M E M O R A N D U M

To:		William Balke, Perrin Quarles Associates, Inc.

cc:		Margaret Sheppard, U.S. EPA

		Erin Birgfield, U.S. EPA

		

From:		Naida Gavrelis, ERG

Date:		June 8, 2004

Subject:	EPA Contract No. 68W02029, Work Assignment No.108: Analysis of
Health and Environmental Impacts of ODS Substitutes—Evaluating
approaches for setting occupational exposure limits

1.0	Introduction

The purpose of this memorandum is to review two aspects of the approach
used by the U.S. Environmental Protection Agency (EPA) in deriving a
proposed Acceptable Exposure Limit (AEL) for 1-bromopropane/n-propyl
bromide (nPB) as part of its Significant New Alternatives Policy (SNAP).
EPA considered decreased sperm motility observed in the offspring
generation in a multi-generation rat inhalation study the most sensitive
endpoint in available experimental studies, which serves as the basis
for the AEL (Federal Register Vol. 68 No. 106). This review was prompted
by a number of public comments received on two topics:

Selecting the most appropriate data from a two-generational study.
Determining whether it is more appropriate to use data from the
offspring (F1) generation, from the parent (F0) generation, or from
whichever yields a more conservative value. 

Applying uncertainty factors to protect sensitive sub-populations.
Evaluating the most appropriate uncertainty factor (UF) to use to
protect potentially sensitive sub-populations in the working population
(e.g., men with abnormal sperm measures, decreased fertility). 

To accomplish this task, ERG reviewed the scientific literature and
regulations and risk assessment guidance from government agencies and
voluntary workplace standard-setting committees. To the extent possible,
we examined risk assessment approaches used for evaluating other
reproductive toxins under similar exposure situations. However, little
or no precedent has been set on either of the two aspects noted above.
Therefore, much of the discussion that follows describes current
scientific views related to the evaluation of workplace exposures, and
traditional and evolving approaches for addressing uncertainty related
to human variability.

Research on the association between occupational exposure and
reproductive effects is relatively new and significant gaps in knowledge
exist, particularly with respect to the factors responsible for
reproductive toxicity and their mechanism of action (Kumar 2004). In
1996, for example, the National Institute of Occupational Safety and
Health placed reproductive health high on its 10-year research agenda
with recommendations to increase the understanding of fundamental
biological processes underlying normal and abnormal reproductive
function or outcomes (NIOSH 1996). Also in 1996, the American College of
Occupational and Environmental Medicine (ACOEM) recognized that (1) the
magnitude of occupational and environmental reproductive and
developmental health risks was not well characterized, (2) scientific,
epidemiological, and toxicological data on reproductive health risks
were limited or nonexistent for many substances, and (3) industrial
exposure limits (e.g., PELs and TLVs) had been established without
consideration of protection from adverse reproductive or developmental
health effects (ACOEM 1996; Fedoruk 1996). As a result, ACOEM developed
Reproductive Hazard Management Guidelines to serve as a guide for
managing possible occupational reproductive health hazards, especially
in the absence of chemical-specific exposure limits. At the same time,
risk assessment guidance for reproductive toxins was evolving. In 1996,
EPA released its Guidelines for Reproductive Toxicity Risk Assessment.
In 1998, the National Toxicology Program (NTP) established its Center
for the Evaluation of Risks to Human Reproduction (CERHR). CERHR
conducts expert panel reviews and develops monographs describing adverse
reproductive and/or developmental health effects associated with
exposure to environmental and/or occupational chemicals. The center
recently published (October 2003) a monograph for nPB. The importance of
examining reproductive outcomes has clearly been identified, though the
information and knowledge required to interpret available data are still
evolving.

In the absence of complete data sets, regulatory agencies and other
scientific bodies uniformly endorse the application of scientific
judgment as a necessary component in determining the appropriateness of
study data and uncertainties associated with chemical exposures. This
requires weighing what is known and not known and presenting data in a
clear, transparent, and logical way. Such an approach is consistent with
EPA policy and guidance, as well as approaches documented by other
groups developing health guidelines and standards. 

The following sections evaluate the scientific analyses and judgments
related to the study endpoint and UF for human variability used in
deriving the nPB AEL. 

Section 2.0 presents analysis evaluating the use of data from a
two-generation reproductive study and, particularly, the use of the F1
generation data. Use of these data would yield an AEL of 18 parts per
million (ppm), assuming application of the same modeling results,
adjustments, and uncertainty considerations presented in the proposed
rule.

Section 3.0 provides an overview of current scientific views and
approaches when evaluating an appropriate UF for human variability.
Available data seem to lend support for choosing a value of less than
10, but greater than 1 (with 100.5 [or 3] being the conventional
default).

 

2.0	Using Data from a Two-Generation Study 

EPA selected what it considered to be the most sensitive endpoint
observed in the available literature in its evaluation of nPB toxicity.
This approach is consistent with traditional risk assessment practice.
It is also consistent with the data review and interpretations presented
in NTP’s (2003) monograph for nPB. However, EPA’s selection of
decreased sperm motility observed in the offspring in a two-generation
inhalation study in rats (WIL 2001) as the critical endpoint for AEL
development was questioned by some who commented on the proposed rule.
More specifically, choosing to focus on effects observed in the second
generation (F1) was challenged, particularly in the context of setting
an occupational exposure limit.

