United States Technical Review FINAL (December 1996)
Environmental Protection Workgroup EPA-540-R-03-001
Agency for Lead January 2003
Recommendations of the
Technical Review Workgroup for Lead for an
Approach to Assessing Risks Associated with Adult
Exposures to Lead in Soil
-------
-------
Preface
This report was updated in 2003 to add two appendices and remove "interim" from the title. The
change in the title reflects publication of the TRW's Adult Lead Model Review Report. With the
exception of the following, the guidance is unchanged from the December 1996 publication.
The report now includes an appendix showing the format for the spreadsheet form of the model
(Appendix B) and an explanation of how the guidance is to be applied (Appendix C). The NHANES
Report (March 2002) should be used in conjunction with this report. Based on the TRW's analysis
of the data collected in the completed NHANES III survey (Phases 1 and 2), updated ranges for the
baseline adult PbB and GSDi adult parameters should be in the EPA ALM spreadsheet. Although
the use of these updated ranges in the EPA ALM spreadsheet would not appreciably change PRGs
calculated with the methodology, it is recommended that data from both phases of NHANES III be
used in all PbB analyses; this is consistent with CDC recommendations. The results of the
NHANES III Report are not included in this document except by reference.
-------
-------
U.S. Environmental Protection Agency
Technical Review Workgroup for Lead
CHAIRPERSONS
Patricia Van Leeuwen
Region 5
Chicago, IL
Paul White
Office of Research and Development
Washington, DC
MEMBERS
Harlal Choudhury
Office of Research and Development
Cincinnati, OH
Barbara Davis
Office of Solid Waste and
Emergency Response
Washington, DC
Robert Elias
Office of Research and Development
Research Triangle Park, NC
Susan Griffin
Region 8
Denver, CO
Karen Hogan
Office of Prevention, Pesticides
and Toxic Substances
Washington, DC
Mark Maddaloni
Region 2
New York, NY
Allan Marcus
Office of Research and Development
Research Triangle Park, NC
Chris Weis
Region 8
Denver, CO
Larry Zaragoza
Office of Solid Waste and
Emergency Response
Washington, DC
in
-------
-------
Adult Lead Risk Assessment Committee
of the
Technical Review Workgroup for Lead
CHAIRPERSON
Mark Maddaloni
Region 2
New York, NY
MEMBERS
Mary Ballew
Region 1
Boston, MA
Cherri Baysinger-Daniel
Missouri Department of Health
Jefferson City, MO
Mark Johnson
Region 5
Chicago, IL
Kevin Koporec
Region 4
Atlanta, GA
Roseanne Lorenzana
Region 10
Seattle, WA
Margaret McDonough
Region 1
Boston, MA
Patricia Van Leeuwen
Region 5
Chicago, IL
Chris Weis
Region 8
Denver, CO
Paul White
Office of Research and Development
Washington, DC
Larry Zaragoza
Office of Solid Waste and
Emergency Response
Washington, DC
-------
-------
1. INTRODUCTION
This report describes a methodology for assessing risks associated with non-residential adult
exposures to lead in soil. The methodology focuses on estimating fetal blood lead concentration in
women exposed to lead contaminated soils. This approach also provides tools that can be used for
evaluating risks of elevated blood lead concentrations among exposed adults. The methodology
is the product of extensive evaluations by the Technical Review Workgroup for Lead (TRW) which
began considering methodologies to evaluate nonresidential adult exposure in 1994 (Balbus-
Kornfeld, 1994; U.S. EPA, 1994a). In 1995, the TRW reviewed a methodology developed by EPA
Region 8 for deriving risk-based remediation goals (RBRGs) for nonresidential soil atthe California
Gulch NPL site (U.S. EPA, 1995). A TRW committee on adult lead risk assessment was formed
in January, 1996 to further develop the ideas and information gathered as part of these previous
efforts into a generic methodology that could be adapted for use in site-specific assessments.
