1.
INTRODUCTION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.1.
PURPOSE
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.2.
INTENDED
AUDIENCE
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.3.
BACKGROUND
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.3.1.
Selection
of
Studies
for
the
Handbook
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1.3.2.
Using
the
Handbook
in
an
Exposure
Assessment
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
1.3.3.
Approach
Used
to
Develop
Recommendations
for
Exposure
Factors
.
.
.
4
1.3.4.
Characterizing
Variability
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
1.4.
GENERAL
EQUATION
FOR
CALCULATING
DOSE
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
11
1.5.
RESEARCH
NEEDS
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14
1.6.
ORGANIZATION
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
1.7.
REFERENCES
FOR
CHAPTER
1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
16
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
1
Purpose
C
Summarize
data
on
human
behaviors
and
characteristics
affecting
exposure
C
Recommend
exposure
factor
values
1.
INTRODUCTION
1.1.
PURPOSE
The
purpose
of
the
Exposure
Factors
Handbook
is
°
update
of
fish
intake
data;
to:
(
1)
summarize
data
on
human
behaviors
and
°
expansion
of
data
for
time
spent
at
residence;
characteristics
which
affect
exposure
to
environmental
°
update
of
body
weight
data;
contaminants,
and
(
2)
recommend
values
to
use
for
these
°
addition
of
body
weight
data
for
infants;
factors.
These
recommendations
are
not
legally
binding
on
°
update
of
population
mobility
data;
any
EPA
program
and
should
be
interpreted
as
suggestions
°
addition
of
new
data
for
average
time
spent
in
which
program
offices
or
individual
exposure
assessors
can
different
locations
and
various
microenvironconsider
and
modify
as
needed.
Most
of
these
factors
are
ments;
best
quantified
on
a
site
or
situation­
specific
basis.
The
°
addition
of
data
for
occupational
mobility;
handbook
has
strived
to
include
full
discussions
of
the
°
addition
of
breast
milk
ingestion;
issues
which
assessors
should
consider
in
deciding
how
to
°
addition
of
consumer
product
use;
and
use
these
data
and
recommendations.
The
handbook
is
°
addition
of
reference
residence
factors.
intended
to
serve
as
a
support
document
to
EPA's
Guidelines
for
Exposure
Assessment
(
U.
S.
EPA,
1992a).
Variation
Among
Studies
The
Guidelines
were
developed
to
promote
consistency
This
handbook
is
a
compilation
of
available
data
among
the
various
exposure
assessment
activities
that
are
from
a
variety
of
different
sources.
With
very
few
carried
out
by
the
various
EPA
program
offices.
This
exceptions,
the
data
presented
are
the
analyses
of
the
handbook
assists
in
this
goal
by
providing
a
consistent
set
individual
study
authors.
Since
the
studies
included
in
this
of
exposure
factors
to
calculate
dose.
handbook
varied
in
terms
of
their
objectives,
design,
scope,

1.2.
INTENDED
AUDIENCE
The
Exposure
Factors
to
study
and
from
factor
to
factor.
Handbook
is
addressed
to
For
example,
some
authors
used
exposure
assessors
inside
the
geometric
means
to
present
their
Agency
as
well
as
outside,
who
results,
while
others
used
need
to
obtain
data
on
standard
arithmetic
means
or
distributions.
factors
needed
to
calculate
human
Authors
have
sometimes
used
exposure
to
toxic
chemicals.
different
terms
to
describe
the
1.3.
BACKGROUND
This
handbook
is
the
original
material
as
accurately
as
update
of
an
earlier
version
prepared
in
1989.
Revisions
possible,
EPA
has
made
an
effort
to
present
discussions
and
have
been
made
in
the
following
areas:
results
in
a
consistent
manner.
Further,
the
strengths
and
°
addition
of
drinking
water
rates
for
children;
with
a
better
understanding
of
the
uncertainties
associated
°
changes
in
soil
ingestion
rates
for
children;
with
the
values
derived
from
the
study.
°
addition
of
soil
ingestion
rates
for
adults;
°
addition
of
tapwater
consumption
for
adults
and
children;
Information
in
this
handbook
has
been
summarized
°
addition
of
mean
daily
intake
of
food
class
and
from
studies
documented
in
the
scientific
literature
and
subclass
by
region,
age
and
per
capita
rates;
°
addition
of
mean
moisture
content
of
selected
fruits,
vegetables,
grains,
fish,
meat,
and
dairy
products;
°
addition
of
food
intake
by
class
in
dry
weight
per
kg
of
body
weight
per
day;
°
update
of
homegrown
food
intake;
°
expansion
of
data
in
the
dermal
chapter;

presentation
of
results,
etc.,
the
level
of
detail,
statistics,
and
terminology
may
vary
from
study
same
racial
populations.
Within
the
constraint
of
presenting
the
limitations
of
each
study
are
discussed
to
provide
the
reader
1.3.1.
Selection
of
Studies
for
the
Handbook
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
2
August
1997
other
available
sources.
Studies
were
chosen
that
were
seen
°
Current
information:
Studies
were
chosen
only
as
useful
and
appropriate
for
estimating
exposure
factors.
if
they
were
sufficiently
recent
to
represent
The
handbook
contains
summaries
of
selected
studies
current
exposure
conditions.
This
is
an
published
through
August
30,
1997.
important
consideration
for
those
factors
that
General
Considerations
Many
scientific
studies
were
reviewed
for
possible
°
Adequacy
of
data
collection
period:
Because
inclusion
in
this
handbook.
Studies
were
selected
based
on
most
users
of
the
handbook
are
primarily
the
following
considerations:
addressing
chronic
exposures,
studies
were
°
Level
of
peer
review:
Studies
were
selected
techniques
for
collecting
data
to
characterize
predominantly
from
the
peer­
reviewed
literature
long­
term
behavior.
and
final
government
reports.
Internal
or
interim
reports
were
therefore
avoided.
°
Validity
of
approach:
Studies
utilizing
°
Accessibility:
Studies
were
preferred
that
the
more
likely
or
closely
capture
the
desired
user
could
access
in
their
entirety
if
needed.
measurement
were
selected.
In
general,
direct
°
Reproducibility:
Studies
were
sought
that
direct
observation,
personal
monitoring
devices,
contained
sufficient
information
so
that
methods
or
other
known
methods
were
preferred
where
could
be
reproduced,
or
at
least
so
the
details
of
available.
If
studies
utilizing
direct
the
author's
work
could
be
accessed
and
measurement
were
not
available,
studies
were
evaluated.
selected
that
rely
on
validated
indirect
°
Focus
on
exposure
factor
of
interest:
Studies
measures
(
such
as
heart
rate
for
inhalation
rate),
were
chosen
that
directly
addressed
the
and
use
of
questionnaires.
If
questionnaires
or
exposure
factor
of
interest,
or
addressed
related
surveys
were
used,
proper
design
and
factors
that
have
significance
for
the
factor
procedures
include
an
adequate
sample
size
for
under
consideration.
As
an
example
of
the
the
population
under
consideration,
a
response
latter
case,
a
selected
study
contained
useful
rate
large
enough
to
avoid
biases,
and
ancillary
information
concerning
fat
content
in
avoidance
of
bias
in
the
design
of
the
instrument
fish,
although
it
did
not
directly
address
fish
and
interpretation
of
the
results.
consumption.

°
Data
pertinent
to
the
U.
S.:
Studies
were
seeking
to
characterize
the
national
population,
selected
that
addressed
the
U.
S.
population.
a
particular
region,
or
sub­
population
were
Data
from
populations
outside
the
U.
S.
were
selected,
if
appropriately
representative
of
that
sometimes
included
if
behavioral
patterns
and
population.
In
cases
where
data
were
limited,
other
characteristics
of
exposure
were
similar.
studies
with
limitations
in
this
area
were
°
Primary
data:
Studies
were
deemed
preferable
handbook.
if
based
on
primary
data,
but
studies
based
on
secondary
sources
were
also
included
where
°
Variability
in
the
population:
Studies
were
they
offered
an
original
analysis.
For
example,
sought
that
characterized
any
variability
within
the
handbook
cites
studies
of
food
consumption
populations.
based
on
original
data
collected
by
the
USDA
National
Food
Consumption
Survey.
°
Minimal
(
or
defined)
bias
in
study
design:
change
with
time.

sought
that
utilized
the
most
appropriate
experimental
procedures
or
approaches
that
exposure
data
collection
techniques,
such
as
measurement
methods
such
as
surrogate
°
Representativeness
of
the
population:
Studies
included
and
limitations
were
noted
in
the
Studies
were
sought
that
were
designed
with
minimal
bias,
or
at
least
if
biases
were
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
3
Key
vs.
Relevant
Studies
C
Key
studies
used
to
derive
recommendations
C
Relevant
studies
included
to
provide
suspected
to
be
present,
the
direction
of
the
bias
°
Guidelines
for
Exposure
Assessment
(
U.
S.
EPA
(
i.
e.,
an
over
or
under
estimate
of
the
1992a);
parameter)
was
either
stated
or
apparent
from
°
Dermal
Exposure
Assessment:
Principles
and
the
study
design.
Applications
(
U.
S.
EPA
1992b);