No prescribed guidance or precedent was identified related to the
appropriateness of using F1 generation data for deriving a health
guideline value or, more specifically, an occupational exposure limit.
NIOSH reports that its exposure limits are based on the most sensitive
toxic endpoint observed in either workers or animals. NIOSH has not
conducted a risk assessment using two-generation reproductive toxicity
data and is not certain how it would address the issue of correlated
responses within a litter. According to NIOSH, a toxicologist and
statistician would presumably examine the data and evaluate the
appropriate model to use based on the available data (Dankovic 2004;
Ahlers 2004).  

General risk assessment practice calls for selecting the dose-response
curve generating the lowest point of departure and offers the strongest
reason for basing the AEL on the F1 data presented in the WIL (2001)
study. In the proposed rule, EPA acknowledges the uncertainty regarding
the mechanism(s) of nPB toxicity and uses this as its primary argument
for using the F1 data. Knowledge of the mechanisms of toxicity would
enhance our understanding of the possible windows of vulnerability to
nPB in the development of the male reproductive system. For example,
alterations in semen quality or fertility may not become evident until
some time after the exposure, particularly if an early stage of male
germ cell development is affected. Without a full understanding of the
mechanism for alterations in male reproductive parameters for nPB and
other chemicals, it does not seem appropriate to exclude the F1 data
from consideration. No data were identified that weakens this argument.
Supporting documentation follows.

In examining data from multi-generation studies and evaluating which
dose-response data to select, it helps to understand the scope and
purpose of reproductive studies in general, as well as the specific
design of the WIL 2001 study. First and foremost, multi-generation
studies are preferred over one-generation studies when evaluating
substance-specific reproductive toxicity (EPA 1996; NAS 2001).
Well-conducted multi-generation reproduction studies determine the
potential of an agent to produce adverse effects on the male and female
reproductive systems, in the embryo and fetus, and in the neonate (NAS
2001). Researchers recognize, however, that the multi-generational study
is among the most complex type undertaken for regulatory purposes (NAS
2001). Because such studies provide information on toxicity that follows
treatment throughout the entire reproductive cycle in both males and
females, the critical window of exposure or the system or mechanism of
action responsible for the observed effect may not be fully understood
(EPA 1996; NAS 2001; Kumar 2004). In addition, sperm evaluations in
smaller test animals (e.g., rats and mice) are limited to the terminal
sacrifice of the test animals, and therefore do not allow for the
assessment of semen parameters (morphology, motility, sperm count) at
successive points. Such factors make interpretation difficult (NAS
2001).

In the case of WIL 2001, male and female F0 rats were exposed at least
70 days prior to mating. Females were not exposed on postnatal days 0-4,
and only the females, not their litters, were exposed during postnatal
days 5-21. Therefore, offspring were exposed to nPB in utero and through
nursing. Those F1 litters selected randomly for propagation were
directly exposed from postnatal day 22 and forward. The results of the
WIL study suggest that both sexes are sensitive to nPB, with male and
female reproductive outcomes both likely contributing to altered
fertility/reduced litter sizes seen at higher doses (WIL 2001, NTP
2002). This is further supported by a more recent study showing ovarian
dysfunction in nPB-exposed rats (Yamada et al. 2003). In fact the Yamada
et al. 2003 study suggests that the reproductive system of F0 females
might be more sensitive to nPB than F1 males. EPA continues to study the
findings of the Yamada study.

The CERHR expert panel that reviewed nPB reproductive toxicity argues
that the consistency of effects observed across the two generations in
the WIL study and lack of significant effects on pubertal indices
suggests that nPB is an adult (not perinatal or juvenile) toxicant with
respect to adverse effects on the reproductive system (NTP 2002).
However, given the unknown mechanism of nPB toxicity and observed
differences in the F0 vs. F1 dose-response curves, the question remains
as to whether spermatic effects seen in F1 rats resulted from pre or
postnatal exposures. Studies looking at 2-bromopropane suggest that
gestational and lactational exposures may impair the development of
reproductive organs (Kang et al. 2002) and 2-bromopropane is believed to
target germ cells such as spermatogonia (Kumar 2004; NTP 2003). Ichihara
et al. (2000) speculated that the mode of nPB toxicity, however, is more
likely failure of spermiation. Because specific mechanisms are not fully
understood, it seems premature to conclude whether or not nPB exposures
could be altering the process of spermatogenesis only after fetal
development in utero. As emphasized by Kumar (2004), mechanisms
underlying reproductive toxicity are complex. More information is needed
to determine whether nPB toxicity is induced by direct effects on the
reproductive organs or indirectly through alterations in hormonal
regulation, or by both. The latter, in particular, could be caused by in
utero exposure. Until such time said information is available, worker
protection is best served by considering the F1 data in the WIL study.

In addition, the study of other chemicals demonstrate the plausibility
of prenatal exposures affecting semen quality in sexually mature men,
with various mechanisms of toxicity hypothesized. Researchers continue
to research specific mechanisms of toxicity and critical windows of
exposure for various chemical types. For example, Fisher et al. (2003)
report that disorders of sperm production may result from abnormal fetal
germ cell production; exposure of rats to dibutyl phthalate (DBP) led to
the abnormal development of Sertoli and Leydig cells, resulting in
testicular abnormalities and preventing the normal production of
testosterone needed to support spermatogenesis after puberty. Similar
conclusions were reached in a study of male rabbits exposed to DBP, with
rabbits considered a better model for humans than rodents because of
their relatively long phase of reproductive development (Higuchi et al.
2003). 