This report provides technical recommendations of the TRW for the assessment of adult lead
risks using this methodology. An overriding objective in the development of this methodology was
the immediate need for a scientifically defensible approach for assessing adult lead risks associated
with nonresidential exposure scenarios. The TRW recognizes that other adult lead models may
provide useful information. In particular, models providing more detailed representations of lead
kinetics may be useful in supporting more detailed predictions about the time course of blood lead
concentrations among individuals who receive brief acute exposures to lead or whose exposures
otherwise change markedly with time. The methodology presented here uses a simplified
representation of lead biokinetics to predict quasi-steady state blood lead concentrations among
adults who have relatively steady patterns of site exposures (as described in this report). The TRW
believes that this approach will prove useful for assessing most sites where places of employment
are (or will be) situated on lead contaminated soils. This information is expected to promote
consistency in assessments of adult lead risks. The methodology described in this report is an
approach that is recommended for use pending further development and evaluation of integrated
exposure biokinetic models for adults. The TRW is undertaking review of other models and will
provide reviews on other approaches as appropriate. The Integrated Exposure Uptake Biokinetic
(IEUBK) Model for Lead in Children (U.S. EPA, 1994b,c) is the recommended approach for
assessing residential lead risks.
The recommended approach for assessing nonresidential adult risks utilizes a methodology
to relate soil lead intake to blood lead concentrations in women of child-bearing age. It is
conceptually similar to a slope factor approach for deriving RBRGs that had been proposed by
Bowers et al. (1994) and which was adapted for use at the California Gulch NPL site in Region 8
(U.S. EPA, 1995). This report describes the basic algorithms that are used in the methodology and
provides a set of default parameter values that can be used in cases where high quality data are not
available to support site-specific estimates. The rationale for each parameter default value is
provided in the Appendix.
-------
2. OVERVIEW OF THE APPROACH
The methodology described in this report relates soil lead concentrations to blood lead
concentrations in the exposed population according to the algorithms described below. Note that
the algorithms may consist of variables that include superscripts and/or subscripts. The convention
adopted in this report is to use superscripts as exponents (i.e., a mathematical operation), whereas
subscripts represent key words that provide additional information to distinguish between similar
variables. The basis for the calculation of the blood lead concentration in women of child-bearing
age is the algorithm given by Equation 1:
PbS'BKSF-IR-'AF-'EF-
~^ (Equation 1)
where:
PbBadult central = Central estimate of blood lead concentrations (|ig/dL) in adults (i.e., women of
child-bearing age) that have site exposures to soil lead at concentration, PbS.
PbBadult 0 = Typical blood lead concentration (|ig/dL) in adults (i.e., women of child-bearing
age) in the absence of exposures to the site that is being assessed.
PbS = Soil lead concentration (i-ig/g) (appropriate average concentration for individual).
BKSF= Biokinetic slope factor relating (quasi-steady state) increase in typical adult blood
lead concentration to average daily lead uptake (i-ig/dL blood lead increase per [ig/day
lead uptake).
IRS = Intake rate of soil, including both outdoor soil and indoor soil-derived dust
(g/day).
AFS = Absolute gastrointestinal absorption fraction for ingested lead in soil and lead in
dust derived from soil (dimensionless).
EFS = Exposure frequency for contact with assessed soils and/or dust derived in part
from these soils (days of exposure during the averaging period); may be taken as
days per year for continuing, long term exposure.
AT = Averaging time; the total period during which soil contact may occur; 365
days/year for continuing long term exposures.
The basis for the RBRG calculation is the relationship between the soil lead concentration and
the blood lead concentration in the developing fetus of adult women that have site exposures. As
a health-based goal, EPA has sought to limit the risk to young children of having elevated blood lead
concentrations. Current Office of Solid Waste and Emergency Response (OSWER) guidance calls
-------
for the establishment of cleanup goals to limit childhood risk of exceeding 10 |ig/dL to 5% (U.S.
EPA, 1994a). Equation 2 describes the estimated relationship between the blood lead concentration
in adult women and the corresponding 95th percentile fetal blood lead concentration (PbB fetal 0 95),
assuming that PbBadult central reflects the geometric mean of a lognormal distribution of blood lead
concentrations in women of child-bearing age. If a similar 95th percentile goal is applied to the
protection of fetuses carried by women who experience nonresidential exposures, Equation 2 can
be rearranged to reflect a risk-based goal for the central estimate of blood lead concentrations in
adult women using Equation 3:
PbBfetal,0.95 = PbBadult,central ' GSDl'a£lt ' ^fetal/maternal (Equation 2)
= PbBfetal,0.95,goal
adult,central,goal j 645 (Equation 3)
^fetal/maternal
where:
adult, central goal = Goal ^or central estimate of blood lead concentration (|ig/dL) in adults (i.e.,
women of child-bearing age) that have site exposures. The goal is intended to
ensure that PbBfetal 0 95 goal does not exceed 10 |ig/dL.
feta] 0 95 goal = Goal for the 95th percentile blood lead concentration (|ig/dL) among fetuses
born to women having exposures to the specified site soil concentration. This
is interpreted to mean that there is a 95% likelihood that a fetus, in a woman
who experiences such exposures, would have a blood lead concentration no
greater than PbBfetal 0 95 goal (i.e., the likelihood of a blood lead concentration
greater than 10 |ig/dL would be less than 5%,for the approach described in this
report).
