Available from ATSDR, Atlanta, GA.
Baker, M.D. and S.C. Bogema. 1986. Ingestion of boric acid by infants. Am. J. Emerg. Med.
4(4):358-361 (as cited in U.S. EPA, 2004a).
Bakke, J.P. 1991. Evaluation of the potential of boric acid to induce unscheduled DNA synthesis
in the in vitro hepatocyte DNA repair assay using the male F-344 rat [unpublished study].
Submitted by U.S. Borax Corp; MRID No. 42038903 (as cited in U.S. EPA, 2004a).
Bear, R.A. 1976. Pregnancy in patients with renal disease. Obstet. Gynecol. 48:13-18 (as cited in
U.S. EPA, 2004a).
Bear, R.A. 1978. Pregnancy in patients with chronic renal disease. CMA J. 118:663-665 (as cited
in U.S. EPA, 2004a).
Benson, W.H., WJ. Birge, and H.W. Dorough. 1984. Absence of mutagenic activity of sodium
borate (borax) and boric acid in the Salmonella preincubation test. Environ. Toxicol. Chem.
3:209-214 (as cited in U.S. EPA, 2004a).
Bertine K.K. and E.D. Goldberg. 1971. Fossil fuel combustion and major sedimentary cycle.
Science 173:233 - 235 (as cited in ATSDR, 1992).
Beyer, K.H., F.W. Bergfeld, W.O. Berndt, et al. 1983. Final report on the safety assessment of
sodium borate and boric acid. J. Am. Coll. Toxicol. 2(7):87-125 (as cited in U.S. EPA, 2004a).
Bingham, F.T., A.L. Page, N.T. Coleman, et al. 1971. Boron adsorption characteristics of
selected amorphous soils from Mexico and Hawaii. Soil Sci. Soc. Am. Proc. 35:546-550 (as
cited in ATSDR, 1992).
Proposal Draft
Boron — September, 2006 10-1
-------
Birmingham, DJ. and M.M. Key. 1963. Preliminary survey, U.S. Borax Plant, Boron, CA
(February 20, 1963). Occupational Health Research and Training Facility, Division of
Occupational Health, Public Health Service, U.S. Dept. Of Health, Education and Welfare,
Cincinnati, OH (as cited in U.S. EPA, 2004a).
Bonn, B.A. 1999. Selected elements and organic chemicals in bed sediment and fish tissue of the
Tualatin River Basin, Oregon, 1992-96. U.S. Geological Survey Water-Resources Investigations
Report 99-4107. Available from: .
Link to document from: .
Brown, K.W., G. B. Evans, Jr., and B.D. Frentrup (eds.). 1983. Hazardous Waste Land
Treatment. Boston, MA: Butterworth Publishers, p. 211 (as cited in HSDB, 2003a,c,d).
Brown, T.F., M.E. McCormick, D.R. Morris, et al. 1989. Effects of dietary boron on mineral
balance in sheep. Nutr. Res. 9:503-512 (as cited in U.S. EPA, 2004a).
Butterwick, L., N. de Oude, and K. Raymond. 1989. Safety assessment of boron in aquatic and
terrestrial environments. Ecotoxicol. Environ. Safety 17:339-371 (as cited in ATSDR, 1992).
Chapin, R.E. and W.W. Ku. 1994. The reproductive toxicity of boric acid. Environ. Health
Perspect. 102(Suppl. 7):87-91 (as cited in U.S. EPA, 2004a).
Chapin, R.E., W.W. Ku, M.A. Kenney, et al. 1997. The effects of dietary boron on bone strength
in rats. Fund. Appl. Toxicol. 35:205-215 (as cited in U.S. EPA, 2004a).
Chemfinder.com. 2004. Available on-line at: http://chemfinder.cambridgesoft.com/.
Cherrington, J.W. and N. Chernoff 2002. Periods of vertebral column sensitivity to boric acid
treatment in CD-I mice in utero. Repro. Toxicol. 16:237-243 (as cited in U.S. EPA, 2004a).
Clarke, W.B. and R.S. Gibson. 1988. Lithium, boron and nitrogen in 1-day diet composites and a
mixed-diet standard. J. Food. Comp. Anal. 1:209-220 (as cited in U.S. EPA, 2004a).
Clark, G.M. and T.R. Maret. 1998. Organochlorine compounds and trace elements in fish tissue
and bed sediments in the lower Snake River Basin, Idaho and Oregon. U.S. Geological Survey
Water-Resources Investigations Report 98-4103. Available from:
. Link to document from:
.
Clayton, G.D. and F.E. Clayton (eds.). 1994. Patty's Industrial Hygiene and Toxicology: Volume
2A, 2B, 2C: Toxicology. 4rd ed. New York: John Wiley Sons. pp. 4413-4426 (as cited in HSDB,
2003a, c,d).
Cox, D. and D. Lindley. 1974. Theoretical Statistics. London: Chapman and Hall (as cited in
U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-2
-------
Craan, A.G., A.W. Myres, and D.W. Green. 1997. Hazard assessment of boric acid in toys. Reg.
Tox. Pharm. 26:271-280 (as cited in U.S. EPA, 2004a).
Culver, B.D., P.T. Shen, T.H. Taylor, et al. 1994. The relationship of blood- and urine-boron to
boron exposure in borax-workers and the usefulness of urine-boron as an exposure marker.
Environ. Health Perspect. 102(Suppl. 7): 133-137 (as cited in U.S. EPA, 2004a).
Culver, B.D. and S.A. Hubbard. 1996. Inorganic boron health effects in humans: an aid to risk
assessment and clinical judgement. J. Trace Elem. Exp. Med. 9:175-184 (as cited in U.S. EPA,
2004a).
Cunningham, F.G., S.M. Cox, and T.W. Harstad. 1990. Chronic renal disease and pregnancy
outcome. Am. J. Obste. Gynecol. 163:453-459 (as cited in U.S. EPA, 2004a).
Davidson, I.W.F., J.C. Parker, and R.P. Beliles. 1986. Biological basis for extrapolation across
mammalian species. Reg. Tox. Pharmacol. 6:211-237 (as cited in U.S. EPA, 2004a).
Dawson, B.J.M. 2001. Ground-water quality in the southeastern Sacramento Valley aquifer,
California. U.S. Geological Survey Water-Resources Investigations Report 01-4125. Available
from: . Link to document from:
.
Demerec, M., G. Bentani, and J. Flint. 1951. A survey of chemicals for mutagenic action onE.
co/7. Am. Nat. 84(821): 119-136 (as cited in U.S. EPA, 2004a).
De Vette, H.Q.M., C. Otto, and J.A. Schoonmade. 2001. A study on the identification and
comparison of the dissociation byproducts of Polybor tech., Borax, manufacturing grade and
boric acid, manufacturing grade in aqueous solutions using Raman spectrometry. TNO Nutrition
and Food Research, Delft, The Netherlands; TNO Report V99.1153 (as cited in U.S. EPA,
2004a).