Studies looking at polychlorinated biphenyls (PCBs), dioxins, and some
pesticides show like findings. Though the number of subjects not overly
compelling, a study of Taiwanese boys exposed prenatally to PCBs
revealed increased abnormal morphology, reduced motility, and reduced
capacity to penetrate hamster oocytes (Guo et al. 2000). Significantly
long-lasting effects on mouse spermatogenesis (e.g., reduced sperm head
counts; changes in enzyme levels) were observed in F1 mice following
exposure to lindane in utero—indicative of testicular germ cell
impairment according to the authors (Traina et al. 2003). 

Lastly, in a study of male productive health among approximately 1,700
European men, men with in utero exposure to maternal smoking were shown
to have reduced total sperm counts and sperm concentrations and lower
percentages of motile and morphologically normal sperm cells (Jensen et
al. 2004); authors recognize the need for further evaluation of
confounding factors not accounted for in the present study (e.g., other
prenatal exposures, paternal effects, etc.), but nonetheless the study
provides additional data suggesting chemical insult during fetal
development may affect normal spermatogenesis in the adult male.

3.0	Applying Uncertainty Factors to Protect Sensitive Subpopulations

In the proposed rule for nPB, EPA employed an UF of 3 as an upper bound
to account for possible differences in susceptibility across the worker
population—to protect for potential ‘unobserved’ reproductive
conditions (e.g., sperm motility, aberrant sperm formation) that is
known to exist among otherwise healthy males of working age (Federal
Register Vol. 68 No. 106).

EPA’s choice of an UF of 3 triggered a wide range of public comments.
Some individuals encouraged EPA to use a default factor of 10, stating
that the working population is no different than the general population
when it comes to reproductive effects. On the contrary, others indicated
that the UF should be 1 because of the lack of evidence for a sensitive
sub-population among workers.  

ERG reviewed EPA risk assessment guidelines, occupational exposure
guidelines, and the general scientific literature to provide a framework
for determining the most appropriate UF to account for possible
sensitive sub-populations. The discussion that follows considers the
following factors:

Comparison of EPA risk assessment and occupational exposure guidelines
for evaluating intrahuman variability.

What is known and unknown about variability of reproductive measures in
the healthy male population.

Evolving approaches in uncertainty analysis: weighing intrahuman
variability based on known similarities or differences in toxicokinetics
and toxicodynamics in the population of interest.

Considered individually and collectively, each discussion piece below
lends support for selecting an UF of less than 10, but more than 1. By
convention, an UF of 3 is a reasonable choice to account for possible
increased sensitivities among some workers in the derivation of the AEL
for nPB.  

3.1	General Risk Assessment Approaches: Accounting for Sensitive
Subpopulations

Selecting the most appropriate uncertainty factor in the development of
any health guideline requires scientific judgment. As described below,
established risk assessment guidelines across sectors allow flexibility
in selecting the most appropriate uncertainty factor for the chemical
and exposure situation of interest. Some groups are more prescriptive
than others in specifying approaches related to this and other issues
(in their guidelines and in their chemical-specific evaluations). Both
regulators and non-regulators acknowledge that the choice of the UF will
vary from agent to agent. 

In the absence of prescriptive guidance or precedent for assessing
variability in reproductive parameters among workers, it seems
reasonable that EPA would defer to the well-established risk assessment
framework in addressing uncertainty related to intrahuman variability.
Established risk assessment guidelines call for a default value of 10 to
account for intraspecies differences, but the ultimate decision depends
on the quality of the studies available, the extent of the database, and
scientific judgment (EPA 2002). A greater understanding of the
mechanisms of toxicity has allowed for a reduction in uncertainties
about human variation, leading to decisions to reduce the UF (e.g., from
10 to 3), even among the general population (EPA 2002; Hogan 2000;
Dourson et al. 1996 and 1998; Moore et al. 1995 and 1997). Specific to
setting safe levels of occupational exposure to chemicals, Burin and
Saunders (1999) report that a reduced intraspecies uncertainty factor
(i.e., less than 10) is often used based on the fact that “the exposed
population consists of adults in a reasonably good state of health.” 

Table 1 summarizes documented approaches for addressing intrahuman
variability of various agencies and organizations. As can be seen, even
those entities with more prescribed guidance emphasize the need to
evaluate chemicals on a case-by-case basis, using professional or
scientific judgment. EPA provides a list of considerations when
determining the appropriate UF for reproductive toxins, such as the
specific type of effect, background incidence of that effect, and
toxicokinetic data (EPA 1996), all of which are relevant to the nPB
evaluation. In a recent white paper on risk assessment, the American
Industrial Hygiene Association (AIHA) acknowledges the sources of
scientific uncertainty, pointing to the natural variability in risk
predictors and the lack of knowledge about them; AIHA emphasizes the
importance of ensuring that any assessment is based on sound study data
and that significant gaps in data information exist are clearly
communicated (AIHA 2002). In its guidelines for TLV documentation, ACGIH
calls for describing uncertainty qualitatively and avoiding the use of
the term “factor” in describing uncertainty (ACGIH 2003). ACGIH’s
draft documentation for the TLV for nPB, however, does not specify its
uncertainty considerations. This prohibits a direct comparison with the
approach used by EPA.

NIOSH addresses population variation on a case-by-case basis, but
acknowledges that the age and general health range of workers is
generally more limited than the general population posing less of a
problem in its risk assessments. There have been specific instances
where susceptible populations were considered in the development of a
Recommended Exposure Limit (REL) (e.g., undiagnosed cardiovascular
disease in developing the carbon monoxide REL; workers with existing
respiratory impairments in developing RELs for various oxides of
nitrogen) (Ahlers 2004). NIOSH developed a REL based on reproductive
toxicity for glycol ethers, but does not specifically address human
variation in its criteria document (NIOSH 1991). OSHA applied a 100-fold
factor to address inter- and intra-species variability in its 1993
proposed rule for glycol ethers (Federal Register 54 March 23, 1993),
but no longer employs such an approach in its assessments.