GSD; adult = Estimated value of the individual geometric standard deviation (dimensionless);
the GSD among adults (i.e., women of child-bearing age) that have exposures
to similar on-site lead concentrations, but that have non-uniform response
(intake, biokinetics) to site lead and non-uniform off-site lead exposures. The
exponent, 1 .645, is the value of the standard normal deviate used to calculate the
95th percentile from a lognormal distribution of blood lead concentration.
-------
R fetai/matemai = Constant of proportionality between fetal blood lead concentration at birth and
maternal blood lead concentration (dimensionless).
The soil lead concentration associated with a given exposure scenario and PbB adult central goal can
be calculated by rearranging Equation 1 and substituting PbB ^central, goal for PbBaduit! central :
^""^adult,centrca,goal
- (Frmatinn
(BKSF'IRS'AF.-EF^ Aquation
It is this form of the algorithm that can be used to calculate a RBRG where the RBRG represents the
soil lead concentration (PbS) that would be expected to result in a specified adult blood lead
concentration (PbB adult central ,) and corresponding 95th percentile fetal blood lead concentration
(PbB fetal, 0.95, goal)-
Equations 1-4 are based on the following assumptions:
1. Blood lead concentrations for exposed adults can be estimated as the sum of an
expected starting blood lead concentration in the absence of site exposure (PbBadult
0) and an expected site-related increase.
2. The site-related increase in blood lead concentrations can be estimated using a linear
biokinetic slope factor (BKSF) which is multiplied by the estimated lead uptake.
3. Lead uptake can be related to soil lead levels using the estimated soil lead
concentration (PbS), the overall rate of daily soil ingestion (IRS), and the estimated
fractional absorption of ingested lead (AFS) The term "soil" is used throughout this
document to refer to that portion of the soil to which adults are most likely to be
exposed. In most cases, exposure is assumed to be predominantly to the top layers
of the soil which gives rise to transportable soil-derived dust. Exposure to soil-
derived dust occurs both in outdoor and indoor environments, the latter occurring
where soil-derived dust has been transported indoors. Other types of dust, in addition
to soil-derived dust, can contribute to adult lead exposure and may even predominate
in the occupational setting; these include dust generated from manufacturing
processes (e.g., grinding, milling, packaging of lead-containing material), road dust,
pavement dust, and paint dust. This methodology, as represented in Equations 1 and
4, does not specifically account for site exposure to dusts that are not derived from
soil. However, the methodology can be modified to include separate variables that
represent exposure to lead in various types of dust. This approach is discussed in
greater detail in the Appendix.
-------
4. As noted above, exposure to lead in soil may occur by ingesting soil-derived dust in
the outdoor and/or indoor environments. The default value recommended for IRS
(0.05 g/day) is intended for occupational exposures that occur predominantly indoors.
More intensive soil contact would be expected for predominantly outdoor activities
such as construction, excavation, yard work, and gardening.
5. A lognormal model can be used to estimate the inter-individual variability in blood
lead concentrations (i.e., the distribution of blood lead concentrations in a population
of individuals who contact similar environmental lead levels).
6. Expected fetal blood lead concentrations are proportional to maternal blood lead
concentrations.
The primary basis for using Equation 4 to calculate a RBRG is that fetuses and neonates are
a highly sensitive population with respect to the adverse effects of lead on development and that 10
1-ig/dL is considered to be a blood lead level of concern from the standpoint of protecting the health
of sensitive populations (U.S. EPA, 1986, 1990; NRC, 1993). Therefore, risk to the fetus can be
estimated from the probability distribution of fetal blood lead concentrations (i.e., the probability
of exceeding 10 |ig/dL), as has been the approach taken for estimating risks to children (U.S. EPA,
1994a,c). Equation 4 can be used to estimate the soil lead concentration at which the probability of
blood lead concentrations exceeding a given value (e.g., 10 |ig/dL) in fetuses of women exposed to
environmental lead is no greater than a specified value (e.g., 0.05).
The methodology can be modified to accommodate different assumptions or to estimate
RBRGs for different risk categories. For example, a RBRG could be estimated for risks to adults
(e.g., hypertension) by substituting an appropriate adult blood lead concentration benchmark.