Devirian, T.A. and S.L. Volpe. 2003. The physiological effects of dietary boron. Crit. Rev. Food.
Sci. Nutr. 43(2):219-231 (as cited in U.S. EPA, 2004a).
Dieter, M.P. 1994. Toxicity and carcinogenicity studies of boric acid in male and female
B6C3F1 mice. Environ. Health Perspect. 102(Suppl. 7):93-97 (as cited in U.S. EPA, 2004a).
Dixon, R.L., IP. Lee, and RJ. Sherins. 1976. Methods to assess reproductive effects of
environmental chemicals. Studies of cadmium and boron administered orally. Environ. Health
Perspect. 13:59-67 (as cited in U.S. EPA, 2004a).
Dixon, R.L., RJ. Sherins, and IP. Lee. 1979. Assessment of environmental factors affecting
male fertility. Environ. Health Perspect. 30:53-68 (as cited in U.S. EPA, 2004a).
Dourson, M., A. Maier, B. Meek, et al. 1998. Boron tolerable intake re-evaluation of
toxicokinetics for data derived uncertainty factors. Biol. Trace Elem. Res. 66(l-3):453-463 (as
cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-3
-------
Draize, J.H. and E.A. Kelley. 1959. The urinary excretion of boric acid preparations following
oral administration and topical applications to intact and damaged skin of rabbits. Toxicol. Appl.
Pharmacol. 1:267-276 (as cited in U.S. EPA, 2004a).
Dunlop, W. 1981. Serial changes in renal haemodynamics during normal human pregnancy. Br.
J. Obstet. Gynecol. 88:1-9 (as cited in U.S. EPA, 2004a).
ECETOC (European Centre for Ecotoxicology and Toxicology of Chemicals). 1994.
Reproductive and general toxicology of some inorganic borates and risk assessment for human
beings. Technical Report No. 65. Brussels: European Centre for Ecotoxicology and Toxicology
of Chemicals (as cited in U.S. EPA, 2004a).
Eckel W.P. and W.D. Langley. 1988. A background-based ranking technique for assessment of
elemental enrichment in soils at hazardous waste sites. In Superfund '88: Proceedings of the 9th
National Conference, Nov 28-30, 1988, Washington, DC. pp. 286-288 (as cited in ATSDR,
1992).
Elwell, M. 1993. Letter to C. Smallwood, U.S. EPA, Cincinnati, OH. March 5 (as cited in U.S.
EPA, 2004a).
Emsley, J. 1989. The Elements. Oxford: Clarendon Press (as cited in U.S. EPA, 2004a).
Fail, P.A., J.D. George, T.B. Grizzle, et al. 1990. Final report on the reproductive toxicity of
boric acid (CAS No. 10043-35-3) in CD-I-Swiss mice. National Toxicology Program, Public
Health Service, U.S. Department of Health and Human Services, Research Triangle Park, NC;
NTP Report 90-105 (as cited in U.S. EPA, 2004a).
Fail, P. A., J.D. George, J.C. Seely, et al. 1991. Reproductive toxicity of boric acid in CD-I Swiss
mice: Assessment using the continuous breeding protocol. Fund. Appl. Toxicol. 17:225-239 (as
cited in U.S. EPA, 2004a).
Fail, P.A., R.E. Chapin, CJ. Price, et al. 1998. General, reproductive, developmental, and
endocrine toxicity of boronated compounds. Reprod. Toxicol. 12:1-18 (as cited in U.S. EPA,
2004a).
Field, E.A., CJ. Price, M.C. Marr, et al. 1989 Final report on the developmental toxicity of
boric acid (CAS No. 10043-35-3) in CD-I-Swiss Mice. National Toxicology Program, Public
Health Service, U.S. Department of Health and Human Services, Research Triangle Park, NC;
NTP Final Report No. 89-250 (as cited in U.S. EPA, 2004a).
Forbes, R. M., A.R. Cooper, and H.H. Mitchell. 1954. On the occurrence of beryllium, boron,
cobalt and mercury in human tissues. J. Biol. Chem. 209:857-865 (as cited in U.S. EPA, 2004a).
Forbes, R.M. and H.H. Mitchell. 1957. Accumulation of dietary boron and strontium in young
adult albino rats. Arch. Ind. Health. 16:489-492 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-4
-------
Frey, M.M., C. Seidel, M. Edwards, et al. 2004. Occurrence Survey for Boron and Hexavalent
Chromium. AwwaRF Report 91044F.
Friis-Hansen, B., B. Aggerbeck, and J.A. Jansen. 1982. Unaffected blood boron levels in
newborn infants treated with a boric acid ointment. Food Chem. Toxicol. 20:451-454 (as cited in
U.S. EPA, 2004a).
Garabrant, D.H., L. Bernstein, J.M. Peters, et al. 1984. Respiratory and eye irritation from boron
oxide and boric acid dusts. J. Occup. Med. 26:584-586 (as cited in U.S. EPA, 2004a).
Garabrant, D.H., L. Bernstein, J.M. Peters, et al. 1985. Respiratory effects of borax dust. Br. J.
Ind. Med. 42:831-837 (as cited in U.S. EPA, 2004a).
Oilman, A.G., T.W. Rail, A.S. Nies, and P. Taylor, (eds.). 1990. Goodman and Oilman's The
Pharmacological Basis of Therapeutics. 8th ed. New York: Pergamon Press (as cited in HSDB,
2003b).
Graedel, T.E. 1978. Inorganic elements, hydrides, oxides, and carbonates. In: Chemical
Compounds in the Atmosphere. New York, NY: Academic Press, pp. 35-49 (as cited in ATSDR,
1992).
Hamilton, P.A., T.L. Miller, and D.N. Myers. 2004. Water Quality in the Nation's Streams and
Aquifers: Overview of Selected Findings, 1991-2001. USGS Circular 1265. Available from:
. Link to document from:
>http://water.usgs.gov/pubs/circ/2004/1265/>.
Haworth, S., T. Lawlor, K. Mortelmans, et al. 1983. Salmonella mutagenicity test results for 250
chemicals. Environ. Mutagen (Suppl.)l:3-142 (as cited in U.S. EPA, 2004a).
Heindel, J.J., CJ. Price, E.A. Field, et al. 1992. Developmental toxicity of boric acid in mice and
rats. Fund. Appl. Toxicol. 18:266-277 (as cited in U.S. EPA, 2004a).
Heindel, J.J., CJ. Price, and B.A. Schwetz, BA. 1994. The developmental toxicity of boric acid
in mice, rats and rabbits. Environ. Health Perspect. 102(Suppl. 7): 107-112 (as cited in U.S. EPA,
2004a).