Sections 3.2 and 3.3 describe the type of scientific data that might
weigh into choosing the appropriate UF for intrahuman variability—that
is, the possible variation in susceptibility to reproductive hazards
even among healthy workers and the expected variation in nPB response
among workers in general. 

Variation in Male Reproductive Parameters Across the “Healthy
Worker” Population (Factors Affecting Susceptibility)

The worker population consists of adults considered to be in a
reasonably good state of health, and excludes the very young or old.
Therefore, less variation in susceptibility or sensitivity to chemical
insult is generally presumed across workers as compared to the general
population (Burin and Saunders 1999; Monson 1987). However, EPA argues
in its proposed rule for nPB that when looking at reproductive outcomes
variation might be expected even among “healthy workers” and
possibly making a subpopulation of workers more susceptible to nPB. EPA
indicates that 6% of adult males are infertile Table 1.  Documented
Approaches for Accounting for Intrahuman Variability

Organization	Standard/Guideline	General Approaches Used	Additional Notes

Environmental Health Standards/Guidelines for the General Population

U.S. EPA (1994, 2002)	Reference Concentration	( UF = 10 (default value)

	( Accounts for sensitive subpopulations within the general population,
largely to account for possible variation in the vulnerability of the
very young or very old, the infirm, or other sensitive populations
depending on the chemical and exposure situation of interest. 

( The UF chosen depends on the quality of the studies available, the
extent of the database, and scientific judgment.

ATSDR	Minimal Risk Level



U.S. EPA (1996)	Reproductive Toxicity Risk Assessment	( UF = 3 or 10

( By convention, a value of 3 is used in place of one-half power (i.e.,
100.5) when appropriate

	( The total size of the UF will vary from agent to agent and requires
scientific judgment, taking into account interspecies differences,
variability within species, the slope of the dose-response curve, the
types of reproductive effects observed, the background incidence of the
effects, the route of administration, and toxicokinetic data. 

California EPA (2003)	Air Exposure Limit

(acute and chronic)	( UF = 10 (based on animal data)

( UF = 3 (based on human study performed in healthy adults)

( UF = 1 (assessment includes sensitive human subpopulations)

	Occupational Health Standards/Guidelines

OSHA	Permissible Exposure Limit (PEL)

	( No prescribed guidance. 

( Size of the interval between the PEL and the no-observed-effect levels
or experimentally derived thresholds depends on professional judgment

( Details regarding OSHA’s toxicological assessments are included in
the preambles to many of the OSHA substance-specific standards	( Since
lawsuits in the early 1990s, OSHA has not incorporated “safety
factors” when establishing its PELs.

( OSHA acknowledges the possibility of heterogeneity in workers and that
there may be a wide variation in individual responses to some toxic
substances (e.g., wide range in odor thresholds)

( As stated in the preamble to 29 CFR 1910, Air Contaminants,
case-by-case assessment is used to establish new and revised limits.

NIOSH

(Ahlers 2004)	Recommended Exposure Limit (REL)	( No prescribed guidance.

 	( NIOSH does not specifically apply “safety factors” to data,
although they have defaulted to a factor of 10 in the past.

ACGIH

(2002 and 2003)	Threshold Limit Value (TLV)	( No prescribed guidance.

	( The basis on which TLVs are established may differ from substance to
substance. The TLV reflects uncertainties in the available data;
adjustments are made to reflect an appropriate degree of conservatism,
using professional judgment. 

( Language referring to these adjustments as “factors” should be
avoided; the adjustments do not need to be quantified, but rather
“explained”

AIHA

(2003) 	Workplace Environmental Exposure Limit (WEEL)	( No prescribed
guidance.

( Schematic of WEEL development process accounts for uncertainty,
including consideration of toxicokinetic data. 

	( Available toxicological data are presented in the areas of acute,
subchronic and chronic toxicity, genotoxicity, carcinogenicity,
reproductive toxicity, metabolism and toxicokinetics, and human use and
experience. Data are evaluated along with historical toxicological
perspectives to rationalize development of a WEEL. Evaluations are
accomplished in the same manner as PELs and TLVs®; therefore, the
resultant WEEL should provide a level to which nearly all workers may be
repeatedly exposed, for a working lifetime, without adverse health
effects.

(citing Purves 1992), with 40% to 90% of the infertility due to
deficient sperm production (citing Griffin 1994). 

Risk assessors acknowledge that little basis currently exists for a
priori identification of susceptible subpopulations for many chemicals
(EPA 2002). For reproductive health, however, variations across the
population have been the subject of increasing study. ERG identified
additional studies that document an overall decline in reproductive
health, as well as variation in sperm parameters across the healthy male
population. Study findings are highlighted below, as well as a brief
discussion of implications of these findings on the nPB AEL derivation. 

Trends of declining male reproductive health have been reported in the
Western world, such as increases in testicular cancer incidence, several
indications of decreasing semen quality, and observations of urogenital
abnormalities. Some researchers hypothesize that hormonally active
agents, which are distributed ubiquitously in the environment, could
play an etiologic role, though epidemiological evidence supporting this
theory is lacking  (Bonde and Storgaard 2002; Moline et al. 2000).
Possible mechanisms are not known, although some suggest that potential
changes may be related to disruption of the hormonal regulation of
testicular development in prenatal life or early childhood (Bonde et al.
1998). 