Similarly, other exposure scenarios can be incorporated into the assessment. Alternative methods
for estimating soil lead risk by partitioning soil into outdoor soil and indoor dust components are
discussed in the Appendix.
Recommended default values for each of the parameters in Equations 1 - 4 are presented in
Table 1. These defaults should not be casually replaced with other values unless the alternatives are
supported by high quality site-specific data to which appropriate statistical analyses have been
applied and that have undergone thorough scientific review. Examples of the output from the
methodology are presented in Figures 1 and 2, which show plots of the calculated PbBfetal 0 95 as a
function of PbS when different combinations of default parameter values are used. The rationale
for each default value listed in Table 1 is summarized in the Appendix.
-------
Table 1. Summary of Default Parameter Values for the Risk Estimation Algorithm (Equations 1-4)
Parameter
PbBfetal, 0.95,goal
GSDiiadult
^-fetal/maternal
°t>Badulti0
BKSF
IRS
EFS
AFS
Unit
H-g/dL
~
~
|ig/dL
Hg/dL
per
l-ig/day
g/day
day/yr
~
Value
10
1.8
2.1
0.9
1.7-2.2
0.4
0.05
219
0.12
Comment
For estimating RBRGs based on risk to the developing fetus.
Value of 1.8 is recommended for a homogeneous population while 2.1 is recommended for
a more heterogeneous population.
Based on Goyer (1990) and Graziano et al. (1990).
Plausible range based on NHANES III phase 1 for Mexican American and non-Hispanic
black, and white women of child bearing age (Brody et al. 1994). Point estimate should be
selected based on site-specific demographics.
Based on analysis of Pocock et al. (1983) and Sherlock et al. (1984) data.
Predominantly occupational exposures to indoor soil-derived dust rather than outdoor soil;
(0.05 g/day = 50 mg/day).
Based on U.S. EPA (1993) guidance for average time spent at work by both full-time and
part-time workers (see Appendix for recommendations on minimum exposure frequency
and duration).
Based on an absorption factor for soluble lead of 0.20 and a relative bioavailability of 0.6
(soil/soluble).
-------
500
1000 1500
PbS(|jg/g)
2000
2500
Figure 1. Example output of risk estimation algorithm (Equation 4) assuming a PbBadult 0 of 2.0
1-ig/dL (mixed racial) and a GSD; adult of either 1.8 (homogeneous population) or 2.1 (heterogeneous
urban population).
-------
500
1000 1500
PbS(|jg/g)
2000
2500
Figure 2. Example output of risk estimation algorithm (Equation 4) assuming plausible default
minimum and maximum values of PbBadult 0 (1.7 and 2.2 pg/dL) and GSD; adult (1.8 and 2.1).
-------
3. REFERENCES
Balbus-Kornfeld, J. 1994. Comments and Recommendations on the Draft Interim Guidance for
Screening Levels of Lead in Soil for Non-Residential Sites. Letter from John Balbus-Kornfeld to
Bruce Means. November 17, 1994.
Bowers, T. S., B.D. Beck and H. S. Karam. 1994. Assessing the relationship between environmental
lead concentrations and adult blood lead levels. Risk Analysis. 14(2): 183-189.
Brody, D.J., J.L. Pirkle, R.A. Kramer, K.M. Flegal, T.D. Matte, E.W. Gunter and D.C. Paschal.
1994. Blood lead levels in the U.S. population. Phase 1 of the third National Health and Nutrition
Examination Survey (NHANES III, 1988 to 1991). JAMA. 272(4): 277-283.
Goyer, R.A. 1990. Transplacental transport of lead. Environ. Health Perspect. 89: 101-105.
Graziano, J.H., D. Popovac, P. Factor-Litvak, P. Shrout, J. Kline, M.J. Murphy, Y. Zhao, A.
Mehmeti, X. Ahmedi, B. Rajovic, Z. Zvicer, D. Nenezic, N. Lolacono and Z. Stein. 1990.
Determinants of elevated blood lead during pregnancy in a population surrounding a lead smelter
in Kosovo, Yugoslavia. Environ. Health Perspect. 89:95-100.
NRC. 1993. Measuring Lead Exposure in Infants, Children and Other Sensitive Populations.
National Academy Press. Washington, DC. ISBN 0-309-04927-X.