HSDB (Hazardous Substance Data Bank). 2003a. Boron. Division of Specialized Information
Services, National Library of Medicine. Available from: .
HSDB (Hazardous Substance Data Bank). 2003b. Boric acid. Division of Specialized
Information Services, National Library of Medicine. Available from:
.
HSDB (Hazardous Substance Data Bank). 2003c. Borax. Division of Specialized Information
Services, National Library of Medicine. Available from: .
Proposal Draft
Boron — September, 2006 10-5
-------
HSDB (Hazardous Substance Data Bank). 2003d. Sodium tetraborate (anhydrous borax).
Division of Specialized Information Services, National Library of Medicine. Available from:
.
HSDB (Hazardous Substance Data Bank). 2003e. Boron oxide. Division of Specialized
Information Services, National Library of Medicine. Available from:
.
Hunt, C.D. 1994. The biochemical effects of physiologic amounts of dietary boron in animal
nutrition models. Environ. Health Perspect. 102(Suppl. 7):35-43 (as cited in U.S. EPA, 2004a).
Hunt, C.D. 1996. Biochemical effects of physiological amounts of dietary boron. J. Trace El em.
Exp. Med. 9:185-213 (as cited in U.S. EPA, 2004a).
IEHR (Institute for Evaluating Health Risks). 1997. An assessment of boric acid and borax using
the IEHR evaluative process for assessing human developmental and reproductive toxicity of
agents. Reprod. Toxicol. 11:123-160 (as cited in U.S. EPA, 2004a).
IOM (Institute of Medicine). 2002. Dietary Reference Intakes for Vitamin A, Vitamin K,
Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Molybdenum, Nickel Silicon,
Vanadium and Zinc. Washington, DC: National Academy Press (as cited in U.S. EPA, 2004a).
IPCS (International Programme on Chemical Safety). 2001. Guidance Document for the Use of
Data on Development of Chemical-specific Adjustment Factors (CSAF) for Interspecies
Differences and Human Variability in Dose/concentration Response Assessment. Geneva,
Switzerland: World Health Organization (as cited in U.S. EPA, 2004a).
Ischii, Y., N. Fujizuka, T. Takahashi, et al. 1993. A fatal case of acute boric acid poisoning. Clin.
Toxicol. 31(2):345-352 (as cited in U.S. EPA, 2004a).
lyengar, G.V., W.B. Clarke, R.G. Downing, et al. 1988. Lithium in biological and dietary
materials. Trace Elem. Anal. Chem. Med. Biol. 5:267-269 (as cited in U.S. EPA, 2004a).
Iyer, V.N. and W. Szybalski. 1958. Two simple methods for the detection of chemical mutagens.
Appl. Microbiol. 6:23-29 (as cited in U.S. EPA, 2004a).
Jansen, J.A., J. Andersen, and J.S. Schou. 1984a. Boric acid single dose pharmacokinetics after
intravenous administration to man. Arch. Toxicol. 55:64-67 (as cited in U.S. EPA, 2004a).
Jansen, J.A., J.S. Schou, and A. Aggerback. 1984b. Gastrointestinal absorption and in vitro
release of boric acid from water emulsifying ointments. Food Chem. Toxicol. 22:49-53 (as cited
in U.S. EPA, 2004a).
Job, C. 1973. Absorption and excretion of orally administered boron. Z. Angew
Bader-Klimahelik 20:137-142 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-6
-------
lungers, P., D. Chauveau, G. Choukroun, et al. 1997. Pregnancy in women with impaired renal
function. Clin. Nephrol. 47(5):281-288 (as cited in U.S. EPA, 2004a).
Kavlock, R.J., B.C. Allen, E,M. Faustman, et al. 1995. Dose response assessments for
developmental toxicity: IV. Benchmark doses for fetal weight changes. Fund. Appl. Toxicol.
26:211-222 (as cited in U.S. EPA, 2004a).
Kent, N.L. and R.A. McCance. 1941. The absorption and excretion of "minor" elements by man.
I. Silver, gold, lithium, boron and vanadium. Biochem. J. 35:837-844 (as cited in U.S. EPA,
2004a).
Keren, R. and U. Mezuman. 1981. Boron adsorption by clay minerals using a phenomenological
equation. Clays and Clay Minerals 29:19-204 (as cited in ATSDR, 1992).
Keren, R., R.G. Gast, and B. Bar-Yosef 1981. pH-dependent boron adsorption by Na-
montmorillonite. Soil Sci. Soc. Am. J. 45:45-48 (as cited in ATSDR, 1992).
Keren, R. and H. Talpaz. 1984. Boron adsorption by montmorillonite as affected by particle size.
Soil Sci. Soc. Am. J. 48:555-559 (as cited in Keren et al., 1994).
Keren, R. and F.T. Bingham. 1985. Boron in water, soils and plants. In: Steward, B.E. (ed.).
Advances in Soil Science, pp. 229-276 (as cited in Keren et al., 1994).
Keren, R., P.R. Gross, and D.L. Sparks. 1994. Equilibrium and kinetics of borate
adsorption-desorption on pyrophyllite in aqueous suspensions. Soil Sci. Soc. Am. J.
58:1116-1122.
Keren, P. and D.L. Sparks. 1994. Effect of pH and ionic strength on boron adsorption by
pyrophyllite. Soil Sci. Soc. Am. J. 58:1095-1100 (as cited in Keren et al., 1994).
Kirk-Othmer. 1984. Encyclopedia of Chemical Technology. 3rd ed., Volumes 1-26. New York,
NY: John Wiley and Sons. p. 4(78) 76 (as cited in HSDB, 2004).
Kroschwitz, J.I. andM. Howe-Grant. 1992. Kirk-Othmer Encyclopedia of Chemical
Technology. 4th ed., Volumes 1-26. New York, NY: John Wiley and Sons. pp. 360-409 (as cited
in HSDB, 2003a,b,c,d,e).
Krutzen, F., P. Olofsson, S.E. Back, et al. 1992. Glomerular filtration rate in pregnancy; a study
in normal subjects and in patients with hypertension, preeclampsia and diabetes. Scand. J. Clin.
Lab. Invest. 52:387-392 (as cited in U.S. EPA, 2004a).
Ku, W.W., R.E. Chapin, R.F. Moseman, et al. 1991. Tissue disposition of boron in male Fischer
rats. Toxicol. Appl. Pharmacol. 111:145-151 (as cited in U.S. EPA, 2004a).
Ku, W.W., R.E. Chapin, R.N. Wine, et al. 1993a. Testicular toxicity of boric acid (BA):
relationship of dose to lesion development and recovery in the F344 rat. Reprod. Toxicol.
7:305-319 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-7
-------
Ku, W.W., L.M. Shih, and R.E. Chapin. 1993b. The effects of boric acid (BA) on testicular cells
in culture. Reprod. Toxicol. 7:321-331 (as cited in U.S. EPA, 2004a).