A meta-analysis of 61 previously published studies (1938 to 1990)
reports a world-wide decline in semen quality, with sperm counts varying
dramatically among different geographic locations (Fisch et al. 1996). 

Variation in semen quality (sperm concentration and motility) has been
reported in fertile males across the United States, with lower semen
quality observed in semi-rural or agricultural areas as compared to more
urban areas (Swan 2000 and 2003).

In Denmark, a decline in human sperm count with increasing year of birth
was observed in 1,196 men participating in occupational sperm studies
from 1986 through 1995 (Bonde et al. 1998).

Two additional European studies showed inter- and intra-individual
variability in multiple seminal parameters in healthy subjects
(concentration, motility, morphology, and vitality) (Auger et al. 2000;
Alvarez et al. 2003).

Evaluating the significance of the reported variations (and ultimately
their relevance in selecting an appropriate UF for nPB) is difficult.
Many of the laboratory methods for assessing male reproductive health
have not been standardized, despite guidelines published by the World
Health Organization (Auger et al. 2000; Moline et al. 2000). This means
that different laboratories may generate different results of the same
sample and that the results of a single sample within the same
laboratory may be different if analyzed at different times. For example,
Alvarez et al. (2003) report that the analytical variation in semen
parameters is dependent on the quality control material used and the
level of semen quality. Auger et al. (2000) report that even when
standardized approaches are used, methods remain subjective.
Improvements in laboratory quality control would enable researchers to
better characterize trends in reproductive health (Auger et al 2000;
Moline et al. 2000). The Swan (2003) study specifically reports use of
standardized methods and strict quality control, giving more credence to
the fact that observed geographical variations in semen quality in this
particular study were not influenced by differences or biases in sperm
evaluation. 

Some question also remains about the definition of “normal semen
quality” and how to use semen measures to assess human fertility. For
example, researchers are exploring which semen variables are the most
sensitive to toxicant exposure, and which are the most predictive of
human fertility (Alvarez et al. 2003; Moline et al. 2000). In addition,
biases introduced by differential participation related to age and
fertility cannot be ruled out in the available studies (Larsen et al.
1998; Fisch et al 2003). 

Despite remaining questions, documentation of the overall decline and
variation in semen quality in tested populations provides evidence of
increased susceptibilities in some workers. It is therefore, conceivable
that subpopulations with already reduced semen quality (subfertile or
intermediate fertility) could have a lower threshold to nPB than those
considered fertile (or “normal”). 

Scientific Basis of UF for Intrahuman Variability: Evolving Approaches

Since the early 1990s, several researchers have been examining the
scientific basis and alternate approaches for the application of
uncertainty factors in quantitative risk assessment. In general they
have concluded that the 10-fold default factors are protective of
greater than 99% of the general population (Dourson et al. 1996; Renwick
and Lazarus 1998; Burin and Saunders 1999; WHO 2001). When NOAELs are
available in a known sensitive human subpopulation, or if human
toxicokinetics or toxicodynamics are known with some certainty,
adjusting or replacing the default value of 10 has been recommended
(Dourson et al. 1996). No data related to occupational cohorts were
identified.

Researchers report that differences in individual response can be due to
many factors, but can generally be broken into two categories:
variability in toxicokinetics (absorption, distribution, metabolism, and
excretion) and differences in toxicodynamics (response at the target
site) (Renwick and Lazarus 1998; WHO 2001). This is true both for
differences between species and differences between individuals within a
species. There is growing support for more data-informed approaches to
account for human variability, which range from default (“presumed
protective”) to more “biological based protective” (based on
substance-specific empirical data) (EPA 2002; WHO 2001). Renwick
proposes dividing the UF for intrahuman variability into a factor of 4.0
for toxicokinetics and a factor of 2.5 for toxicodynamics (Renwick and
Lazarus 1998). WHO (2001) calls for a factor of 3.16 (100.5) for
toxicokinetics and 3.16 for toxicodynamics, and has used this approach
for several of its recent risk assessments. The use of such
chemical-specific adjustment factors (CSAFs) is based principally on the
availability of data (e.g., toxicokinetic and toxicodynamic data,
biological models). 

EPA has not yet established guidance for the use of chemical-specific
data for deriving UFs, but division of UFs into toxicokinetic and
toxicodynamic components is in its RfC methodology (EPA 1994 and 2002).
A growing number of assessments, however, have reduced uncertainty
factors for intraspecies variability by half a log (i.e., 3). For
example, Moore et al. (1995, 1997) did so in their evaluations of the
more data-rich chemicals lithium and boric acid. NIOSH indicated that it
would consider whether some portion of the 10-fold factor could be
equated to toxicokinetic differences and replaced by the use of a
validated physiologically-based toxicokinetic model, but notes that
replacing the toxicodynamic portion of the UF would require an excellent
understanding of the mechanism and good human data which NIOSH
recognizes is not often available (Dankovic 2004).