Pocock, S.J., A.G. Shaper, M. Walker, C.J. Wale, B. Clayton, T. Delves, R.F. Lacey, R.F. Packham
and P. Powell. 1983. Effects of tap water lead, water hardness, alcohol, and cigarettes on blood lead
concentrations. J. Epi. Comm. Health. 37: 1-7.
Sherlock, J.C., D. Ashby, H.T. Delves, G.I. Forbes, M.R. Moore, W.J. Patterson, S.J. Pocock, M.J.
Quinn, W.N. Richards and T.S. Wilson. 1984. Reduction in exposure to lead from drinking water
and its effect on blood lead concentrations. Human Toxicol. 3: 383-392.
U.S. EPA. 1986. Air Quality Criteria for Lead Volumes I - IV. Environmental Criteria and
Assessment Office, Office of Research and Development, RTF, NC. EPA 600/8-83-028 a-d.
U.S. EPA. 1990. Supplement to the 1986 EPA Air Quality Criteria Document for Lead - Volume
1 Addendum. Office of Research and Development, Office of Health and Environmental
Assessment, Washington, DC. EPA-600/8-89/049A.
U.S. EPA. 1993. Superfund's Standard Default Exposure Factors for the Central Tendency and
RME-Draft. Working Draft, November 1993.
U.S. EPA. 1994a. Revised Interim Soil Lead Guidance for CERCLA Sites and RCRA Corrective
Action Facilities. OSWER Directive No. 9355.4-12. Office of Emergency and Remedial Response,
Washington, D.C. EPA/540/F-94/043, PB94-963282.
-------
U.S. EPA. 1994b. Technical Support Document: Parameters and Equations Used in the Integrated
Exposure Uptake Biokinetic Model for Lead in Children (v. 0.99d). Office of Emergency and
Remedial Response, Washington, D.C. EPA/540/R-94/040, PB94-963505.
U. S. EPA. 1994c. Guidance Manual for the Integrated Exposure Uptake Biokinetic Model for Lead
in Children. Office of Emergency and Remedial Response, Washington, D.C. EPA/540/R-93/081,
PB93-963510.
U.S. EPA. 1995. A TRW Report: Review of a Methodology for Establishing Risk-Based Soil
Remediation Goals for the Commercial Areas of the California Gulch Site. Technical Review
Workgroup for Lead, October, 1995.
10
-------
APPENDIX A
Equations and Rationale for Default Values
Assigned to Parameters in the Slope Factor Approach and
Exposure Model for Assessing Risk Associated with Adult
Exposures to Lead in Soil
-------
-------
Equations and Rationale for Default Values Assigned to Parameters in the
Slope Factor Approach and Exposure Model for Assessing Risk Associated
with Adult Exposures to Lead in Soil
1. Equations for the Adult Lead Model A-3
2. Individual Blood Lead Geometric Standard Deviation (GSD;) A-6
3. Fetal/Maternal Blood Lead Concentration Ratio (Rfetai/maternai) A-8
4. Baseline Blood Lead Concentration (PbBadult 0) A-8
5. Biokinetic Slope Factor (BKSF) A-10
6. Soil Lead Absorption Factor (AFS) A-15
7. Daily Soil Ingestion Rate (IRS) A-19
8. Exposure Frequency (EFS) A-22
9. Applying Monte Carlo Analysis to the Adult Lead Methodology A-23
10. References A-25
A-l
-------
-------
1. Equations for the Adult Lead Model
The format of the equations used in the adult lead methodology follows the approach used
in the IEUBK Model for Lead in Children (IEUBK Model). Note that the equations may consist of
variables that include superscripts and/or subscripts. The convention adopted in this report is to use
superscripts as exponents (i.e., a mathematical operation), whereas subscripts represent key words
that provide additional information to distinguish between similar variables. The term "soil" refers
to that portion of the soil to which adults are most likely to be exposed. In most cases, exposure is
assumed to be predominantly to the top layers of the soil which gives rise to transportable soil-
derived dust. Exposure to soil-derived dust occurs both in outdoor and indoor environments, the
latter occurring where soil-derived dust has been transported indoors. Other types of dust, in
addition to soil-derived dust, can contribute to adult lead exposure and may even predominate in
some occupational settings; these include dust generated from manufacturing processes (e.g.,
grinding, milling, packaging of lead-containing material), road dust, pavement dust, and paint dust.
Exposure to lead from soil (direct and through indoor soil-derived dust) and lead
intake:
PbS IR~ EF,
INTAKE = (Equation A-1)
J*. L
INTAKE = Daily average intake (ingestion) of lead from soil taken over averaging time AT
(Hg/day).