Ku, W.W. and R.E. Chapin. 1994. Mechanism of the testicular toxicity of boric acid in rats. In
vivo and in vitro studies. Environ. Health Perspect. 102(Suppl. 7):99-105 (as cited in U.S. EPA,
2004a).
Landolph, J.R. 1985. Cytotoxicity and negligible genotoxicity of borax and borax ores to
cultured mammalian cells. Am. J. Ind. Med. 7:31-43 (as cited in U.S. EPA, 2004a).
Lee, IP., RJ. Sherins, and R.L. Dixon. 1978. Evidence for induction of germinal aplasia in male
rats by environmental exposure to boron. Toxicol. Appl. Pharmacol. 45:577-590 (as cited in U.S.
EPA, 2004a).
Levey, A.S., R.D. Perrone, and N.E. Madaias. 1988. Serum creatinine and renal function. Ann.
Rev. Med. 39:465-490 (as cited in U.S. EPA, 2004a).
Linder, R.E., L.F. Strader, and G.L. Rehnberg. 1990. Effect of acute exposure to boric acid on
the male reproductive system of the rat. J. Toxicol. Environ. Health 31:133-146 (as cited in U.S.
EPA, 2004a).
Litovitz, T.L., W. Klein-Schwartz, G.M. Oderda, et al. 1988. Clinical manifestations of toxicity
in a series of 784 boric acid ingestions. Am. J. Emerg. Med. 6:209-213 (as cited in U.S. EPA,
2004a).
Locksley, H.B. and W.H. Sweet. 1954. Tissue distribution of boron compounds in relation to
neutron-capture therapy of cancer. Proc. Soc. Exp. Biol. Med. 86:56-63 (as cited in U.S. EPA,
2004a).
Magour, S., P. Schramel, J. Ovcar, et al. 1982. Uptake and distribution of boron in rats:
interaction with ethanol and hexobarbital in the brain. Arch. Environ. Contam. Toxicol.
11:521-525 (as cited in U.S. EPA, 2004a).
Malins, D.C., B.B. McCain, D.W. Brown, et al. 1984. Chemical pollutants in sediments and
disease of bottom-dwelling fish in Puget Sound, Washington. Environ. Sci. Technol.18:705-713
(as cited in ATSDR, 1992).
Mann, D.L. 1988. Analysis of boron in contaminated shrimp by inductively coupled plasma
spectroscopy. Spectroscopy 3(3):37-39 (as cited in HSDB, 2003a).
Meister, R.T. (ed.) 2001. Farm Chemicals Handbook. Willoughby, OH: Meister Publishing Co.
pp. B13, C-57 (as cited in HSDB, 2003b,c).
Mertz, W. 1993. Essential trace metals: new definitions based on new paradigms. Nutr. Rev.
51(10):287-295 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-8
-------
Moore, J.W. 1991. Inorganic Contaminants of Surface Water: Research and Monitoring
Priorities. New York, NY: Springer-Verlag.
Murray, FJ. 1995. A human health risk assessment of boron (boric acid and borax) in drinking
water. Reg. Toxicol. Pharmacol. 22:221-230 (as cited in U.S. EPA, 2004a).
Murray, FJ. 1996. Issues in boron risk assessment: pivotal study, uncertainty factors and ADI's.
J. Trace Elem. Exp. Med. 9:231 (as cited in U.S. EPA, 2004a).
Murray, FJ. 2002. The dose-dependent contribution of renal tubular reabsorption in determining
the renal elimination of compounds primarily eliminated by glomerular filtration. Prepared for
U.S. Environmental Protection Agency, National Center for Environmental Assessment,
Washington DC (as cited in U.S. EPA, 2004a).
Murray, F J. and M.E. Andersen. 2001. Data-derived uncertainty factors: boric acid (BA) as a
case study. Human Eco. Risk Assess. 7(1): 125-138 (as cited in U.S. EPA, 2004a).
Naghii, M.R. and S. Samman. 1996. The effect of boron supplementation on the distribution of
boron in selected tissues and on testosterone synthesis in rats. Nutr. Biochem. 7:507-512 (as
cited in U.S. EPA, 2004a).
Naghii, M.R. and S. Samman. 1997. The effect of boron supplementation on its urinary excretion
and selected cardiovascular risk factors in healthy male subjects. Biol. Trace Element Res.
56:273-286 (as cited in U.S. EPA, 2004a).
Narotsky, M.G., N. Wery, B.T. Hamby, et al. 2003. Effects of boric acid on hox gene expression
and the axial skeleton in the developing rat. In: Massaro, E J. and J.M. Rogers (eds.). The
Skeleton: Biochemical, Genetic and Molecular Interactions in Development and Homeostasis.
Totowa, NJ: Humana Press (in press) (as cited in U.S. EPA, 2004a).
National Fire Protection Association. 1997. Fire Protection Guide to Hazardous Materials. 12th
Edition, NFPA Publications. National Fire Protection Association, 1 Batterymarch Park, Quincy,
MA. pp.491-508 (as cited in HSDB, 2003a,b,c,e).
Nielsen, F.H. 1970. Percutaneous absorption of boric acid from boron-containing preparations in
rats. Acta Pharmacol. Toxicol. 28:413-424 (as cited in U.S. EPA, 2004a).
Nielsen, F.H. 1991. Nutritional requirements for boron, silicon, vanadium, nickel, and arsenic:
Current knowledge and speculation. FASEB J. 5:2661-2667 (as cited in U.S. EPA, 2004a).
Nielsen, F.H. 1992. Facts and fallacies about boron. Nutr. Today 27:6-12 (as cited in U.S. EPA,
2004a).
Nielsen, F.H. 1994. Biochemical and physiologic consequences of boron deprivation in humans.
Environ. Health Perspect. 102(Suppl. 7):59-63 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-9
-------
Nielsen, F.H. 1996. Evidence for the nutritional essentiality of boron. J. Trace Elem. Exp. Med.
9:215-229 (as cited in U.S. EPA, 2004a).
Nielsen, F.H, C.D. Hunt, L.M. Mullen, et al. 1987. Effect of dietary boron on mineral, estrogen,
and testosterone metabolism in post-menopausal women. FASEB J. 1:394-397.
NRC (National Research Council). 1983. Risk Assessment in the Federal Government:
Managing the Process. Washington, DC: National Academy Press.
NRC (National Research Council). 1989. Recommended dietary allowances. 10th ed.
Washington, DC: National Academy Press, p. 267 (as cited in U.S. EPA, 2004a).
Nriagu, J.O. 1979. Copper in the atmosphere and precipitation. In: Nriagu, J.O. (ed.). Copper in
the Environment. Part I: Ecological Cycling. New York, N Y: John Wiley and Sons, Inc. (as
cited in ATSDR, 1992).