3.4	Selecting an Appropriate UF

In the case of nPB, insufficient data are available to develop CSAFs. No
human kinetic or metabolism data were identified (NTP 2003). To some
extent, however, the considerations described above inform the nPB
decision at least from a qualitative perspective. Regarding the
toxicokinetics of nPB, the presumption of the healthy worker effect
points to some reduction of uncertainty across the worker population
compared to the general population—implying that a UF less than 10 is
appropriate for differences within the working population. Because
mechanism/mode of action data are inadequate and some workers may have
altered sperm parameters to begin with (see Section 3.2), accounting for
uncertainty/variability in the worker population by applying an UF of
greater than 1 is appropriately protective. EPA’s default value
(100.5) or, roughly 3, seems reasonable because it accounts for the
occupational nature of the exposure (shorter exposure) and the absence
of children and elderly in the workplace (potentially more sensitive
receptors), but factors in the possible variability in fertility
parameters across the healthy worker population. References Cited

ACGIH (American Conference of Governmental Industrial Hygienists). 2003.
TLVs( and BEIs( Based on the Documentation of the Threshold Limit Values
for Chemical Substances and Physical Agents & Biological Exposure
Indices.

ACOEM (American College of Occupational and Environmental Medicine)
1996. ACOEM reproductive hazard management guidelines [committee
report]. J Occup Environ Med. 38(1):83-90. 

Ahlers HW. Acting Branch Chief. Document Development Branch. Education
and Information Division. National Institute of Occupational Safety and
Health (NIOSH). Personal communication. April 13, 2004.

AIHA. 2003. WEEL Development Flowchart. Workplace Environmental Exposure
Levels Committee, American Industrial Hygiene Association. Available at:
  HYPERLINK
"http://www.aiha.org/Committees/documents/weeldevelopmentchart.pdf" 
http://www.aiha.org/Committees/documents/weeldevelopmentchart.pdf 

AIHA. 2002. White Paper on Risk Assessment and Risk Management. American
Industrial Hygiene Association. Adopted by Board of Directors: December
7, 1994. Amended and Adopted by Board:  February 1, 2002 Available at:  
HYPERLINK "http://www.aiha.org/GovernmentAffairs-PR/html/GAWPrisk.htm" 
http://www.aiha.org/GovernmentAffairs-PR/html/GAWPrisk.htm 

Alvarez et al. 2003. Biological variation of seminal parameters in
health subjects. Hum Reprod. Oct;18(10):2082-8.

Auger et al. 2000. Intra- and inter-individual variability in human
sperm concentration, motility and vitality assessment during a workshop
involving ten laboratories. Hum. Reprod. 15(11):2360-8.

Bonde JP, Storgaard L. 2002. How work-place conditions, environmental
toxicants and lifestyle affect male reproductive function. Int J Androl.
25(5):262-8.

Bonde JP, Kold Jensen T, Brixen Larsen S, Abell A, Scheike T, Hjollund
NH, Kolstad HA, Ernst E, Giwercman A, Skakkebaek NE, Keiding N, Olsen J.
1998. Year of birth and sperm count in 10 Danish occupational studies.
Scand J Work Environ Health. 24(5):407-13.  

Burin GJ and Saunders DR. 1999. Addressing human variability in risk
assessment -- the robustness of the intraspecies uncertainty factor.
Regul Toxicol Pharmacol. Dec;30(3):209-16.

CA EPA (California Environmental Protection Agency). 2003. The Air
Toxics Hot Spots Program Guidance Manual for Preparation of Health Risk
Assessments. Office of Environmental Health Hazard Assessment. August
2003. Available at:   HYPERLINK
"http://www.oehha.ca.gov/air/hot_spots/pdf/HRAguidefinal.pdf" 
http://www.oehha.ca.gov/air/hot_spots/pdf/HRAguidefinal.pdf 

Dankovic D. Acting Chief, Risk Evaluation Branch, National Institute of
Occupational Safety and Health (NIOSH). Personal communication. March
30, 2004.

Dourson ML, Maier A, Meek B, Renwick A, Ohanian E, and Poirer K. 1998.
Boron tolerable intake: re-evaluation of toxicokinetics for data-derived
uncertainty factors. Biol. Trace Element Res. 66: 453-463. Available at:

  HYPERLINK "http://www.tera.org/pubs/Boron1998.pdf" 
http://www.tera.org/pubs/Boron1998.pdf  

Dourson ML, Felter SP, and Robinson D. 1996. Evolution of science-based
uncertainty factors in noncancer risk assessment. Reg. Tox. Pharmacol.,
24: 108-120. Available at:   HYPERLINK
"http://www.tera.org/pubs/UF%20in%20Noncancer.pdf" 
http://www.tera.org/pubs/UF%20in%20Noncancer.pdf  

EPA. 2002. A review of the reference dose and reference concentration
processes. Final Report. Risk Assessment Forum. Reference/Dose/Reference
Concentration (RfD/RfC) Technical Panel. EPA/630/P-02/002F. December
2002. Available at:   HYPERLINK "http://www.epa.gov/iris/backgr-d.htm" 
http://www.epa.gov/iris/backgr-d.htm 

EPA. 1996. Guidelines for reproductive toxicity risk assessment. Risk
Assessment Forum. Washington. DC. 630/R-96/009. Available at:  
HYPERLINK "http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2838" 
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2838 

EPA 1994. Methods for derivation inhalation reference concentrations and
application of inhalation dosimetry. EPA/600/8-90/066F. October 1994.
Available at:   HYPERLINK "http://www.epa.gov/iris/backgr-d.htm" 
http://www.epa.gov/iris/backgr-d.htm 

Fedoruk MJ 1996. ACOEM reproductive hazard management guidelines.
American College of Occupational and Environmental Medicine. J Occup
Environ Med Jan;38(1):83-90.  

Fisch H and Goluboff ET. 1996. Geographic variations in sperm counts: a
potential cause of bias in studies of semen quality. Fertil Steril.
65(5):1044-6.