°
Minimal
(
or
defined)
uncertainty
in
the
data:
Associated
with
Indirect
Exposure
to
Studies
were
sought
with
minimal
uncertainty
in
Combustor
Emissions
(
U.
S.
EPA,
1990);
the
data,
which
was
judged
by
evaluating
all
the
°
Risk
Assessment
Guidance
for
Superfund
(
U.
S.
considerations
listed
above.
At
least,
studies
EPA,
1989);
were
preferred
that
identified
uncertainties,
°
Estimating
Exposures
to
Dioxin­
Like
such
as
those
due
to
inherent
variability
in
Compounds
(
U.
S.
EPA,
1994);
environmental
and
exposure­
related
parameters
°
Superfund
Exposure
Assessment
Manual
(
U.
S.
or
possible
measurement
error.
Studies
that
EPA,
1988a);
documented
Quality
Assurance/
Quality
Control
°
Selection
Criteria
for
Mathematical
Models
measures
were
preferable.
Used
in
Exposure
Assessments
(
U.
S.
EPA
Key
versus
relevant
studies
°
Selection
Criteria
for
Certain
studies
described
Mathematical
Models
in
this
handbook
are
designated
Used
in
Exposure
as
"
key,"
that
is,
the
most
useful
Assessments
(
U.
S.
EPA
for
deriving
exposure
factors.
The
1987);
recommended
values
for
most
°
Standard
Scenarios
for
exposure
factors
are
based
on
the
Estimating
Exposure
to
results
of
the
key
studies.
Other
Chemical
Substances
studies
are
designated
"
relevant,"
During
Use
of
Consumer
meaning
applicable
or
pertinent,
Products
(
U.
S.
EPA
but
not
necessarily
the
most
1986a);
important.
This
distinction
was
made
on
the
strength
of
the
°
Pesticide
Assessment
Guidelines,
Subdivisions
K
attributes
listed
in
the
"
General
Considerations."
For
and
U
(
U.
S.
EPA,
1984,
1986b);
and
example,
in
Chapter
14
of
Volume
III,
one
set
of
studies
is
°
Methods
for
Assessing
Exposure
to
Chemical
deemed
to
best
address
the
attributes
listed
and
is
Substances,
Volumes
1­
13
(
U.
S.
EPA,
1983­
1989).
designated
as
"
key."
Other
applicable
studies,
including
foreign
data,
believed
to
have
value
to
handbook
users,
but
These
documents
may
serve
as
valuable
information
having
fewer
attributes,
are
designated
"
relevant."
resources
to
assist
in
the
assessment
of
exposure.
The
1.3.2.
Using
the
Handbook
in
an
Exposure
Assessment
Some
of
the
steps
for
performing
an
exposure
handbook
discusses
the
recommendations
provided
by
the
assessment
are
(
1)
determining
the
pathways
of
exposure,
American
Industrial
Health
Council
(
AIHC)
­
Exposure
(
2)
identifying
the
environmental
media
which
transports
Factors
Sourcebook
(
May
1994)
for
some
of
the
major
the
contaminant,
(
3)
determining
the
contaminant
exposure
factors.
The
AIHC
Sourcebook
summarizes
and
concentration,
(
4)
determining
the
exposure
time,
evaluates
statistical
data
for
various
exposure
factors
used
frequency,
and
duration,
and
(
5)
identifying
the
exposed
in
risk
assessments.
Probability
distributions
for
population.
Many
of
the
issues
related
to
characterizing
exposure
from
selected
exposure
pathways
have
been
addressed
in
a
number
of
existing
EPA
guidance
documents.
These
include,
but
are
not
limited
to
the
following:
°
Methodology
for
Assessing
Health
Risks
1988b);

reader
is
encouraged
to
refer
to
them
for
more
detailed
discussion.
In
addition
to
the
references
listed
above,
this
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
4
August
1997
Recommendations
and
Confidence
Ratings
C
Recommendations
based
on
data
from
single
or
multiple
key
studies
C
Variability
and
limitation
of
the
data
evaluated
C
Recommendations
rated
as
low,
medium,
and
specific
exposure
factors
were
derived
from
the
available
consistent
with
the
exposure
factors
for
a
population
of
scientific
literature
using
@
Risk
simulation
software.
Each
interest.
This
should
serve
as
a
guide
for
when
this
issue
is
factor
is
described
by
a
specific
term,
such
as
lognormal,
a
concern.
normal,
cumulative
type,
or
triangular.
Other
distributions
included
Weibull,
beta
logistic,
and
gamma.
Unlike
this
handbook,
however,
the
Sourcebook
does
not
provide
a
description
and
evaluation
of
every
study
available
on
each
As
discussed
above,
EPA
first
reviewed
all
literature
exposure
factor.
pertaining
to
a
factor
and
determined
relevant
and
key
Most
of
the
data
presented
in
this
handbook
are
studies.
The
key
studies
were
used
to
derive
derived
from
studies
that
targeted
(
1)
the
general
population
recommendations
for
the
values
of
each
factor.
The
(
e.
g.,
USDA
food
consumptin
surveys);
and
(
2)
a
sample
recommended
values
were
derived
solely
from
EPA's
population
from
a
specific
area
or
group
(
e.
g.,
Calabrese's
interpretation
of
the
available
data.
Different
values
may
be
et
al.
(
1989)
soil
ingestion
study
using
children
from
the
appropriate
for
the
user
to
select
in
consideration
of
policy,
Amherst,
Massachusetts,
area).
Due
to
unique
activity
precedent,
strategy,
or
other
factors
such
as
site­
specific
patterns,
preferences,
practices
and
biological
differences,
information.
EPA's
procedure
for
developing
various
segments
of
the
population
may
experience
recommendations
was
as
follows:
exposures
that
are
different
from
those
of
1.
Key
studies
were
the
general
population,
evaluated
in
terms
which,
in
many
cases,
of
both
quality
and
may
be
greater.
It
is
relevance
to
necessary
for
risk
or
specific
populaexposure
assessors
tions
(
general
U.
S.
characterizing
a
diverse
population,
age
population,
to
identify
groups,
gender,
and
enumerate
certain
etc.).
The
criteria
groups
within
the
for
assessing
the
general
population
who
quality
of
studies
is
are
at
risk
for
greater
described
in
Section
contaminant
exposures
1.3.1.
or
exhibit
a
heightened
sensitivity
to
particular
chemicals.
For
further
guidance
on
2.
If
only
one
study
was
classified
as
key
for
a
particular
addressing
susceptible
populations,
it
is
recommended
to
factor,
the
mean
value
from
that
study
was
selected
as
consult
the
EPA,
National
Center
for
Environmental
the
recommended
central
value
for
that
population.
If
Assessment
document
Socio­
demographic
Data
Used
for
Identifying
Potentially
Highly
Exposed
Subpopulations
(
to
be
released
as
a
final
document
in
the
Fall
of
1997).
Most
users
of
the
handbook
will
be
preparing
estimates
of
exposure
which
are
to
be
combined
with
doseresponse
factors
to
estimate
risk.
Some
of
the
exposure
factors
(
e.
g.,
life
time,
body
weight)
presented
in
this
document
are
also
used
in
generating
dose­
response
relationships.
In
order
to
develop
risk
estimates
properly,
assessors
must
use
dose­
response
relationships
in
a
manner
consistent
with
exposure
conditions.
Although,
it
is
beyond
the
scope
of
this
document
to
explain
in
detail
how
assessors
should
address
this
issue,
a
discussion
(
see
Appendix
A
of
this
chapter)
has
been
included
which
describes
how
dose­
response
factors
can
be
modified
to
be
1.3.3.
Approach
Used
to
Develop
Recommendations
for
Exposure
Factors
there
were
multiple
key
studies,
all
with
reasonably
equal
quality,
relevance,
and
study
design
information
were
available,
a
weighted
mean
(
if
appropriate,
considering
sample
size
and
other
statistical
factors)
of
the
studies
were
chosen
as
the
recommended
mean
value.
If
the
key
studies
were
judged
to
be
unequal
in
quality,
relevance,
or
study
design,
the
range
of
means
were
presented
and
the
user
of
this
handbook
must
employ
judgment
in
selecting
the
most
appropriate
value
for
the
population
of
interest.
In
cases
where
the
national
population
was
of
interest,
the
mid­
point
of
the
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
5
range
was
usually
judged
to
be
the
most
appropriate
when
determining
usual
intake
of
foods
in
a
population.
On
value.
the
other
hand,
it
is
not
as
important
for
factors
where
long­

3.
The
variability
of
the
factor
across
the
population
was
the
case
of
tapwater
intake,
the
currency
of
the
data
was
a
discussed.
If
adequate
data
were
available,
the
critical
element
in
determining
the
final
rating.
In
addition,
variability
was
described
as
either
a
series
of
some
exposure
factors
are
more
easily
measured
than
percentiles
or
a
distribution.
others.
For
example,
soil
ingestion
by
children
is
estimated
4.
Limitations
of
the
data
were
discussed
in
terms
of
data
found
in
soil.
Body
weight,
however,
can
be
measured
limitations,
the
range
of
circumstances
over
which
the
directly
and
it
is,
therefore,
a
more
reliable
measurement.
estimates
were
(
or
were
not)
applicable,
possible
This
is
reflected
in
the
confidence
rating
given
to
both
of
biases
in
the
values
themselves,
a
statement
about
these
factors.
In
general,
the
better
the
methodology
used
parameter
uncertainties
(
measurement
error,
sampling
to
measure
the
exposure
factor,
the
higher
the
confidence
in
error)
and
model
or
scenario
uncertainties
if
models
or
the
value.
scenarios
have
been
used
in
the
derivation
of
the
recommended
value.