NTP (National Toxicology Program). 1987. Toxicology and carcinogenesis studies of boric acid
(CAS No. 10043-35-3) in B6C3F1 mice (feed studies). Public Health Service, U.S. Department
of Health and Human Services; NTP TR-324. Available from:
(as cited in U.S. EPA, 2004a).
Odunola, O.A. 1997. Individual and combined genotoxic response of boric acid and aflatoxin Bl
in Escherichia coli PQ37. East Afir. Med. J. 74:499-502 (as cited in U.S. EPA, 2004a).
O'Flaherty, EJ. 1981. Toxicants and Drugs: Kinetics and Dynamics. New York: John Wiley and
Sons, Inc. pp. 237-241 (as cited in U.S. EPA, 2004a).
O'Loughlin, K.G. 1991. Bone marrow erythrocyte micronucleus assay of boric acid in
Swiss-Webster mice [unpublished study]. Submitted by U.S. Borax Corp. MRID No. 42038904
(as cited in U.S. EPA, 2004a).
O'Neil, M.J., A. Smith, and P.E. Heckelman (eds.). 2001. The Merck Index: An Encyclopedia of
Chemicals, Drugs, and Biologicals. 13th ed. Whitehouse Station, NJ: Merck & Co., Inc. pp.1326,
1327, 1333, 8662, and 8664.
Osol, A. (ed.). 1980. Remington's Pharmaceutical Sciences. 16th ed. Easton, Pennsylvania:
Mack Publishing Co. p. 1257 (as cited in HSDB, 2003a,b,c,d,e).
O'Sullivan, K. and M. Taylor. 1983. Chronic boric acid poisoning in infants. Arch. Dis. Child
58:737-739 (as cited in U.S. EPA, 2004a).
Owen, E.C. 1944. The excretion of borate by the dairy cow. J. Dairy Res. 13(3):243-248 (as
cited in U.S. EPA, 2004a).
Pahl, M.V., B.D. Culver, P.L. Strong, et al. 2001. The effect of pregnancy on renal clearance of
boron in humans: a study based on normal dietary intake of boron. Toxicol. Sci. 60(2):252-256
(as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-10
-------
Parks, W.L. and J.L. White. 1952. Boron retention by clay and humus systems saturated with
various cations. Soil Sci. Soc. Am. Proc. 298-300 (as cited in ATSDR, 1992).
Penland, J.G. 1994. Dietary boron, brain function and cognitive performance. Environ. Health
Perspect. 102(7):65-72 (as cited in U.S. EPA, 2004a).
Peterson, D.A. and R.B. Zelt. 1999. Element concentrations in bed sediment of the Yellowstone
River Basin, Montana, North Dakota, and Wyoming—A retrospective analysis. U.S. Geological
Survey Water-Resources Investigations Report 99-4185. Available from:
. Link to document from:
.
Price, C.J., E.A. Field, M.C. Marr, et al. 1990. Developmental toxicity of boric acid (CAS No.
10043-35-3) in Sprague Dawley rats. National Toxicology Program, Public Health Service, U.S.
Department of Health and Human Services, Research Triangle Park, NC; NTP Report No.
90-105 (and Report Supplement No. 90-105A) (as cited in U.S. EPA, 2004a).
Price, C.J., M.C. Marr, C.B. Myers, et al. 1991. Developmental toxicity of boric acid (CAS No.
10043-35-3) in New Zealand White rabbits. National Toxicology Program, Public Health
Service, U.S. Department of Health and Human Services, Research Triangle Park, NC; NTP
TER90003 (and Laboratory Supplement No. TER-90003) (as cited in U.S. EPA, 2004a).
Price, C.J., M.C. Marr, and C.B. Myers. 1994. Determination of the
no-observable-adverse-effect level (NOAEL) for developmental toxicity in Sprague-Dawley
(CD) rats exposed to boric acid in feed on gestational days 0 to 20, and evaluation of postnatal
recovery through postnatal day 21 [final report]. Research Triangle Institute, Center for Life
Science, Research Triangle Park, NC; RTI Identification No. 65C-5657-200 (as cited in U.S.
EPA, 2004a).
Price, C.J., P.L. Strong, M.C. Marr, et al. 1996a. Developmental toxicity NOAEL and postnatal
recovery in rats fed boric acid during gestation. Fund. Appl. Toxicol. 32:179-193 (as cited in
U.S. EPA, 2004a).
Price, C.J., M.C. Marr, C.B. Myers, et al. 1996b. The developmental toxicity of boric acid in
rabbits. Fund. Appl. Toxicol. 34:176-187 (as cited in U.S. EPA, 2004a).
Price, C.J., P.L. Strong, FJ. Murray, et al. 1997. Blood boron concentrations in pregnant rats fed
boric acid throughout gestation. Reprod. Toxicol. 11(6):833-842 (as cited in U.S. EPA, 2004a).
Rai, D., J.M. Zachara, A.P. Schwab, et al. 1986. Chemical attenuation rates, coefficients, and
constants in leachate migration. Vol. 1. A critical review. Report to Electric Power Research
Institute, Palo Alto, CA by Battelle, Pacific Northwest Laboratories, Richland, WA. Research
Project 2198-1 (as cited in ATSDR, 1992).
Rainey C.J., L.A. Nyquist, R.E. Christensen, et al. 1999. Daily boron intake from the American
diet. J. Am. Diet Assoc. 99(3):335-40.
Proposal Draft
Boron — September, 2006 10-11
-------
Renwick, A.G. 1993. Data-derived safety factors for the evaluation of food additives and
environmental contaminants. Food Add. Contam. 10(3):275-30 (as cited in U.S. EPA, 2004a).
Rope, S.K., WJ. Arthur, T.H. Craig, et al. 1988. Nutrient and trace elements in soil and desert
vegetation of southern Idaho. Environ. Monit. Assess. 10:1-24 (as cited in ATSDR, 1992).
Rossoff, IS. 1974. Handbook of Veterinary Drugs. New York: Springer Publishing Company (as
cited in HSDB, 2003a).
Rudd, CJ. 1991. Mouse lymphoma cell mutagenesis assay (tK+/-/tK-/-) of boric acid
[unpublished study]. Submitted by U.S. Borax Corp. MRID No. 420390 (as cited in U.S. EPA,
2004a).
Sakata, M. 1987. Relationship between adsorption of arsenic (III) and boron by soil and soil
properties. Environ. Sci. Technol. 21:1126-1130 (as cited in ATSDR, 1992).
Sax, N.I. and RJ. Lewis, Sr. (eds.). 1987. Hawley's Condensed Chemical Dictionary, llth ed.
New York: Van Nostrand Reinhold Co. (as cited in HSDB, 2003a,b,c,d,e).
Sayli, B.S., E. Tuccar, and A.H. Elhan. 1998. An assessment of fertility in boron-exposed
Turkish subpopulations. Reprod. Toxicol. 12(3):297-304 (as cited in U.S. EPA, 2004a).