PbS = Soil lead concentration (i-ig/g) (appropriate average concentration for individual).
IRS = Intake rate of soil, including outdoor soil and indoor soil-derived dust (g/day).
EFS = Exposure frequency for contact with assessed soils and/or dust derived in part from
these soils (days of exposure during the averaging period); may be taken as days per
year for continuing, long term exposures.
AT = Averaging time; the total period during which soil contact may occur; 365 days/year for
continuing long term exposures.
Lead uptake:
UPTAKE = AFS INTAKE (Equation A-2)
A-3
-------
UPTAKE = Daily average uptake of lead from the gastrointestinal tract into the systemic
circulation (|j,g/day).
AFS = Absolute gastrointestinal absorption fraction for ingested lead in soil and lead in dust
derived from soil (dimensionless).
Central estimate of adult blood lead concentration:
PbBaduit,centrai = PbBadult + BKSF-UPTAKE (Equation A-3)
PbBadult central = Central estimate of blood lead concentrations (|ig/dL) in adults (i.e., women of
child-bearing age) that have site exposures to soil lead at concentration, PbS.
PbBadulto = Typical blood lead concentration (|ig/dL) in adults (i.e., women of child-bearing
age) in the absence of exposures to the site that is being assessed.
BKSF = Biokinetic slope factor relating (quasi-steady state) increase in typical adult
blood lead concentration to average daily lead uptake (i-ig/dL blood lead increase
per [ig/day lead uptake).
Distributional model for adult blood lead:
In this methodology, variability in blood lead concentrations among a population is
mathematically described by a lognormal distribution defined by two parameters, the geometric
mean (GM) and the geometric standard deviation (GSD):
PbBadult ~ Lognormal(GM,GSD)
PbBadult= Adult blood lead concentration (which is a variable quantity having the specified
probability distribution).
GM = Geometric mean blood lead concentration (|ig/dL) for adults having site exposure.
The central estimate of adult blood lead, PbBadult central, constructed in Equation A-3
is treated as a plausible estimate of the geometric mean.
GSD = Geometric standard deviation for blood lead concentrations among adults having
exposures to similar on-site lead concentrations, but having non-uniform response
(intake, biokinetics) to site lead and non-uniform off-site lead exposures. The
individual blood lead concentration geometric standard deviation, GSD;, is
A-4
-------
substituted for GSD. As described below (Section 2 of the Appendix), GSD; is
assumed to address sources of variability in blood lead concentrations among the
exposed population.
Parameter estimates for the geometric mean (GM) and geometric standard deviation (GSD) of the
lognormal distribution are described below. Note that blood lead concentrations for site exposures
can be quantified at any percentile of the population using these parameters. For example, the 95th
percentile blood lead concentration can be calculated by Equation A-4:
PbBadult,0.95 = PbB astral GSD,1*5 (Equation A-4)
PbBadult095= 95th percentile blood lead concentration (|ig/dL) among individuals having
exposures to the specified site soil lead concentrations. This is interpreted to mean
that there is a 95% likelihood that an adult exposed to the specified soil lead
concentrations would have a blood lead concentration less than or equal to PbBadult 0 95
Distributional model for fetal blood lead:
PbBfetal ~ Rfetal!maternal ' PbB adult (Equation A-5)
PbBfetal = Fetal blood lead concentration (|ig/dL) (which, like PbBadult, is a variable quantity
having the specified probability distribution).
Rfetai/matemai = Constant of proportionality between fetal and maternal blood lead concentrations.
PbBadult= Adult blood lead concentration (|ig/dL), estimated with parameters appropriate to women
of child bearing age.
Note that this relationship implies a deterministic (non-random) relationship between maternal and
fetal blood lead concentrations. This assumption omits a source of variability (varying individual-
specific ratios of fetal to maternal blood lead) that would tend to increase the variance of fetal blood
lead concentrations. The assumption of proportionality implies that fetal blood lead concentrations
also are lognormally distributed:
PbBfetal ~ Lognormal(GM,GSD)
A-5
-------
GM = Geometric mean blood lead concentration (|ig/dL) for fetuses, equal to Rfetai/maternai
multiplied by PbBadult central.
GSD = Geometric standard deviation of blood lead concentration among adults, GSD;
(Section 2 of the Appendix).