5.
Finally,
EPA
assigned
a
confidence
rating
of
low,
each
of
the
factors.
Variability
is
characterized
in
one
or
medium
or
high
to
each
recommended
value.
This
more
of
three
ways:
(
1)
as
tables
with
various
percentiles
or
rating
is
not
intended
to
represent
an
uncertainty
ranges
of
values;
(
2)
as
analytical
distributions
with
analysis,
rather
it
represents
EPA's
judgment
on
the
specified
parameters;
and/
or
(
3)
as
a
qualitative
discussion.
quality
of
the
underlying
data
used
to
derive
the
Analyses
to
fit
standard
or
parametric
distributions
(
e.
g.,
recommendation.
This
judgment
was
made
using
the
normal,
lognormal)
to
the
exposure
data
have
not
been
guidelines
shown
in
Table
1­
1.
Table
1­
1
is
an
performed
by
the
authors
of
this
handbook,
but
have
been
adaptation
of
the
General
Considerations
discussed
reproduced
in
this
document
wherever
they
were
found
in
earlier
in
Section
1.3.1.
Clearly
this
is
a
continuum
the
literature.
Recommendations
on
the
use
of
these
from
low
to
high
and
judgment
was
used
to
determine
distributions
are
made
where
appropriate
based
on
the
these
ratings.
Recommendations
given
in
this
adequacy
of
the
supporting
data.
The
list
of
exposure
handbook
are
accompanied
by
a
discussion
of
the
factors
and
the
way
that
variability
has
been
characterized
rationale
for
their
rating.
(
i.
e.,
average,
upper
percentiles,
multiple
percentiles,
fitted
Table
1­
2
summarizes
EPA's
recommendations
and
percentile
is
used
throughout
this
handbook
and
it
is
confidence
ratings
for
the
various
exposure
factors.
intended
to
represent
values
in
the
upper
tail
(
i.
e.,
between
It
is
important
to
note
that
the
study
elements
listed
90th
and
99.9th
percentile)
of
the
distribution
of
values
for
in
Table
1­
1
do
not
have
the
same
weight
when
arriving
at
a
particular
exposure
factor.
the
overall
confidence
rating
for
the
various
exposure
An
attempt
was
made
to
present
percentile
values
in
factors.
The
relative
weight
of
each
of
these
elements
the
recommendations
that
are
consistent
with
the
exposure
depend
on
the
exposure
factor
of
interest.
Also,
the
relative
estimators
defined
in
the
Exposure
Guidelines
(
i.
e.,
mean,
weights
given
to
the
elements
for
the
various
factors
were
50th,
90th,
95th,
98th,
and
99.9th
percentile).
This
was
not,
subjective
and
based
on
the
professional
judgement
of
the
however,
always
possible
because
either
the
data
available
authors
of
this
handbook.
In
general,
most
studies
would
were
limited
for
some
factors,
or
the
authors
of
the
study
did
rank
high
with
regard
to
"
level
of
peer
review,"
not
provide
such
information.
It
is
important
to
note,
"
accessibility,"
"
focus
on
the
factor
of
interest,"
and
"
data
however,
that
these
percentiles
were
discussed
in
the
pertinent
to
the
U.
S."
These
elements
are
important
for
the
Exposure
Guidelines
within
the
context
of
risk
descriptors
study
to
be
included
in
this
handbook.
However,
a
high
and
not
individual
exopusure
factors.
For
example,
the
score
of
these
elements
does
not
necessarily
translate
into
a
Guidelines
stated
high
overall
score.
Other
elements
in
Table
1­
1
were
also
examined
to
determine
the
overall
score.
For
example,
the
adequacy
of
data
collection
period
may
be
more
important
term
variability
may
be
small
such
as
tapwater
intake.
In
by
measuring,
in
the
feces,
the
levels
of
certain
elements
1.3.4.
Characterizing
Variability
This
document
attempts
to
characterize
variability
of
distribution)
are
presented
in
Table
1­
3.
The
term
upper
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
6
August
1997
Table
1­
1.
Considerations
Used
to
Rate
Confidence
in
Recommended
Values
CONSIDERATIONS
HIGH
CONFIDENCE
LOW
CONFIDENCE
Study
Elements
Level
of
peer
review
The
studies
received
high
level
of
peer
review
The
studies
received
limited
peer
review.
(
e.
g.,
they
appear
in
peer
review
journals).

Accessibility
The
studies
are
widely
available
to
the
public.
The
studies
are
difficult
to
obtain
(
e.
g.,
draft
reports,
unpublished
data).

Reproducibility
The
results
can
be
reproduced
or
methodology
The
results
cannot
be
reproduced,
the
can
be
followed
and
evaluated.
methodology
is
hard
to
follow,
and
the
author(
s)
cannot
be
located.

Focus
on
factor
of
interest
The
studies
focused
on
the
exposure
factor
of
The
purpose
of
the
studies
was
to
characterize
a
interest.
related
factor.

Data
pertinent
to
U.
S.
The
studies
focused
on
the
U.
S.
population.
The
studies
focused
on
populations
outside
the
U.
S.

Primary
data
The
studies
analyzed
primary
data.
The
studies
are
based
on
secondary
sources.

Currency
The
data
were
published
after
1990.
The
data
were
published
before
1980.

Adequacy
of
data
collection
period
The
study
design
captures
the
measurement
of
The
study
design
does
not
very
accurately
interest
(
e.
g.,
usual
consumption
patterns
of
a
capture
the
measurement
of
interest.
population).

Validity
of
approach
The
studies
used
the
best
methodology
There
are
serious
limitations
with
the
approach
available
to
capture
the
measurement
of
used.
interest.

Study
sizes
The
sample
size
is
greater
than
100
samples.
The
sample
size
is
less
than
20
samples.

The
sample
size
depends
on
how
the
target
population
is
defined.
As
the
size
of
a
sample
relative
to
the
total
size
of
the
target
population
increases,
estimates
are
made
with
greater
statistical
assurance
that
the
sample
results
reflect
actual
characteristics
of
the
target
population.

Representativeness
of
the
population
The
study
population
is
the
same
as
population
The
study
population
is
very
different
from
the
of
interest.
population
of
interest.
a
Variability
in
the
population
The
studies
characterized
variability
in
the
The
characterization
of
variability
is
limited.
population
studied.

Lack
of
bias
in
study
design
Potential
bias
in
the
studies
are
stated
or
can
be
The
study
design
introduces
biases
in
the
results.
(
a
high
rating
is
desirable)
determined
from
the
study
design.

Response
rates
The
response
rate
is
less
than
40
percent.
In­
person
interviews
The
response
rate
is
greater
than
80
percent.
The
response
rate
is
less
than
40
percent.
Telephone
interviews
The
response
rate
is
greater
than
80
percent.
The
response
rate
is
less
than
40
percent.
Mail
surveys
The
respnose
rate
is
greater
than
70
percent.

Measurement
error
The
study
design
minimizes
measurement
Uncertainties
with
the
data
exist
due
to
errors.
measurement
error.

Other
Elements
Number
of
studies
The
number
of
studies
is
greater
than
3.
The
number
of
studies
is
1.

Agreement
between
researchers
The
results
of
studies
from
different
researchers
The
results
of
studies
from
different
researchers
are
in
agreement.
are
in
disagreement.

Differences
include
age,
sex,
race,
income,
or
other
demographic
parameters.
a
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
7
Table
1­
2.
Summary
of
Exposure
Factor
Recommendations
and
Confidence
Ratings
EXPOSURE
FACTOR
RECOMMENDATION
CONFIDENCE
RATING
Drinking
water
intake
rate
21
ml/
kg­
day/
1.4
L/
day
(
average)
Medium
34
ml/
kg­
day/
2.3
L/
day
(
90th
percentile)
Medium
Percentiles
and
distribution
also
included
Means
and
percentiles
also
included
for
pregnant
and
lactating
women
Total
fruit
intake
rate
3.4
g/
kg­
day
(
per
capita
average)
Medium
12.4
g/
kg­
day
(
per
capita
95th
percentile)
Low
Percentiles
also
included
Means
presented
for
individual
fruits
Total
vegetable
intake
rate
4.3
g/
kg­
day
(
per
capita
average)
Medium
10
g/
kg­
day
(
per
capita
95th
percentile)
Low
Percentiles
also
included
Means
presented
for
individual
vegetables
Total
meat
intake
rate
2.1
g/
kg­
day
(
per
capita
average)
Medium
5.1
g/
kg­
day
(
per
capita
95th
percentile)
Low
Percentiles
also
included
Percentiles
also
presented
for
individual
meats
Total
dairy
intake
rate
8.0
g/
kg­
day
(
per
capita
average)
Medium
29.7
g/
kg­
day
(
per
capita
95th
percentile)
Low
Percentiles
also
included
Means
presented
for
individual
dairy
products
Grain
intake
4.1
g/
kg­
day
(
per
capita
average)
High
10.8
g/
kg­
day
(
per
capita
95th
percentile)
Low
in
long­
term
upper
percentiles
Percentiles
also
included
Breast
milk
intake
rate
742
ml/
day
(
average)
Medium
1,033
ml/
day
(
upper
percentile)
Medium
Fish
intake
rate
General
Population
20.1
g/
day
(
total
fish)
average
High
14.1
g/
day
(
marine)
average
High
6.0
g/
day
(
freshwater/
estuarine)
average
High
63
g/
day
(
total
fish)
95th
percentile
long­
term
Medium
Percentiles
also
included
Serving
size
High
129
g
(
average)
High
326
g
(
95th
percentile)
Recreational
marine
anglers
Medium
2
­
7
g/
day
(
finfish
only)
Recreational
freshwater
Medium
8
g/
day
(
average)
Medium
25
g/
day
(
95th
percentile)
Native
American
Subsistence
Population
Medium
70
g/
day
(
average)
Low
170
g/
day
(
95th
percentile)
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
8
August
1997
Table
1­
2.
Summary
of
Exposure
Factor
Recommendations
and
Confidence
Ratings
(
continued)