Schou, J.S., J.A. Jansen, and B. Aggerbeck. 1984. Human pharmacokinetics and safety of boric
acid. Arch. Toxicol. 7:232-235 (as cited in U.S. EPA, 2004a).
Schroeder, H.A. and M. Mitchener. 1975. Life-term effects of mercury, methyl mercury, and
nine other trace metals on mice. J. Nutr. 105:453-458 (as cited in U.S. EPA, 2004a).
Seal, B.S. and HJ. Weeth. 1980. Effect of boron in drinking water on the male laboratory rat.
Bull. Environ. Contam. Toxicol. 25:782-789 (as cited in U.S. EPA, 2004a).
Secor, R.B. and CJ. Radke. 1985. Spillover of the diffuse double layer on montmorillonite
particles. J. Colloid Interface Sci. 103:237-244 (as cited in Keren et al., 1994).
Seidel, C. 2006. Personal communication to Brent Ranalli at The Cadmus Group, Inc.
[concerning boron data from AWWARF-sponsored study, with data in attached spreadsheet].
Denver, CO: McGuire Malcolm Pirnie. May 19.
Seller, H.G., H. Sigel and A. Sigel (eds.). 1988. Handbook on the Toxicity of Inorganic
Compounds. New York, NY: Marcel Dekker, Inc. p. 130 (as cited in HSDB, 2003a,c,d).
Settimi, L., E. Elovaara, and H. Savolainen. 1982. Effects of extended peroral borate ingestion
on rat liver and brain. Toxicol. Lett. 10:219-223 (as cited in U.S. EPA, 2004a).
Shemesh, O., H. Golbetz, J.P. Kriss, et al. 1985. Limitations of creatinine as a filtration marker
in glomerulopathic patients. Kidney Int. 28:830-838 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-12
-------
Singh, V. and S.P. Singh. 1984. Studies on boron toxicity in lentil and barley. Indian J. Agron.
29:545-546 (as cited in HSDB, 2003c,d).
Skjaerven, R., AJ. Wilcox, and R.T. Lie. 2002. The interval between pregnancies and the risk of
preeclampsia. N. Engl. J. Med. 346(l):33-38 (as cited in U.S. EPA, 2004a).
Smith, H.W. 1951. The kidney structure and function in health and disease. New York: Oxford
University Press, pp 231-238 (as cited in U.S. EPA, 2004a).
Spencer, E. Y. 1982. Guide to the Chemicals Used in Crop Protection. 7th ed. Publication 1093.
Research Institute, Agriculture Canada, Ottawa, Canada: Information Canada, p. 48 (as cited in
HSDB, 2003a,b).
Stewart, K.R. 1991. Salmonella/microtome plate incorporation assay of boric acid [unpublished
study]. Submitted by U.S. Borax Corporation; MRID No. 4203901 (as cited in U.S. EPA,
2004a).
Sturgiss, S.N., R. Wilkinson, and J.M. Davison. 1996. Renal reserve during human pregnancy.
Am. J. Physiol. 271:F16-F20 (as cited in U.S. EPA, 2004a).
Stuttgen, G., T. Siebel, and B. Aggerbeck. 1982. Absorption of boric acid through human skin
depending on the type of vehicle. Arch. Dermatol. Res. 272:21-29 (as cited in U.S. EPA, 2004a).
Swan, S.H., JJ. Beaumont, S.K. Hammond, et al. 1995. Historical cohort study of spontaneous
abortion among fabrication workers in the semiconductor health study: agent-level analysis. Am.
J. Ind. Med. 28:751-769 (as cited in U.S. EPA, 2004a).
Szybalski, W. 1958. Special microbiological system. II. Observations on chemical mutagenesis
in microorganisms. Ann. NY Acad. Sci 76:475-489 (as cited in U.S. EPA, 2004a).
Tarasenko, N.Y., A.A. Kasparov, and O.M. Strongina. 1972. The effect of boric acid on the
reproductive function of the male organism. Gig. Tr. Prof. Zabol. 16(11): 13-16 (as cited in
Whorton et al., 1994b; U.S. EPA, 2004a).
Taylor, M. 1997. Letter to M. Dourson, TERA, Cincinnati, OH. August 28 (as cited in U.S. EPA,
2004a).
Treinen, K.A. and R.E. Chapin. 1991. Development of testicular lesions in F344 rats after
treatment with boric acid. Toxicol. Appl. Pharmacol. 107:325-335 (as cited in U.S. EPA, 2004a).
Ury, H.K. 1966. Interim Report on the 1963 respiratory disease survey at Boron, CA. Air
Pollution Medical Studies Unit, Bureau of Chronic Diseases, California State Department of
Public Health, Sacramento, CA (as cited U.S. EPA, 2004a).
U.S. Borax and Chemical Corporation. 1991. Comments on the draft toxicological profile for
boron. Submitted to ATSDR February 14, 1991 (as cited in ATSDR, 1992).
Proposal Draft
Boron — September, 2006 10-13
-------
U.S. Borax Research Corporation. 1963. MRID No. 00068026; HED Doc. No. 009301.
Available from EPA. Write to FOI, EPA, Washington, DC, 20460 (as cited in U.S. EPA, 2004a).
U.S. Borax Research Corporation. 1966. MRID No. 00005622, 00068021, 00068881; HED Doc.
No. 009301. Available from EPA. Write to FOI, EPA, Washington, DC, 20460 (as cited in U.S.
EPA, 2004a).
U.S. Borax Research Corporation. 1967. MRID No. 00005623, 005624; HED Doc. No. 009301.
Available from EPA. Write to FOI, EPA, Washington, DC, 20460 (as cited in U.S. EPA, 2004a).
U.S. Borax. 2000. UCI Boric acid clearance study reports and associated data: rat and human
studies (as cited in U.S. EPA, 2004a).
U.S. EPA (United States Environmental Protection Agency). 1980. Guidelines and methodology
used in the preparation of health effect assessment chapters of the consent decree water criteria
documents. Fed. Reg. 45(231):79347-79357.
U.S. EPA (United States Environmental Protection Agency). 1986a. Guidelines for the health
risk assessment of chemical mixtures. Fed. Reg. 51(185):34014-34025. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1986b. Guidelines for mutagenicity
risk assessment. Fed. Reg. 51(185):34006-34012. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1987. Toxic air pollutant/source
crosswalk: A screening tool for locating possible sources emitting toxic air pollutants. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards. EPA-450/4-87-023a. (as cited in ATSDR, 1992).
U.S. EPA (United States Environmental Protection Agency). 1988. Recommendations for and
documentation of biological values for use in risk assessment. Prepared by the Office of Health
and Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH
for the Office of Solid Waste and Emergency Response, Washington, DC; EPA 600/6-87/008.
Available from: National Technical Information Service, Springfield, VA; PB-88179874.