Fisher JS, Macpherson S, Marchetti N, Sharpe RM. 2003. Human
‘testicular dysgenesis syndrome’: a possible model using in-utero
exposure of the rat to dibutyl phthalate. Hum Reprod. Jul;18(7):1383-94.

Griffin JE, Wilson JD. 1994. Disorders of the testes. In: Isselbacher
KJ, Braunwald E, Wilson JD, Martin JB, Fauci AS, Kasper DL (eds),
Harrison’s Principles of Internal Medicine, 13th Edition, McGraw-Hill,
New York, Chapter 339, pp 2006-2017.

Guo YL, Hsu PC, Hsu CC, Lambert GH. 2000. Semen quality after prenatal
exposure to polychlorinated biphenyls and dibenzofurans. Lancet.
Oct;356(9237):1240-1.

Higuchi TT, Palmer JS, Gray LE Jr, Veeramachaneni DN. 2003. Effects of
dibutyl phthalate in male rabbits following in utero, adolescent, or
postpubertal exposure. Toxicol Sci. Apr;72(2):301-13.

Hogan KA. 2000. Characterization of data variability and uncertainty:
health effects assessments in the Integrated Risk Information System
(IRIS). U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment. Washington,
DC. EPA/635/R-00/005A.

Ichihara G, Yu X, Kitoh J, Asaeda N, Kumazawa T, Iwai H et al. 2000.
Reproductive toxicity of 1-bromopropane, a newly introduced alternative
to ozone layer depleting solvents, in male rats. Toxicol Sci.
54(2):416-23. Erratum in: Toxicol Sci. 55(2).

Jensen TK, Jorgensen N, Punab M, Haugen TB, Suominen J, et al. 2004.
Association of in utero exposure to maternal smoking with reduced semen
quality and testis size in adulthood: a cross-sectional study of 1,770
young men from the general population in five European countries. Am J
Epidemiol. Jan;159(1):49-58.

Kang KS, Li GX, Che JH, Lee YS. 2002. Impairment of male rat
reproductive function in F1 offspring from dams exposed to 2-bromoprpane
during gestation and lactation. Reprod Toxicol. 16(2):151-9.

Kumar S. 2004. Review. Occupation exposure associated with reproductive
dysfunction. J Occup Health 46:1-19.

Larsen SB, Abell A, and Bonde JP. 1998. Selection bias in occupational
sperm studies. Am J Epidemiol. 147(7):681-5.

Moline JM, Golden AL, Bar-Chama N, Smith E, Rauch ME et al. 2000.
Exposure to hazardous substances and male reproductive health: a
research framework. Environ Health Perspect. 108(9)9:803-13. 

Available at:   HYPERLINK
"http://ehp.niehs.nih.gov/members/2000/108p803-813moline/moline-full.htm
l" 
http://ehp.niehs.nih.gov/members/2000/108p803-813moline/moline-full.html


Monson RR. 1987. Observations on the healthy worker effect. Journal of
Occupational Medicine. 28:425-33.

Moore, JA. and IEHR Expert Scientific Committee. 1995. An assessment of
lithium using the IEHR evaluative process for assessing human
developmental and reproductive toxicity of agents. Reprod Toxicol.
9(2):175-210. 

Moore, JA and an Expert Scientific Committee. 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(1):123-60. 

NAS. 2001. Evaluating Chemical and Other Agent Exposures for
Reproductive and Developmental Toxicity. Subcommittee on Reproductive
and Developmental Toxicity, Committee on Toxicology, Board on
Environmental Studies and Toxicology, National Research Council.
Available at:   HYPERLINK
"http://books.nap.edu/books/0309073162/html/index.html" 
http://books.nap.edu/books/0309073162/html/index.html 

Nayernia K, Li M, Engel W. 2004. Methods in Molecular Biology. Chapter
8: Spermatogonial Stem Cells. 253:105-20. March 2004.

NIOSH. 1996. National Institute of Occupational Safety and Health.
National Occupational Research Agenda.  April 1996.

NIOSH. 1991. NIOSH Criteria Documents. Criteria for a Recommended
Standard: Occupational Exposure to Ethylene Glycol Monomethyl Ether,
Ethylene Glycol Monoethyl Ether, and Their Acetates. DHHS (NIOSH)
Publication No. 91-119. September 1991. Available at:

  HYPERLINK "http://www.cdc.gov/niosh/91-103.html" 
http://www.cdc.gov/niosh/91-103.html 

NTP. 2003. NTP-CERHR monograph on the potential human reproductive and
development effects of 1-bromopropane. National Toxicology Program. U.S.
Department of Health and Human Services. Center for the Evaluation of
Risk to Human Reproduction. NIH Publication No. 04-4479. October 2003. 

NTP. 2002. NTP-CERHR expert panel report on the reproductive and
developmental toxicity of 1-bromopropane. National Toxicology Program.
U.S. Department of Health and Human Services. Center for the Evaluation
of Risk to Human Reproduction. NTP-CERHR-1-BP-02. March 2002. 

Ohta H, Wakayama T, Nishimune Y. 2004. Commitment of fetal male germ
cells to spermatogonial stem cells during mouse embryonic development.
Biol Reprod. May;70(5):1286-91.

Purves K and Christiansene E. 1992. Male infertility: current concepts.
Ann Med. 24:258-72.

Price PS, Keenan RE, Swartout JC, Gillis CA et al. 1997. An approach for
modeling noncancer dose responses with and emphasis on uncertainty. Risk
Analysis; 17(4): 427-37.