EXPOSURE
FACTOR
RECOMMENDATION
CONFIDENCE
RATING
Home
produced
food
intake
Total
Fruits
Medium
(
for
means
and
short­
term
2.7
g/
kg­
day
(
consumer
only
average)
distributions)
11.1
g/
kg­
day
(
consumer
only
95th
percentile)
Low
(
for
long­
term
distributions)
Percentiles
also
included
Total
vegetables
2.1
g/
kg­
day
(
consumer
only
average)
7.5
g/
kg­
day
(
consumer
only
95th
percentile)
Percentiles
also
included
Total
meats
2.2
g/
kg­
day
(
consumer
only
average)
6.8
g/
kg­
day
(
consumer
only
95th
percentile)
Percentiles
also
included
Total
dairy
products
14
g/
kg­
day
(
consumer
only
average)
44
g/
kg­
day
(
consumer
only
95th
percentile)
Percentiles
also
included
Inhalation
rate
Children
(<
1
year)
High
4.5
m
/
day
(
average)
3
Children
(
1­
12
years)
High
8.7
m
/
day
(
average)
3
Adult
Females
High
11.3
m
/
day
(
average)
3
Adult
Males
High
15.2
m
/
day
(
average)
3
Surface
area
Water
contact
(
bathing
and
swimming)
High
Use
total
body
surface
area
for
children
in
Tables
6­
6
through
6­
8;
for
adults
use
Tables
6­
2
through
6­
4
(
percentiles
are
included)
Soil
contact
(
outdoor
activities)
High
Use
whole
body
part
area
based
on
Table
6­
6
through
6­
8
for
children
and
6­
2
through
6­
4
for
adults
(
percentiles
are
included)

Soil
adherence
Use
values
presented
in
Table
6­
16
depending
on
activity
Low
and
body
part
(
central
estimates
only)

Soil
ingestion
rate
Children
Medium
100
mg/
day
(
average)
400
mg/
day
(
upper
percentile)
Adults
Low
50
mg/
day
(
average)
Pica
child
Low
10
g/
day
Life
expectancy
75
years
High
Body
weight
for
adults
71.8
kg
High
Percentiles
also
presented
in
tables
7­
4
and
7­
5
Body
weights
for
children
Use
values
presented
in
Tables
7­
6
and
7­
7
(
mean
and
High
percentiles)

Body
weights
for
infants
(
birth
to
6
Use
values
presented
in
Table
7­
1
(
percentiles)
High
months)
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Table
1­
2.
Summary
of
Exposure
Factor
Recommendations
and
Confidence
Ratings
(
continued)

EXPOSURE
FACTOR
RECOMMENDATION
CONFIDENCE
RATING
Exposure
Factors
Handbook
Page
August
1997
1­
9
Showering/
Bathing
Showering
time
High
10
min/
day
(
average)
35
min/
day
(
95th
percentile)
(
percentiles
are
also
included)
Bathing
time
High
20
min/
event
(
median)
45
min/
event
(
90th
percentile)
Bathing/
showering
frequency
High
1
shower
event/
day
Swimming
Frequency
High
1
event/
month
Duration
High
60
min/
event
(
median)
180
min/
event
(
90th
percentile)

Time
indoors
Children
(
ages
3­
11)
Medium
19
hr/
day
(
weekdays)
17
hr/
day
(
weekends)
Adults
(
ages
12
and
older)
Medium
21
hr/
day
Residential
High
16.4
hrs/
day
Time
outdoors
Children
(
ages
3­
11)
Medium
5
hr/
day
(
weekdays)
7
hr/
day
(
weekends)
Adults
Medium
1.5
hr/
day
Residential
High
2
hrs/
day
Time
spent
inside
vehicle
Adults
1
hr
20
min/
day
Medium
Occupational
tenure
6.6
years
(
16
years
old
and
older)
High
Population
mobility
9
years
(
average)
Medium
30
years
(
95th
percentile)
Medium
Residence
volume
369
m
(
average)
Medium
3
217
m
(
conservative)
Medium
3
Residential
air
exchange
0.45
(
median)
Low
0.18
(
conservative)
Low
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
10
August
1997
Table
1­
3.
Characterization
of
Variability
in
Exposure
Factors
Exposure
Factors
Average
Upper
percentile
Multiple
Percentiles
Fitted
Distributions
Drinking
water
intake
rate
T
T
T
T
Total
fruits
and
total
vegetables
intake
rate
T
T
T
Qualitative
discussion
for
longterm
Individual
fruits
and
individual
vegetables
T
intake
rate
Total
meats
and
dairy
products
intake
rate
T
T
T
Qualitative
discussion
for
longterm
Individual
meats
and
dairy
products
intake
T
rate
Grains
intake
T
T
T
Breast
milk
intake
rate
T
T
Fish
intake
rate
for
general
population,
T
T
recreational
marine,
recreational
freshwater,
and
native
american
Serving
size
for
fish
T
T
T
Homeproduced
food
intake
rates
T
T
T
Soil
intake
rate
T
Qualitative
discussion
for
longterm
Inhalation
rate
T
T
Surface
area
T
T
T
Soil
adherence
T
Life
expectancy
T
Body
weight
T
T
T
Time
indoors
T
Time
outdoors
T
Showering
time
T
T
T
Occupational
tenure
T
Population
mobility
T
T
T
Residence
volume
T
Residential
air
exchange
T
that
the
assessor
may
derive
a
high­
end
estimate
of
C
The
exposure
assessor
should
only
consider
exposure
by
using
maximum
or
near
maximum
values
for
using
probabilistic
analysis
when
there
are
one
or
more
sensitive
exposure
factors,
leaving
others
at
credible
distribution
data
(
or
ranges)
for
the
their
mean
value.
factor
under
consideration.
Even
if
these
The
use
of
Monte
Carlo
or
other
probabilistic
distributions
are
known,
it
may
not
be
necessary
analysis
require
a
selection
of
distributions
or
histograms
to
apply
this
technique.
For
example,
if
only
for
the
input
parameters.
Although
this
handbook
is
not
average
exposure
values
are
needed,
these
can
intended
to
provide
a
complete
guidance
on
the
use
of
often
be
computed
accurately
by
using
average
Monte
Carlo
and
other
probabilistic
analyses,
the
following
values
for
each
of
the
input
parameters.
should
be
considered
when
using
such
techniques:
Probabilistic
analysis
is
also
not
necessary
when
conducting
assessments
for
screening
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
11
purposes,
i.
e.,
to
determine
if
unimportant
pathways
can
be
to
statistically
determine
the
distributions
that
fit
eliminated.
In
this
case,
bounding
estimates
can
be
the
data.
calculated
using
maximum
or
near
maximum
values
for
each
of
the
input
parameters.
C
If
only
a
range
of
values
is
known
for
an
C
It
is
important
to
note
that
the
selection
of
options.
distributions
can
be
highly
site
specific
and
will
always
involve
some
degree
of
judgment.
­
keep
that
variable
constant
at
its
central
Distributions
derived
from
national
data
may
value;
not
represent
local
conditions.
To
the
extent
­
assume
several
values
within
the
range
of
possible,
an
assessor
should
use
distributions
or
values
for
the
exposure
factor;
frequency
histograms
derived
from
local
­
calculate
a
point
estimate(
s)
instead
of
using
surveys
to
assess
risks
locally.
When
probabilistic
analysis;
and
distributional
data
are
drawn
from
national
or
­
assume
a
distribution
(
The
rationale
for
the
other
surrogate
population,
it
is
important
that
selection
of
a
distribution
should
be
discussed
the
assessor
address
the
extent
to
which
local
at
length.)
There
are,
however,
cases
where
conditions
may
differ
from
the
surrogate
data.
assuming
a
distribution
is
not
recommended.

In
addition
to
a
qualitative
statement
of
­­
data
are
missing
or
very
limited
for
a
key
uncertainty,
the
representativeness
assump­
tion
parameter
­
examples
include:
soil
should
be
appropriately
addressed
as
part
of
a
ingestion
by
adults;
sensitivity
analysis.
­­
data
were
collected
over
a
short
time
C
Distribution
functions
to
be
used
in
Monte
trends
(
the
respondent
usual
behavior)
­
Carlo
analysis
may
be
derived
by
fitting
an
examples
include:
food
consumption
appropriate
function
to
empirical
data.
In
doing
surveys;
activity
pattern
data;
this,
it
should
be
recognized
that
in
the
lower
­­
data
are
not
representative
of
the
and
upper
tails
of
the
distribution
the
data
are
population
of
interest
because
sample
size
scarce,
so
that
several
functions,
with
radically
was
small
or
the
population
studied
was
different
shapes
in
the
extreme
tails,
may
be
selected
from
a
local
area
and
was
consistent
with
the
data.
To
avoid
introducing
therefore
not
representative
of
the
area
of
errors
into
the
analysis
by
the
arbitrary
choice
of
interest
­
examples
include:
soil
ingestion
an
inappropriate
function,
several
techniques
by
children;
and
can
be
used.
One
way
is
to
avoid
the
problem
­­
ranges
for
a
key
variable
are
uncertain
by
using
the
empirical
data
itself
rather
than
an
due
to
experimental
error
or
other
analytic
function.
Another
is
to
do
separate
limitations
in
the
study
design
or
analyses
with
several
functions
which
have
methodology
­
examples
include:
soil
adequate
fit
but
form
upper
and
lower
bounds
to
ingestion
by
children.
the
empirical
data.
A
third
way
is
to
use
truncated
analytical
distributions.
Judgment
must
be
used
in
choosing
the
appropriate
goodness
of
fit
test.
Information
on
the
The
definition
of
exposure
as
used
in
the
Exposure
theoretical
basis
for
fitting
distributions
can
Guidelines
(
U.
S.
EPA,
1992a)
is
"
condition
of
a
be
found
in
a
standard
statistics
text
such
as
Statistical
Methods
for
Environmental
Pollution
Monitoring,
Gilbert,
R.
O.,
1987,
Van
Nostrand
Reinhold;
off­
the­
shelf
computer
software
such
as
Best­
Fit
by
Palisade
Corporation
can
be
used
exposure
factor,
the
assessor
has
several
These
include:

period
and
may
not
represent
long
term
1.4.
GENERAL
EQUATION
FOR
CALCULATING
DOSE
ADD
pot
'
Total
Potential
Dose
Body
Weight
x
Averaging
Time
Total
Potential
Dose
'
C
x
IR
x
ED
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
12
August
1997
(
Eqn.
1­
1)