U.S. EPA (United States Environmental Protection Agency). 1991. Guidelines for
developmental toxicity risk assessment. Fed. Reg. 56(234):63798-63826. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1994a. Interim policy for particle
size and limit concentration issues in inhalation toxicity: notice of availability. Fed. Reg.
59(206):53799. Available from: .
U.S. EPA (United States Environmental Protection Agency). 1994b. Methods for derivation of
inhalation reference concentrations and application of inhalation dosimetry. Office of
Proposal Draft
Boron — September, 2006 10-14
-------
Environmental Health Assessment, Washington, DC; EPA/600/8-90/066F. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1994c.United States Environmental
Protection Agency. Reregi strati on Eligibility Decision (RED) - Boric Acid and its Sodium Salts.
Washington, DC: Office of Prevention, Pesticides, and Toxic Substances, U.S. EPA. Available
from: .
U.S. EPA (United States Environmental Protection Agency). 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 (United States Environmental Protection Agency). 1996. Guidelines for reproductive
toxicity risk assessment. Fed. Reg. 61(212):56274-56322. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1997a. Exposure Factors
Handbook. Vol. I: General factors. National Center for Environmental Assessment, Office of
Research and Development, Washington, DC; EPA/600/P-95/002Fa. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1997b. Draft Contaminant
Candidate List (CCL). Fed. Reg. 62:52193-52219.
U.S. EPA (United States Environmental Protection Agency). 1998a. Guidelines for neurotoxicity
risk assessment. Fed. Reg. 63(93):26926-26954. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1998b. Science policy council
handbook: peer review. Office of Science Policy for the Office of Research and Development,
Washington, DC; EPA/100/B-98/001. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 1998c. Final Contaminant
Candidate List (CCL). Fed. Reg. 63:10273-10287.
U.S. EPA (United States Environmental Protection Agency). 2000a. Science policy council
handbook: peer review [second edition]. Office of Science Policy for the Office of Research and
Development, Washington, DC; EPA/100/B-OO/OOl. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 2000b. Science policy council
handbook: risk characterization. Office of Science Policy for the Office of Research and
Development, Washington, DC; EPA/100/B-00/002. Available from:
.
Proposal Draft
Boron — September, 2006 10-15
-------
U.S. EPA (United States Environmental Protection Agency). 2000c. United States
Environmental Protection Agency. Benchmark dose technical guidance document [external
review draft]. Risk Assessment Forum, Washington, DC; EPA/630/R-00/001. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 2000d. Supplementary guidance
for conducting health risk assessment of chemical mixtures. Risk Assessment Forum,
Washington, DC; EPA/630/R-00/002. Available from: .
U.S. EPA (United States Environmental Protection Agency). 2000e. Water Industry Baseline
Handbook, 2nd Edition.
U.S. EPA (United States Environmental Protection Agency). 2002a. A review of the reference
dose and reference concentration processes. Risk Assessment Forum Washington, DC;
EPA/630/P-02/002F. Available from: .
U.S. EPA (United States Environmental Protection Agency). 2002b. Announcement of
preliminary regulatory determinations for priority contaminants on the Drinking Water
Contaminant Candidate List. Fed. Reg. 67:38222-38244.
U.S. EPA (United States Environmental Protection Agency). 2003. Announcement of regulatory
determinations for priority contaminants on the Drinking Water Contaminant Candidate List.
Fed. Reg. 68:42897-42906.
U.S. EPA (United States Environmental Protection Agency). 2004a. Toxicological Review of
Boron and Compounds. In: Support of Summary Information on the Integrated Risk Information
System (IRIS). EPA 635/04/052, June, 2004.
U.S. EPA (United States Environmental Protection Agency). 2004b. TRI Explorer: Trends.
Searches for boron trichloride, boron trichloride. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 2004c. Drinking Water
Contaminant Candidate List 2; Notice. Fed. Reg. 69:17406-17415.
U.S. EPA (United States Environmental Protection Agency). 2004d. Integrated Risk Information
System (IRIS) - Boron and Compounds; CASRN 7440-42-8. Available from:
.
U.S. EPA (United States Environmental Protection Agency). 2005. Guidelines for carcinogen
risk assessment. Risk Assessment Forum, Washington, DC; EPA/630/P-03/001B. Available
from: .
U.S. FDA (United Stated Food and Drug Administration). 1988. Title 21. Part 181, Section
181.30. Substances used in the manufacture of paper and paperboard products used in food
packaging. Washington, DC: U.S. Government Printing Office.
Proposal Draft
Boron — September, 2006 10-16
-------
USGS (United States Geological Survey). 2001. Summary publications from 51 NAWQA study
units sampled in 1991-2001. Available from: .
USGS (United States Geological Survey). 2004. Mineral Commodity Summaries, January 2004:
Boron. Available from:
. Link to document
from: .
Usuda, K., K. Kono, K. Nishiuraet al. 1997. Boron diffusion across the dialysis membrane
during hemodialysis. Miner Electrolyte Metab. 23(2): 100-104 (as cited in U.S. EPA, 2004a).
Usuda, K., K. Kono, Y. Orita, et al. 1998. Serum and urinary boron levels in rats after single
administration of sodium tetraborate. Arch. Toxicol. 72:468-474 (as cited in U.S. EPA, 2004a).
Vanderpool, R.A., D. Hof, and P.E. Johnson. 1994. Use of inductively coupled plasma-mass
spectrometry in boron-10 stable isotope experiments with plants, rats, and humans. Environ.
Health Perspect. 102(Suppl. 7): 13-20 (as cited in U.S. EPA, 2004a).
Vaziri, N.D., F. Oveisi, B.D. Culver, et al. 2001. The effect of pregnancy on renal clearance of
boron in rats given boric acid orally. Toxicol. Sci. 60(2):257-263 (as cited in U.S. EPA, 2004a).
Ventura, S.J., W.D. Mosher, S.C. Curtain, et al. 1999. Highlights of trends in pregnancies and
pregnancy rates by outcome: estimates for the United States, 1976-96. Natl. Vital Stat. Rep.
47(29): 1-12. Available from: (as
cited in U.S. EPA, 2004a).
Vignec, A.J. and R. Ellis. 1954. Inabsorbability of boric acid in infant powder. Am. J. Dis. Child
88:72-80.
Waggot, A. 1969. An investigation of the potential problem of increasing boron concentrations
in rivers and water courses. Water Res. 3:749-765 (as cited in ATSDR, 1992).
Warner, K.L. 1999. Analysis of nutrients, selected inorganic constituents, and trace elements in
water from Illinois community-supply wells, 1984-91. U.S. Geological Survey Water-Resources
Investigations Report 99-4152. Available from:
. Link to document from:
.
Weast, R.C. (ed.). 1988-1989. CRC Handbook of Chemistry and Physics. 68th ed. Boca Raton,
FL: CRC Press Inc. (as cited in HSDB, 2003a).