Renwick AG and Lazarus NR. 1998. Human variability and noncancer risk
assessment: An analysis of the default uncertainty factor. Regul Toxicol
Pharmacol. 27(1 Part 1): 3-20. 

Swan SH, Elkin EP and Fenster L. 2000. The question of declining sperm
density revisited: an analysis of 101 studies published 1934-1996.
Environ Health Perspect. 108(10):961-6.

Swan et al. 2003 Geographic differences in semen quality of fertile U.S.
males. Environ Health Perspect. 111(4):414-20.

Traina ME, Rescia M, Urbani E, Mantovani A, Macri C, Ricciardi C et al.
2003. Long-lasting effects of lindane on mouse spermatogenesis induced
by in utero exposure. Reprod Toxicol. Jan-Feb;17(1):25-35.

WIL. 2001. WIL Research Laboratories. An inhalation two-generation
reproductive toxicity study of 1-bromopropane in rats. Sponsored by the
Brominated Solvent Consortium. May 24, 2001.

WHO 2001. Guidance document for the use of data in development of
chemical-specific adjustment factors (CSAFs) for interspecies difference
and human variability in dose/concentration - response assessment. World
Health Organization. International Programme on Chemical Safety.
WHO/PCS/01.4. July 2001. [This document reports the collective view of
an international group of experts and does not necessarily represent the
decisions or the stated policy of UNEP, ILO, or WHO]

Yamada T, Ichihara G, Wang H, Yu X et al. 2003. Exposure to
1-bromopropane causes ovarian dysfunction in rats. Toxicol. Sci.
71:96-103.

References Consulted 

ACGIH. 2002. Operations Manual.  Threshold Limit Values for Chemical
Substances Committee. American Conference of Governmental Industrial
Hygienists. Available at:   HYPERLINK
"http://www.acgih.org/TLV/OpsManual.pdf" 
http://www.acgih.org/TLV/OpsManual.pdf 

Barnes, D.G., and M.L. Dourson. 1988. Reference dose (RfD): Description
and use in health risk assessments. Regul. Toxicol. Pharmacol.
8:471-486. Available at:   HYPERLINK
"http://www.tera.org/pubs/RFD%20and%20use%20article.pdf" 
http://www.tera.org/pubs/RFD%20and%20use%20article.pdf 

Compilation of papers on the “healthy worker” effect:

  HYPERLINK "http://www.canoshweb.org/odp/html/jul1988.htm#H3" 
http://www.canoshweb.org/odp/html/jul1988.htm#H3 

Felter SP and Dourson M. 1998. The inexact science of risk assessment
(and implications for risk management). Human and Ecol. Risk Assessment:
Vol. 4(2):245-51. Available at:   HYPERLINK
"http://www.tera.org/pubs/inexact%20science.pdf" 
http://www.tera.org/pubs/inexact%20science.pdf 

Gaylor DW and Kodell RL. 2002. A procedure for developing risk-based
reference doses. Regul Toxicol Pharmacol. 35(2 Pt 1):137-41. 

Gaylor DW and Kodell RL. 2000. Percentiles of the product of uncertainty
factors for establishing probabilistic reference doses. Risk Analysis.
20(2):245-50.

Griffin E and Wilson D. Disorders of the testes. In: Isselbacher K,
Baunwald E, Wilson J. Martin B et al. eds. Harrison’s Principles of
Internal Medicine. 13th ed. New York, McGraw Hill. pp. 2006-17.

Jankovic J and Drake F. A screening method for occupational reproductive
health risk. Am Ind Hyg Assoc J 1996 Jul;57(7):641-9. 

Zhao Q, Unrine J, and Dourson M. 1999. Replacing the default values of
10 with data-derived values: a comparison of two different data derived
uncertainty factors for boron. Human and Ecological Risk Assessment.
5(5):973-83. Available at:   HYPERLINK
"http://www.tera.org/pubs/boron.pdf"  http://www.tera.org/pubs/boron.pdf


Potentially relevant references not reviewed

Guzick DS, Overstreet JW, Factor-Litvak P, Brazil CK et al. 2001. Sperm
morphology, motility, and concentration in fertile and infertile men. N.
Engl. J. Med 345:1388-93.

Menkveld R, Wong WY. Lombard CJ, Wetzels AMM et al. 2001. Semen
parameters, including WHO and strict criteria morphology, in a fertile
and subfertile population: an effort towards standardization of in-vivo
thresholds. Hum. Reprod. 16:1165-71.

 The continuous production of mammalian sperm at puberty is maintained
by the proliferation and differentiation of spermatogonial stem cells
(SSCs) that originate from primordial germ cells (PGCs) in the early
embryo (Ohta et al. 2004; Nayernia et al. 2004; Traina et al. 2003;
Moline et al. 2000). SSCs are responsible for initiating the first round
of spermatogenesis that will produce sperm upon the onset of puberty. 

 NIOSH noted that safety factors used in the past (e.g., 10 for both
inter- and intra-species differences) “often lead to unreasonably low
occupational exposure limits and the process is frowned on by the
courts” (Ahlers 2004). OSHA evaluations have followed similar trends. 

 Susceptible in this context would mean a greater response at the same
internal dose in a particular subpopulation due to intrinsic factors. 

 The highest proportion of reductions of the traditional 10-fold factor
has occurred with the interspecies factor for EPA’s RfCs (i.e.,
dosimetric adjustments as was done in the case of nPB).  

 PAGE   

 PAGE   8 

 PAGE   

 PAGE   18 