(
Eqn.
1­
2)

Where:

C
=
Contaminant
Concentration
IR
=
Intake
Rate
ED
=
Exposure
Duration
chemical
contacting
the
outer
boundary
of
a
human."
This
Contaminant
concentration
is
the
concentration
of
means
contact
with
the
visible
exterior
of
a
person
such
as
the
contaminant
in
the
medium
(
air,
food,
soil,
etc.)
the
skin,
and
openings
such
as
the
mouth,
nostrils,
and
contacting
the
body
and
has
units
of
mass/
volume
or
lesions.
The
process
of
a
chemical
entering
the
body
can
be
mass/
mass.
described
in
two
steps:
contact
(
exposure),
followed
by
The
intake
rate
refers
to
the
rates
of
inhalation,
entry
(
crossing
the
boundary).
The
magnitude
of
exposure
ingestion,
and
dermal
contact
depending
on
the
route
of
(
dose)
is
the
amount
of
agent
available
at
human
exchange
exposure.
For
ingestion,
the
intake
rate
is
simply
the
boundaries
(
skin,
lungs,
gut)
where
absorption
takes
place
amount
of
food
containing
the
contaminant
of
interest
that
during
some
specified
time.
An
example
of
exposure
and
an
individual
ingests
during
some
specific
time
period
(
units
dose
for
the
oral
route
as
presented
in
the
the
EPA
of
mass/
time).
Much
of
this
handbook
is
devoted
to
rates
of
Exposure
Guidelines
is
shown
in
Figure
1­
1.
Starting
with
ingestion
for
some
broad
classes
of
food.
For
inhalation,
the
a
general
integral
equation
for
exposure
(
U.
S.
EPA
1992a),
intake
rate
is
the
rate
at
which
contaminated
air
is
inhaled.
several
dose
equations
can
be
derived
depending
upon
Factors
that
affect
dermal
exposure
are
the
amount
of
boundary
assumptions.
One
of
the
more
useful
of
these
material
that
comes
into
contact
with
the
skin,
and
the
rate
derived
equations
is
the
Average
Daily
Dose
(
ADD).
The
at
which
the
contaminant
is
absorbed.
ADD,
which
is
used
for
many
noncancer
effects,
averages
The
exposure
duration
is
the
length
of
time
that
exposures
or
doses
over
the
period
of
time
over
which
contaminant
contact
lasts.
The
time
a
person
lives
in
an
exposure
occurred.
The
ADD
can
be
calculated
by
area,
frequency
of
bathing,
time
spent
indoors
versus
averaging
the
potential
dose
(
D
)
over
body
weight
and
an
outdoors,
etc.
all
affect
the
exposure
duration.
The
Activity
pot
averaging
time.
Factors
Chapter
(
Volume
III,
Chapter
15)
gives
some
For
cancer
effects,
where
the
biological
response
is
at
the
concentration
and
intake
rate
specified
by
the
other
usually
described
in
terms
of
lifetime
probabilities,
even
parameters
in
the
equation.
though
exposure
does
not
occur
over
the
entire
lifetime,
Dose
can
be
expressed
as
a
total
amount
(
with
units
doses
are
often
presented
as
lifetime
average
daily
doses
of
mass,
e.
g.,
mg)
or
as
a
dose
rate
in
terms
of
mass/
time
(
LADDs).
The
LADD
takes
the
form
of
the
Equation
1­
1
(
e.
g.,
mg/
day),
or
as
a
rate
normalized
to
body
mass
(
e.
g.,
with
lifetime
replacing
averaging
time.
The
LADD
is
a
with
units
of
mg
of
chemical
per
kg
of
body
weight
per
day
very
common
term
used
in
carcinogen
risk
assessment
(
mg/
kg­
day)).
The
LADD
is
usually
expressed
in
terms
of
where
linear
non­
threshold
models
are
employed.
mg/
kg­
day
or
other
mass/
mass­
time
units.
The
total
exposure
can
be
expressed
as
follows:
In
most
cases
(
inhalation
and
ingestion
exposure)
the
examples
of
population
behavior
patterns,
which
may
be
useful
for
estimating
exposure
durations
to
be
used
in
the
exposure
calculations.
When
the
above
parameter
values
remain
constant
over
time,
they
are
substituted
directly
into
the
exposure
equation.
When
they
change
with
time,
a
summation
approach
is
needed
to
calculate
exposure.
In
either
case,
the
exposure
duration
is
the
length
of
time
exposure
occurs
dose­
response
parameters
for
carcinogen
risks
have
been
adjusted
for
the
difference
in
absorption
across
body
barriers
between
humans
and
the
experimental
animals
used
to
derive
such
parameters.
Therefore,
the
exposure
assessment
in
these
cases
is
based
on
the
potential
dose
with
no
explicit
correction
for
the
fraction
absorbed.
However,
the
exposure
assessor
needs
to
make
such
an
adjustment
when
calculating
dermal
exposure
and
in
other
specific
cases
when
current
information
indicates
that
the
human
absorption
factor
Organ
Chemical
Effect
Exposure
Internal
Dose
Biologically
Effective
Dose
Metabolism
Applied
Dose
Potential
Dose
G.
I.
Tract
Uptake
Mouth
Intake
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
13
Figure
1­
1.
Schematic
of
Dose
and
Exposure:
Oral
Route
Source:
U.
S.
EPA,
1992a
used
in
the
derivation
of
the
dose­
response
factor
is
included
in
the
"
intake
rate"
term
in
Equation
1­
2
and
the
inappropriate.
exposure
assessor
does
not
need
to
explicitly
include
body
The
lifetime
value
used
in
the
LADD
version
of
The
units
of
intake
in
this
handbook
for
the
ingestion
Equation
1­
1
is
the
period
of
time
over
which
the
dose
is
of
fish,
breast
milk,
and
the
inhalation
of
air
are
not
averaged.
For
carcinogens,
the
derivation
of
the
dose­
normalized
to
body
weight.
In
this
case,
the
exposure
response
parameters
usually
assumes
no
explicit
number
of
assessor
needs
to
use
(
in
Equation
1­
1)
the
average
weight
years
as
the
duration
of
a
lifetime,
and
the
nominal
value
of
of
the
exposed
population
during
the
time
when
the
75
years
is
considered
a
reasonable
approximation.
For
exposure
actually
occurs.
If
the
exposure
occurs
exposure
estimates
to
be
used
for
assessments
other
than
continuously
throughout
an
individual's
life
or
only
during
carcinogenic
risk,
various
averaging
periods
have
been
the
adult
ages,
using
an
adult
weight
of
71.8
kg
should
used.
For
acute
exposures,
the
administered
doses
are
provide
sufficient
accuracy.
If
the
body
weight
of
the
usually
averaged
over
a
day
or
a
single
event.
For
individuals
in
the
population
whose
risk
is
being
evaluated
nonchronic
noncancer
effects,
the
time
period
used
is
the
is
non­
standard
in
some
way,
such
as
for
children
or
for
actual
period
of
exposure.
The
objective
in
selecting
the
first­
generation
immigrants
who
may
be
smaller
than
the
exposure
averaging
time
is
to
express
the
exposure
in
a
way
national
population,
and
if
reasonable
values
are
not
which
can
be
combined
with
the
dose­
response
relationship
available
in
the
literature,
then
a
model
of
intake
as
a
to
calculate
risk.
function
of
body
weight
must
be
used.
One
such
model
is
The
body
weight
to
be
used
in
the
exposure
Equation
discussed
in
Appendix
1A
of
this
chapter.
Some
of
the
1­
1
depends
on
the
units
of
the
exposure
data
presented
in
parameters
(
primarily
concentrations)
used
in
estimating
this
handbook.
For
food
ingestion,
the
body
weights
of
the
exposure
are
exclusively
site
specific,
and
therefore
default
surveyed
populations
were
known
in
the
USDA
surveys
and
recommendations
could
not
be
used.
they
were
explicitly
factored
into
the
food
intake
data
in
The
food
ingestion
rate
values
provided
in
this
order
to
calculate
the
intake
as
grams
per
day
per
kilogram
handbook
are
generally
expressed
as
"
as
consumed"
since
body
weight.
In
this
case,
the
body
weight
has
already
been
this
is
the
fashion
in
which
data
are
reported
by
survey
weight.
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
14
August
1997
(
129
g/
meal)(
5
meals/
mo)(
mo/
30
d)(
365
d/
yr)(
30
yrs)
=
235,425
g
(
21.5
g/
day)(
365
d/
yr)(
30
yrs)
=
235,425
g
respondents.
This
is
of
importance
because
concentration
The
objective
is
to
define
the
terms
so
that
when
data
to
be
used
in
the
dose
equation
are
generally
measured
multiplied,
they
give
the
appropriate
estimate
of
mass
of
in
uncooked
food
samples.
In
most
situations,
the
only
contaminant
contacted.
This
can
be
accomplished
by
practical
choice
is
to
use
the
"
as
consumed"
ingestion
rate
basing
the
intake
rate
on
either
a
long­
term
average
(
chronic
and
the
uncooked
concentration.
However,
it
should
be
exposure)
or
an
event
(
acute
exposure)
basis,
as
long
as
the
recognized
that
cooking
generally
results
in
some
reductions
duration
value
is
selected
appropriately.
Consider
the
case
in
weight
(
e.
g.,
loss
of
moisture),
and
that
if
the
mass
of
the
in
which
a
person
eats
a
129­
g
fish
meal
approximately
five
contaminant
in
the
food
remains
constant,
then
the
times
per
month
(
long­
term
average
is
21.5
g/
day)
for
30
concentration
of
the
contaminant
in
the
cooked
food
item
years;
or
21.5
g/
day
of
fish
every
day
for
30
years.
will
increase.
Therefore,
if
the
"
as
consumed"
ingestion
rate
and
the
uncooked
concentration
are
used
in
the
dose
equation,
dose
may
be
underestimated.
On
the
other
hand,
cooking
may
cause
a
reduction
in
mass
of
contaminant
and
other
ingredients
such
that
the
overall
concentration
of
contaminant
does
not
change
significantly.
In
this
case,
combining
cooked
ingestion
rates
and
uncooked
Thus,
a
frequency
of
either
60
meals/
year
or
a
duration
of
concentration
will
provide
an
appropriate
estimate
of
dose.
365
days/
year
could
be
used
as
long
as
it
is
matched
with
Ideally,
food
concentration
data
should
be
adjusted
to
the
appropriate
intake
rate.
account
for
changes
after
cooking,
then
the
"
as
consumed"
intake
rates
are
appropriate.
In
the
absence
of
data,
it
is
reasonable
to
assume
that
no
change
in
contaminant
In
an
earlier
draft
of
this
handbook,
reviewers
were
concentration
occurs
after
cooking.
Except
for
general
asked
to
identify
factors
or
areas
where
further
research
is
population
fish
consumption
and
home
produced
foods,
needed.
The
following
list
is
a
compilation
of
areas
for
uncooked
intake
rate
data
were
not
available
for
presention
future
research
identified
by
the
peer
reviewers
and
authors
in
this
handbook.
Data
on
the
general
population
fish
of
this
document:
consumption
have
been
presented
in
this
handbook
(
Section
10.2)
in
both
"
as
consumed"
and
uncooked
basis.
It
is
°
The
data
and
information
available
with
respect
important
for
the
assessor
to
be
aware
of
these
issues
and
to
occupational
exposures
are
quite
limited.
choose
intake
rate
data
that
best
matches
the
concentration
Efforts
need
to
be
directed
to
identify
data
or
data
that
is
being
used.
references
on
occupational
exposure.
The
link
between
the
intake
rate
value
and
the
exposure
duration
value
is
a
common
source
of
confusion
°
Further
research
is
necessary
to
refine
estimates
in
defining
exposure
scenarios.
It
is
important
to
define
the
of
fish
consumption,
particularly
by
duration
estimate
so
that
it
is
consistent
with
the
intake
rate:
subpopulations
of
subsistence
fishermen.