Similarly, percentiles of the fetal blood lead distribution can be estimated (for fetuses carried by
women exposed to the specified concentration of lead at the assessed site). For example:
PAR If PAR
r0£>fetal,0.95 ~ ^fetal/maternal f OD adult,central
i,adult
(Equation A-6)
PbBfetal095 = 95th percentile blood lead concentration (|ig/dL) among fetuses born to women
having exposures to the specified site soil lead concentrations. This is interpreted
to mean that there is a 95% likelihood that a fetus born, in a woman who experiences
such exposures, would have a blood lead concentration no greater than PbBfetal 0 95.
Note that when the expressions for PbBadult central, INTAKE, and UPTAKE (Equations A-1, A-2 and
A-3) are substituted into Equation A-6, we obtain the complete expression for PbBfetal095 that is
presented in the fact sheet (Overview of the Approach, Equations 1 and 2):
PbB
'fetal,0.95
^-
-fetal/maternal
1.645
(PbS-BKSF-IRs-AFs-EFs)
AT
+ PbB
'adult,0
(Equation A-7)
Equation A-7 represents variability in blood lead concentration arising from two main factors: 1)
exposure variables, including inter-individual variability in activity-weighted ingestion rates, and
2) inter-individual variability in physiology, including factors affecting lead biokinetics.
2. Individual Blood Lead Geometric Standard Deviation (GSDj)
The GSD; is a measure of the inter-individual variability in blood lead concentrations in a
population whose members are exposed to the same nonresidential environmental lead levels.
Ideally, the value(s) for GSD; used in the methodology should be estimated in the population of
concern at the site. This requires data on blood lead concentration and exposure in a representative
sample of sufficient size to yield statistically meaningful estimates of GSD in subsamples stratified
by nonresidential exposure level. In the absence of high quality data for the site, GSD; may be
extrapolated from estimates for other surrogate populations. In making such extrapolations, factors
that might contribute to higher or lower variability in the surrogate population than among similarly
exposed individuals in the population of concern, should be evaluated. These factors include
variability in exposure (level and pathways), and biokinetics (see Section 6 of Appendix),
socioeconomic and ethnic characteristics, degree of urbanization and geographical location. Such
A-6
-------
extrapolations, therefore, are site-specific and are a potentially important source of uncertainty in
the methodology.
GSD values measured in populations (GSDp) reflect the combined effect of 1) variability in
environmental concentration levels; and 2) activity-weighted exposures and lead biokinetics. Thus,
estimates of GSDp can be considered a surrogate for estimating the GSD;. Site data on blood lead
concentrations collected from populations of varying homogeneity may be useful for establishing
a plausible range of values of GSD; , provided that the data are of adequate quality and can be
stratified by nonresidential exposure level. The lowest values of GSDp are expected among
homogeneous populations (e.g., individuals with similar socioeconomic and ethnic characteristics
living within a relatively small geographic area) exposed to a single, dominant source of lead (e.g.,
lead mining or smelter sites). For example, a GSDp of 1.8 was recently calculated among adult
women living in Leadville, CO (U.S. EPA, 1995). This relatively low GSD is consistent with an
analysis of blood lead concentration data in mining communities in the United States and Canada,
which suggest that GSDp ranges from 1.6 - 1.8 at active mining sites where blood lead
concentrations are less than 15 |ig/dL (U.S. EPA, 1992). By contrast, higher values of GSDp might
be expected from a national survey. Although lead exposures among the general population are
likely to be more greatly impacted by diet than soil (e.g., compared with populations exposed at a
waste site), the national population is very heterogeneous, in that it includes individuals with
different socioeconomic and ethnic characteristics living in distinct geographic areas.
The TRW has conducted a preliminary analysis of blood lead concentration data collected in
NHANES III Phase 1 from 1988 to 1991 and found that the GSDp for women ages 17 to 45 years
may range from 1.9-2.1 (Table A-l). Because of the complex survey design used in NHANES III
(e.g., large oversampling of young children, older persons, black persons, and Mexican-Americans),
this analysis used sampling weights included in the NHANES III Phase 1 data file to produce
population estimates for blood lead concentration. The weighting factor "WTPEXMH1" was used
to reflect the non-random sampling of individuals in both the mobile examination units (MEC) and
the home examinations. The analysis did not account for the design effects associated with the
selection of strata and primary sampling units (PSUs), which may result in an underestimation of
sampling variance. Since this bias is not likely to greatly impact the GSDp (Brody, personal
communication), the amount of underestimation of the GSDp by the values given in Table A-l is
likely to be small. Geometric mean blood lead concentrations listed in Table A-l are within 0.2
1-ig/dL of these reported in Brody et al. (1994).