Weed Science Society of America. 1983. Herbicide Handbook. 5th ed. Champaign, Illinois:
Weed Science Society of America, (as cited in HSDB, 2003a,c,d).
Weeth, H.J., C.F. Speth, and D.R. Hanks, DR. 1981. Boron content of plasma and urine as
indicators of boron intake in cattle. Am. J. Vet. Res. 42(3):474-477.
Proposal Draft
Boron — September, 2006 10-17
-------
Wegman, D.H. E.A. Eisen, X. Hu, et al. 1994. Acute and chronic respiratory effects of sodium
borate particulate exposures. Environ. Health Perspect. 102(Suppl. 7): 119-128 (as cited in U.S.
EPA, 2004a).
Weir, RJ. and R.S. Fisher. 1972. Toxicologic studies on borax and boric acid. Toxicol. Appl.
Pharmacol. 23:351-364 (as cited in U.S. EPA, 2004a).
Wery, N., M.G. Narotsky, N. Pacico, et al. 2003. Defects in cervical vertebrae in boric
acid-exposed rat embryos are associated with anterior shifts of hox gene expression domains.
Birth Defects Res. (Part A) 67:59-67 (as cited in U.S. EPA, 2004a).
Wesson, L.G. 1969. Physiology of the human kidney. New York: Grune and Stratton (as cited in
U.S. EPA, 2004a).
WHO (World Health Organization). 1998a. Environmental Health Criteria 204: Boron. Geneva,
Switzerland: World Health Organization (as cited in U.S. EPA, 2004a).
WHO (World Health Organization). 1998b. Guidelines for Drinking Water Quality. Geneva,
Switzerland: World Health Organization (as cited in U.S. EPA, 2004a).
WHO (World Health Organization). 1994. Environmental Health Criteria 170: Assessing Human
Health Risks of Chemicals: Derivation of Guidance Values for Health-based Exposure Limits.
Geneva, Switzerland: World Health Organization (as cited in U.S. EPA, 2004a).
Whorton, D., J. Haas, and L. Trent. 1992. Reproductive effects of inorganic borates on male
employees: birth rate assessment report. Prepared for United States Borax and Chemical
Corporation; Document No. 6966001 (as cited in U.S. EPA, 2004a).
Whorton, D., J. Haas, and L. Trent. 1994a. Reproductive effects of inorganic borates on male
employees: Birth rate assessment. Environ. Health Perspect. 102(Suppl. 7): 129-131 (as cited in
U.S. EPA, 2004a).
Whorton, D., J. Haas, and L. Trent, et al. 1994b. Reproductive effects of sodium borates on male
employees: birth rate assessment. Occup. Environ. Med. 51:761-767 (as cited in U.S. EPA,
2004a).
Wilding, J.L., WJ. Smith, P. Yevich, et al. 1959. The toxicity of boron oxide. Am. Ind. Hyg.
Assoc. J. 20:284-289 (as cited in U.S. EPA, 2004a).
Wong, L.C., M.D. Heimbach, D.R. Truscott, et al. 1964. Boric acid poisoning: report of 11
cases. Can. Med. Assoc. J. 90:1018-1023 (as cited in U.S. EPA, 2004a).
Woods, W.G. 1994. An introduction to boron: history, sources, uses, and chemistry. Environ.
Health Perspect. 102(Suppl. 7):5-l 1 (as cited in U.S. EPA, 2004a).
Woods, W.G. 1996. Review of possible boron speciation relating to its essentiality. J. Trace
Elem. Exp. Med. 9:153-163 (as cited in U.S. EPA, 2004a).
Proposal Draft
Boron — September, 2006 10-18
-------
Worthing, C.R. and S.B. Walker (eds.). 1987. The Pesticide Manual - A World Compendium.
8th ed. Thornton Heath, UK: The British Crop Protection Council (as cited in HSDB, 2003b).
Ysart, G., P. Miller, H. Crews, et al. 1999. Dietary exposure estimated of 30 elements from the
UK Total Diet Study. Food Additives and Contamination 16(9):391-403.
Proposal Draft
Boron — September, 2006 10-19
-------
APPENDIX A: Abbreviations and Acronyms
ABP androgen binding protein
ANO VA analy si s of vari ance
ATSDR Agency for Toxic Substances and Disease Registry
AUC area under the curve
B boron
BA boric acid
BMD benchmark dose, maximum likelihood estimate of dose corresponding to BMR
BMDL the 95% lower confidence limit on the benchmark dose
BMR benchmark response
bw body weight
cAMP cyclic adenosine monophosphate
CAS Chemical Abstracts Registry
CCL Contaminant Candidate List
CFSII Continuing Survey of Food Intakes
CNS central nervous system
CSAF chemical-specific adjustment factors
CV coefficient of variation
ECETOC European Centre for Ecotoxicology and Toxicology of Chemicals
FEVj forced expiratory volume in 1 sec
FR Federal Register
FSH follicle stimulating hormone
FVC forced vital capacity
g gram
gd gestation day
GFR glomerular filtration rate
FIRL health reference level
HSDB Hazardous Sub stances Database
ICPMS inductively coupled plasma-mass spectrometry
IEFIR Institute for Evaluating Health Risks
IOC inorganic compounds
IOM Institute of Medicine
IPCS International Programme on Chemical Safety
IRIS Integrated Risk Information System
kg kilogram
L liter
LH luteinizing hormone
LOAEL lowest observed adverse effect level
m meter
MCLG Maximum Contaminant Level Goal
mg milligram
mL milliliter
MRL minimum reporting level
MTD maximum tolerated dose
NAWQA National Water Quality Assessment
NOW AC National Drinking Water Advisory Council
Proposal Draft
Boron — September, 2006
Appendix A-l
-------
NIOSH National Institute for Occupational Safety and Health
MRS National Inorganic and Radionuclide Survey
NOAEL no observed adverse effect level
NPDWR National Primary Drinking Water Regulation
NTP National Toxicology Program
PA plasminogen activators
pnd postnatal day
ppm parts per million
PWS public water systems
RfC reference concentration
RfD reference dose
SBR standardized birth ratio
SD standard deviation
SDWA Safe Drinking Water Act
TD toxicodynamics
TDI tolerable daily intake
TK toxicokinetics
TRI Toxic Release Inventory
TWA time-weighted average
UCM unregulated contaminant monitoring
UF uncertainty factor
UFA interspecies variability (animal-to-human) uncertainty factor
UFH interindividual variability (sensitive humans) uncertainty factor
UL upper intake level
U.S. FDA U.S. Food and Drug Administration
USGS U.S. Geological Service
U.S. EPA U.S. Environmental Protection Agency
VOC volatile organic compound
WHO World Health Organization
Proposal Draft
Boron — September, 2006
Appendix A-2
------- |