°
The
intake
rate
can
be
based
on
an
individual
°
Research
is
needed
to
better
estimate
soil
intake
event,
such
as
129
g
of
fish
eaten
per
meal
rates,
particularly
how
to
extrapolate
short­
term
(
U.
S.
EPA,
1996).
The
duration
should
be
data
to
chronic
exposures.
Data
on
soil
intake
based
on
the
number
of
events
or,
in
this
case,
rates
by
adults
are
very
limited.
Research
in
meals.
this
area
is
also
recommended.
Research
is
also
°
The
intake
rate
also
can
be
based
on
a
long­
rate
(
i.
e.,
inconsistencies
among
tracers
and
term
average,
such
as
10
g/
day.
In
this
case
the
input/
output
misalignment
errors
indicate
a
duration
should
be
based
on
the
total
time
fundamental
problem
with
the
methods).
interval
over
which
the
exposure
occurs.
Research
is
also
needed
to
obtain
more
data
to
1.5.
RESEARCH
NEEDS
needed
to
refine
methods
to
calculate
soil
intake
better
estimate
soil
adherence.
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Exposure
Factors
Handbook
Page
August
1997
1­
15
°
In
cases
where
several
studies
of
equal
quality
evaluate
and
present
the
uncertainty
and
data
collection
procedures
are
available
for
associated
with
exposure
scenario
an
exposure
factor,
procedures
need
to
be
estimates.
developed
to
combine
the
data
in
order
to
create
a
single
distribution
of
likely
values
for
that
Chapter
3
Provides
factors
for
estimating
human
factor.
exposure
through
ingestion
of
water.

°
Reviewers
recommended
that
the
handbook
be
Chapter
4
Provides
factors
for
estimating
made
available
in
CD
ROM
and
that
the
data
exposure
through
ingestion
of
soil.
presented
be
made
available
in
a
format
that
will
allow
the
users
to
conduct
their
own
Chapter
5
Provides
factors
for
estimating
analysis.
The
intent
is
to
provide
a
exposure
as
a
result
of
inhalation
of
comprehensive
factors
tool
with
interactive
vapors
and
particulates.
menu
to
guide
users
to
areas
of
interest,
word
searching
features,
and
data
base
files.
Chapter
6
Presents
factors
for
estimating
dermal
°
Reviewers
recommended
that
EPA
derive
contaminants
that
come
in
contact
distribution
functions
using
the
empirical
data
with
the
skin.
for
the
various
exposure
factors
to
be
used
in
Monte
Carlo
or
other
probabilistic
analysis.
Chapter
7
Provides
data
on
body
weight.

°
Research
is
needed
to
derive
a
methodology
to
Chapter
8
Provides
data
on
life
expectancy.
extrapolate
from
short­
term
data
to
long­
term
or
chronic
exposures.

°
Reviewers
recommended
that
the
consumer
Chapter
9
Provides
factors
for
estimating
products
chapter
be
expanded
to
include
more
exposure
through
ingestion
of
fruits
products.
A
comprehensive
literature
search
and
vegetables.
needs
to
be
conducted
to
investigate
other
sources
of
data.
Chapter
10
Provides
factors
for
estimating
°
Breastmilk
intake.

°
More
recent
data
on
tapwater
intake.
exposure
through
ingestion
of
meats
°
SAB
recommended
analysis
of
1994
and
1995
CSFII
data.
Chapter
12
Presents
data
for
estimating
exposure
1.6.
ORGANIZATION
The
handbook
is
organized
into
three
volumes
as
Chapter
13
Presents
factors
for
estimating
follows:
exposure
through
ingestion
of
home
Volume
I
­
General
Factors
Chapter
1
Provides
the
overall
introduction
to
through
ingestion
of
breast
milk.
the
handbook
Chapter
2
Presents
an
analysis
of
uncertainty
and
discusses
methods
that
can
be
used
to
exposure
to
environmental
Volume
II
­
Ingestion
Factors
exposure
through
ingestion
of
fish.

Chapter
11
Provides
factors
for
estimating
and
dairy
products.

through
ingestion
of
grain
products.

produced
food.

Chapter
14
Presents
data
for
estimating
exposure
Volume
I
­
General
Factors
Chapter
1
­
Introduction
Page
Exposure
Factors
Handbook
1­
16
August
1997
Volume
III
­
Activity
Factors
Chapter
15
Presents
data
on
activity
factors
models.
Exposure
Assessment
Group,
Office
of
(
activity
patterns,
population
mobility,
Health
and
Environmental
Assessment,
Washington,
and
occupational
mobility).
DC.
WPA/
600/
8­
87/
042.
Available
from
NTIS,

Chapter
16
Presents
data
on
consumer
product
U.
S.
EPA.
(
1988a)
Superfund
exposure
assessment
use.
manual.
Office
of
Emergency
and
Remedial
Chapter
17
Presents
factors
used
in
estimating
Available
from
NTIS,
Springfield,
VA;
PB­
89­
residential
exposures.
135859.

Figure
1­
2
provides
a
roadmap
to
assist
users
of
this
models
used
in
exposure
assessments:
groundwater
handbook
in
locating
recommended
values
and
confidence
models.
Exposure
Assessment
Group,
Office
of
ratings
for
the
various
exposure
factors
presented
in
these
Health
and
Environmental
Assessment,
Washington,
chapters.
A
glossary
is
provided
at
the
end
of
Volume
III.
DC.
EPA/
600/
8­
88/
075.
Available
from
NTIS,

1.7.
REFERENCES
FOR
CHAPTER
1
AIHC.
(
1994)
Exposure
factors
sourcebook.
A.
Interim
Final.
Office
of
Solid
Waste
and
Washington,
DC:
American
Industrial
Health
Emergency
Response,
Washington,
DC.
Available
Council.
from
NTIS,
Springfield,
VA;
PB­
90­
155581.
Calabrese,
E.
J.;
Pastides,
H.;
Barnes,
R.;
Edwards,
C.;
U.
S.
EPA.
(
1990)
Methodology
for
assessing
health
risks
Kostecki,
P.
T.;
et
al.
(
1989)
How
much
soil
do
associated
with
indirect
exposure
to
combustor
young
children
ingest:
an
epidemiologic
study.
In:
emissions.
EPA
600/
6­
90/
003.
Available
from
Petroleum
Contaminated
Soils,
Lewis
Publishers,
NTIS,
Springfield,
VA;
PB­
90­
187055/
AS.
Chelsea,
MI.
pp.
363­
397.
U.
S.
EPA.
(
1992a)
Guidelines
for
exposure
assessment.
Gilbert,
R.
O.
(
1987)
Statistical
methods
for
Washington,
DC:
Office
of
Research
and
environmental
pollution
monitoring.
New
York:
Development,
Office
of
Health
and
Environmental
Van
Nostrand
Reinhold.
Assessment.
EPA/
600/
Z­
92/
001.
U.
S.
EPA.
(
1983­
1989)
Methods
for
assessing
exposure
U.
S.
EPA.
(
1992b)
Dermal
exposure
assessment:
to
chemical
substances.
Volumes
1­
13.
Washington,
principles
and
applications.
Washington,
DC:
Office
DC:
Office
of
Toxic
Substances,
Exposure
of
Health
and
Environmental
Assessments.
Evaluation
Division.
EPA/
600/
8­
9/
011F.
U.
S.
EPA.
(
1984)
Pesticide
assessment
guidelines
U.
S.
EPA.
(
1994)
Estimating
exposures
to
dioxin­
like
subdivision
K,
exposure:
reentry
protection.
Office
of
compounds.
(
Draft
Report).
Office
of
Research
and
Pesticide
Programs,
Washington,
DC.
EPA/
540/
9­
Development,
Washington,
DC.
EPA/
600/
6­
48/
001.
Available
from
NTIS,
Springfield,
VA;
PB­
88/
005Cb.
85­
120962.
U.
S.
EPA.
(
1996)
Daily
average
per
capita
fish
U.
S.
EPA.
(
1986a)
Standard
scenarios
for
estimating
consumption
estimates
based
on
the
combined
1989,
exposure
to
chemical
substances
during
use
of
1990,
and
1999
continuing
survey
of
food
intakes
by
consumer
products.
Volumes
I
and
II.
Washington,
individuals
(
CSFII)
1989­
91
data.
Volumes
I
and
II.
DC:
Office
of
Toxic
Substance,
Exposure
Evaluation
Preliminary
Draft
Report.
Washington,
DC:
Office
Division.
of
Water.
U.
S.
EPA.
(
1986b)
Pesticide
assessment
guidelines
subdivision
U,
applicator
exposure
monitoring.
Office
of
Pesticide
Programs,
Washington,
DC.
EPA/
540/
9­
87/
127.
Available
from
NTIS,
Springfield,
VA;
PB­
85­
133286.
U.
S.
EPA.
(
1987)
Selection
criteria
for
mathematical
models
used
in
exposure
assessments:
surface
water
Springfield,
VA;
PB­
88­
139928/
AS.

Response,
Washington,
DC.
EPA/
540/
1­
88/
001.

U.
S.
EPA.
(
1988b)
Selection
criteria
for
mathematical
Springfield,
VA;
PB­
88­
248752/
AS.
U.
S.
EPA.
(
1989)
Risk
assessment
guidance
for
Superfund.
Human
health
evaluation
manual:
part
Volume
I
­
General
Factors
Appendix
1A
Exposure
Factors
Handbook
Page
August
1997
1A­
1
APPENDIX
1A
RISK
CALCULATIONS
USING
EXPOSURE
FACTORS
HANDBOOK
DATA
AND
DOSE­
RESPONSE
INFORMATION
FROM
THE
INTEGRATED
RISK
INFORMATION
SYSTEM
(
IRIS)
Volume
I
­
General
Factors
Appendix
1A
Exposure
Factors
Handbook
Page
August
1997
1A­
3
APPENDIX
1A
RISK
CALCULATIONS
USING
EXPOSURE
FACTORS
HANDBOOK
DATA
AND
DOSE­
RESPONSE
INFORMATION
FROM
IRIS
1.
INTRODUCTION
When
calculating
risk
estimates
for
a
specific
population,
whether
the
entire
national
population
or
some
sub­
population,
the
exposure
information
(
either
from
this
handbook
or
from
other
data)
must
be
combined
with
dose­
response
information.
The
latter
typically
comes
from
the
IRIS
data
base,
which
summarizes
toxicity
data
for
each
agent
separately.
Care
must
be
taken
that
the
assumptions
about
population
parameters
in
the
dose­
response
analysis
are
consistent
with
the
population
parameters
used
in
the
exposure
analysis.
This
Appendix
discusses
procedures
for
insuring
this
consistency.

In
the
IRIS
derivation
of
threshold
based
dose­
response
relationships
(
U.
S.
EPA,
1996),
such
as
the
RfD
and
the
RfCs
based
on
adverse
systemic
effects,
there
has
generally
been
no
explicit
use
of
human
exposure
factors.
In
these
cases
the
numerical
value
of
the
RfD
and
RfC
comes
directly
from
animal
dosing
experiments
(
and
occasionally
from
human
studies)
and
from
the
application
of
uncertainty
factors
to
reflect
issues
such
as
the
duration
of
the
experiment,
the
fact
that
animals
are
being
used
to
represent
humans
and
the
quality
of
the
study.
However
in
developing
cancer
dose­
response
(
D­
R)
assessments,
a
standard
exposure
scenario
is
assumed
in
calculating
the
slope
factor
(
i.
e.,
human
cancer
risk
per
unit
dose)
on
the
basis
of
either
animal
bioassay
data
or
human
data.
This
standard
scenario
has
traditionally
been
assumed
to
be
typical
of
the
U.
S.
population:
1)
body
weight
=
70
kg;
2)
air
intake
rate
=
20
m
/
day;
3)
drinking
water
intake
=
2
liters/
day;
4)
3
lifetime
=
70
years.
In
RfC
derivations
for
cases
involving
an
adverse
effect
on
the
respiratory
tract,
the
air
intake
rate
of
20
m
/
day
is
assumed.
The
use
of
these
specific
values
has
depended
on
whether
the
slope
factor
was
derived
from
animal
3
or
human
epidemiologic
data:

C
Animal
Data:
For
dose­
resopnse
(
D­
R)
studies
based
on
animal
data,
scale
animal
doses
to
human
equivalent
doses
using
a
human
body
weight
assumption
of
70
kg.
No
explicit
lifetime
adjustment
is
necessary
because
the
assumption
is
made
that
events
occurring
in
the
lifetime
animal
bioassay
will
occur
with
equal
probability
in
a
human
lifetime,
whatever
that
might
happen
to
be.

C
Human
Data
­
In
the
analysis
of
human
studies
(
either
occupational
or
general
population),
the
Agency
has
usually
made
no
explicit
assumption
of
body
weight
or
human
lifetime.
For
both
of
these
parameters
there
is
an
implicit
assumption
that
the
population
usually
of
interest
has
the
same
descriptive
parameters
as
the
population
analyzed
by
the
Agency.
In
the
rare
situation
where
this
assumption
is
known
to
be
wrong,
the
Agency
has
made
appropriate
corrections
so
that
the
dose­
response
parameters
represent
the
national
average
population.

When
the
population
of
interest
is
different
than
the
national
average
(
standard)
population,
the
dose­
response
parameter
needs
to
be
adjusted.
In
addition,
when
the
population
of
interest
is
different
than
the
population
from
which
the
exposure
factors
in
this
handbook
were
derived,
the
exposure
factor
needs
to
be
adjusted.
Two
generic
examples
of
situations
where
these
adjustments
are
needed
are
as
follows:

A)
Detailed
study
of
recent
data,
such
as
are
presented
in
this
handbook,
show
that
EPA's
standard
assumptions
(
i.
e.,
70
kg
body
weight,
20
m
/
day
air
inhaled,
and
2
L/
day
water
intake)
are
inaccurate
for
the
national
population
and
may
be
3
inappropriate
for
sub­
populations
under
consideration.
The
handbook
addresses
most
of
these
situations
by
providing
gender­
and
age­
specific
values
and
by
normalizing
the
intake
values
to
body
weight
when
the
data
are
available,
but
it
may
not
have
covered
all
possible
situations.
An
example
of
a
sub­
population
with
a
different
mean
body
weight
would
be
females,
with
an
average
body
weight
of
60
kg
or
children
with
a
body
weight
dependent
on
age.
Another
example
of
a
nonstandard
sub­
population
would
be
a
sedentary
hospital
population
with
lower
than
20
m
/
day
air
intake
rates.
3
Volume
I
­
General
Factors
Appendix
1A
Page
Exposure
Factors
Handbook
1A­
4
August
1997
B)
The
population
variability
of
these
parameters
is
of
interest
and
it
is
desired
to
estimate
percentile
limits
of
the
population
variation.
Although
the
detailed
methods
for
estimating
percentile
limits
of
exposure
and
risk
in
a
population
are
beyond
the
scope
of
this
document,
one
would
treat
the
body
weight
and
the
intake
rates
discussed
in
Sections
2
to
4
of
this
appendix
as
distributions,
rather
than
constants.

2.
CORRECTIONS
FOR
DOSE­
RESPONSE
PARAMETERS
The
correction
factors
for
the
dose­
response
values
tabulated
in
the
IRIS
data
base
for
carcinogens
are
summarized
in
Table
1A­
1.
Use
of
these
correction
parameters
is
necessary
to
avoid
introducing
errors
into
the
risk
analysis.
The
second
column
of
Table
1A­
1
shows
the
dependencies
that
have
been
assumed
in
the
typical
situation
where
the
human
doseresponse
factors
have
been
derived
from
the
administered
dose
in
animal
studies.
This
table
is
applicable
in
most
cases
that
will
be
encountered,
but
it
is
not
applicable
when:
a)
the
effective
dose
has
been
derived
with
a
pharmacokinetic
model
and
b)
the
dose­
response
data
has
been
derived
from
human
data.
In
the
former
case,
the
subpopulation
parameters
need
to
be
incorporated
into
the
model.
In
the
latter
case,
the
correction
factor
for
the
dose­
response
parameter
must
be
evaluated
on
a
case­
by
case
basis
by
examining
the
specific
data
and
assumptions
in
the
derivation
of
the
parameter.

Table
1A­
1.
Procedures
for
Modifying
IRIS
Risk
Values
for
Non­
standard
Populationsa,
b
IRIS
Risk
Measure
IRIS
Risk
Measure
is
Proportional
to:
Correction
Factor
(
CF)
for
modifying
[
Units]
IRIS
Risk
Measures:
b
c
Slope
Factor
(
W
)
=
(
70)
(
W
/
70)
[
per
mg/(
kg/
day)]
S
1/
3
1/
3
P
1/
3
Water
Unit
Risk
I
/[(
W
)
]
=
2/[(
70)
]
(
I
)/
2
x
[
70/(
W
)]
[
per
µ
g/
l]
W
S
S
2/
3
2/
3
W
P
P
2/
3
Air
Unit
Risk:
I
/[(
W
)
]
=
20/[(
70)
]
(
I
)/
20
x
[
70/(
W
)]
A.
Particles
or
aerosols
[
per
µ
g/
m
],
air
concentration
by
3
weight
A
S
S
2/
3
2/
3
A
P
P
2/
3
Air
Unit
Risk:
No
explicit
proportionality
to
body
1.0
B.
Gases
weight
or
air
intake
is
assumed.
ppm
by
volume
is
assumed
to
be
the
[
per
parts
per
million],
air
effective
dose
in
both
animals
and
concentration
by
volume,
humans.

W
=
Body
weight
(
kg)
a
I
=
Drinking
water
intake
(
liters
per
day)
W
I
=
Air
intake
(
cubic
meters
per
day)
A
W
,
I
,
I
denote
standard
parameters
assumed
by
IRIS
b
S
S,
S
W
A
Modified
risk
measure
=
(
CF)
x
IRIS
value
c
W
,
I
,
I
denote
non­
standard
parameters
of
the
actual
population
P
P
P
W
A
As
one
example
of
the
use
of
Table
1A­
1,
the
recommended
value
for
the
average
consumption
of
tapwater
for
adults
in
the
U.
S.
population
derived
in
this
document
(
Chapter
3),
is
1.4
liters
per
day.
The
drinking
water
unit
risk
for
dichlorvos,
as
given
in
the
IRIS
information
data
base
is
8.3
x
10
per
µ
g/
l,
and
was
calculated
from
the
slope
factor
­
6
assuming
the
standard
intake,
I
,
of
2
liters
per
day.
For
the
United
States
population
drinking
1.4
liters
of
tap
water
per
W
S
Volume
I
­
General
Factors
Appendix
1A
Exposure
Factors
Handbook
Page
August
1997
1A­
5
day
the
corrected
drinking
water
unit
risk
should
be
8.3
x
10
x
(
1.4/
2)
=
5.8
x
10
per
F
g/
l.
The
risk
to
the
average
­
6
­
6
individual
is
then
estimated
by
multiplying
this
by
the
average
concentration
in
units
of
F
g/
l.

Another
example
is
when
the
risk
for
women
drinking
water
contaminated
with
dichlorvos
is
to
be
estimated.
If
the
women
have
an
average
body
weight
of
60
kg,
the
correction
factor
for
the
drinking
water
unit
risk
is
(
disregarding
the
correction
discussed
in
the
above
paragraph),
from
Table
1A­
1,
is
(
70/
60)
=
1.11.
Here
the
ratio
of
70
to
60
is
raised
to
2/
3
the
power
of
2/
3.
The
corrected
water
unit
risk
for
dichlorvos
is
8.3
x
10
x
1.11
=
9.2
x
10
per
F
g/
l.
As
before,
the
risk
­
6
­
6
to
the
average
individual
is
estimated
by
multiplying
this
by
the
water
concentration.

When
human
data
are
used
to
derive
the
risk
measure,
there
is
a
large
variation
in
the
different
data
sets
encountered
in
IRIS,
so
no
generalizations
can
be
made
about
global
corrections.
However,
the
typical
default
exposure
values
used
for
the
air
intake
of
an
air
pollutant
over
an
occupational
lifetime
are:
air
intake
is
10
m
/
day
for
an
8­
hour
shift,
240
days
per
3
year
with
40
years
on
the
job.
If
there
is
continuous
exposure
to
an
ambient
air
pollutant,
the
lifetime
dose
is
usually
calculated
assuming
a
70­
year
lifetime.

3.
CORRECTIONS
FOR
INTAKE
DATA
When
the
body
weight,
W
,
of
the
population
of
interest
differs
from
the
body
weight,
W
,
of
the
population
from
which
P
E
the
exposure
values
in
this
handbook
were
derived,
the
following
model
furnishes
a
reasonable
basis
for
estimating
the
intake
of
food
and
air
(
and
probably
water
also)
in
the
population
of
interest.
Such
a
model
is
needed
in
the
absence
of
data
on
the
dependency
of
intake
on
body
size.
This
occurs
for
inhalation
data,
where
the
intake
data
are
not
normalized
to
body
weight,
whereas
the
model
is
not
needed
for
food
and
tap
water
intakes
if
they
are
given
in
units
of
intake
per
kg
body
weight.

The
model
is
based
on
the
dependency
of
metabolic
oxygen
consumption
on
body
size.
Oxygen
consumption
is
directly
related
to
food
(
calorie)
consumption
and
air
intake
and
indirectly
to
water
intake.
For
mammals
of
a
wide
range
of
species
sizes
(
Prosser
and
Brown,
1961),
and
also
for
individuals
of
various
sizes
within
a
species,
the
oxygen
consumption
and
calorie
(
food)
intake
varies
as
the
body
weight
raised
to
a
power
between
0.65
and
0.75.
A
value
of
0.667
=
2/
3
has
been
used
in
EPA
as
the
default
value
for
adjusting
cross­
species
intakes,
and
the
same
factor
has
been
used
for
intra­
species
intake
adjustments.

[
NOTE:
Following
discussions
by
an
interagency
task
force
(
Federal
Register,
1992),
the
agreement
was
that
a
more
accurate
and
defensible
default
value
would
be
to
choose
the
power
to
3/
4
rather
than
2/
3.
A
recent
article
(
West
et
al.,
1997)
has
provided
a
theoretical
basis
for
the
3/
4
power
scaling.
This
will
be
the
standard
value
to
be
used
in
future
assessments,
and
all
equations
in
this
Appendix
will
be
modified
in
future
risk
assessments.
However,
because
risk
assessors
now
use
the
current
IRIS
information,
this
discussion
is
presented
with
the
previous
default
assumption
of
2/
3].

With
this
model,
the
relation
between
the
daily
air
intake
in
the
population
of
interest,
I
=
(
m
/
day)
,
and
the
intake
A
P
3
P
in
the
population
described
in
this
handbook,
I
=
(
m
/
day)
is:
A
E
3
E
I
=
I
x
(
W
/
W
)
.
A
A
P
E
P
E
2/
3
4.
CALCULATION
OF
RISKS
FOR
AIR
CONTAMINANTS
The
risk
is
calculated
by
multiplying
the
IRIS
air
unit
risk,
corrected
as
described
in
Table
1A­
1,
by
the
air
concentration.
But
since
the
correction
factor
involves
the
intake
in
the
population
of
interest
(
I
),
that
quantity
must
be
A
P
included
in
the
equation,
as
follows:

(
Risk)
=
(
air
unit
risk)
x
(
air
concentration)
P
P
=
(
air
unit
risk)
x
(
I
/
20)
x
(
70/
W
)
x
(
air
concentration)
S
P
P
2/
3
A
Volume
I
­
General
Factors
Appendix
1A
Page
Exposure
Factors
Handbook
1A­
6
August
1997
=
(
air
unit
risk)
x
[(
I
x
(
W
/
W
)
/
20)]
x
(
70/
W
)
x
(
air
concentration)
S
E
P
E
2/
3
P
2/
3
A
=
(
air
unit
risk)
x
(
I
/
20)
x
(
70/
W
)
x
(
air
concentration)
S
E
E
2/
3
A
In
this
equation
the
air
unit
risk
from
the
IRIS
data
base
(
air
unit
risk)
,
the
air
intake
data
in
the
handbook
for
the
S
populations
where
it
is
available
(
I
)
and
the
body
weight
of
that
population
(
W
)
are
included
along
with
the
standard
IRIS
A
E
E
values
of
the
air
intake
(
20
m
/
day)
and
body
weight
(
70
kg).
3
For
food
ingestion
and
tap
water
intake,
if
body
weight­
normalized
intake
values
from
this
handbook
are
used,
the
intake
data
do
not
have
to
be
corrected
as
in
Section
3
above.
In
these
cases,
corrections
to
the
dose­
response
parameters
in
Table
1A­
1
are
sufficient.

5.
REFERENCES
Federal
Register.
(
1992)
Cross­
species
scaling
factor
for
carcinogen
risk
assessments
based
on
equivalence
of
(
mg/
kgday
.
Draft
report.
Federal
Register,
57(
109):
24152­
24173,
June
5,
1992.
3/
4
Prosser,
C.
L.;
Brown,
F.
A.
(
1961)
Comparative
Animal
physiology,
2nd
edition.
WB
Saunders
Co.
p.
161.

U.
S.
EPA.
(
1996)
Background
Documentation.
Integrated
Risk
Information
System
(
IRIS).
Online.
National
Center
for
Environmental
Assessment,
Cincinnati,
Ohio.
Background
Documentation
available
from:
Risk
Information
Hotline,
National
Center
for
Environmental
Assessment,
U.
S.
EPA,
26
W.
Martin
Luther
King
Dr.
Cincinnati,
OH
45268.
(
513)
569­
7254
West,
G.
B.;
Brown,
J.
H.;
Enquist,
B.
J.
(
1997)
A
general
model
of
the
origin
of
allometric
scaling
laws
in
biology.
Science
276:
122­
126.