The TRW estimates that 1.8 - 2.1 is a plausible range for GSD;, based on an evaluation of
available blood lead concentration data for different types of populations. In cases where site-
specific data are not available, a value within this range should be selected based on an assessment
as to whether the population at the site would be expected to be more or less heterogeneous than the
U.S. population with respect to racial, ethnic, cultural and socioeconomic factors that may affect
exposure.
A-7
-------
Table A-l. NHANES III Phase 1 Summary Statistics for Blood Lead Concentration Among U.S.
Women by Age and Ethnic/Racial Characteristicsa.
Age Group
(years)
20-49
50-69
>69
20 +
17-45
Non-Hispanic White
No. GM GSD
728 1.9 1.90
476 3.2 1.88
562 3.5 1.82
1,766 2.4 2.01
742 1.7 1.89
Non-Hispanic Black
No. GM GSD
622 2.3 2.01
256 4.2 1.80
135 4.1 1.86
1,013 2.7 2.07
658 2.1 1.98
Mexican American
No. GM GSD
729 2.1 2.10
255 3.3 2.12
75 2.9 2.03
1,059 2.3 2.14
763 2.0 2.10
"Analysis of data weighted by MEC and home weighting factor (WTPEXMH1), excluding samples
missing data on blood lead concentration or age. GM PbB (|ig/dL) = exp(|iln); GSD PbB = exp(oln).
3. Fetal/Maternal Blood Lead Concentration Ratio
The TRW recommends a default value of 0.9 based on studies that have explored the relationship
between umbilical cord and maternal blood lead concentrations (Goyer, 1990; Graziano et al., 1990).
The Goyer (1990) estimate of an average fetal/maternal blood lead concentration ratio of 0.9 is
supported by a large body of data that has been summarized in Agency documents (U. S. EPA, 1 986,
1990). Graziano et al. (1990) compared maternal and umbilical cord blood lead concentrations at
delivery in 888 mother-infant pairs who were between 28 and 44 weeks of gestation. The
relationship was linear with a slope of 0.93 |ig/dL cord blood per |ig/dL maternal blood; the
correlation coefficient was 0.92. The slope of 0.93 from the Graziano et al. (1990) study supports
0.9 as a point estimate for Rfetal/matemal.
Although average fetal/maternal blood lead concentration ratios, as reflected in cord blood, tend
to show consistent trends (Goyer, 1990; Graziano et al., 1990), the trends may not reflect significant
inter-individual variability in maternal and possibly fetal blood lead concentrations due to
physiological changes associated with pregnancy. For example, mobilization of bone lead stores
during pregnancy may be more substantial in some women, and iron and calcium deficiency
associated with poor nutritional status, as well as pregnancy, may enhance gastrointestinal
absorption of lead (U.S. EPA, 1990; Franklin et al., 1995). Conversely, maternal blood lead
concentration may decrease during the later stages of pregnancy because of the dilution effect
associated with a 30% rise in plasma volume, as well as an increased rate of transfer of lead to the
placenta or to fetal tissues (Alexander and Delves, 1981). These changes may give rise to
fetal/maternal blood lead concentration ratios that are different from 0.9.
4. Baseline Blood Lead Concentration (PbBadulto)
The baseline blood lead concentration (PbBadulto) is intended to represent the best estimate of a
reasonable central value of blood lead concentration in women of child-bearing age who are not
exposed to lead-contaminated nonresidential soil or dust at the site. In this analysis, geometric mean
blood lead concentrations are used for this purpose. Ideally, the value(s) for PbBadulto used in the
A-8
-------
methodology should be estimated in the population of concern at the site. This requires data on
blood lead concentrations in a representative sample of adult women who are not exposed to
nonresidential soil or soil-derived dust at the site, but who may experience exposures to other
environmental sources of lead that are similar in magnitude to exposures experienced by the
population of concern. This would include exposure to lead in food and drinking water as well as
residential soil and dust (dust derived from soil and all other non-site related sources). The sample
must be of sufficient size to yield statistically meaningful estimates of PbBadulto.
In the absence of high quality data for the site, PbBadult 0 may be extrapolated from estimates for
other surrogate populations that would be expected to have a similar PbBadult 0 distribution as that of
the population of concern. In making such extrapolations, factors that might contribute to
differences between the geometric mean PbBadulto in the surrogate population and population of
concern should be evaluated. These factors include differences in the residential exposure (level and
pathways), socioeconomic, ethnic and racial demographics, housing stock, degree of urbanization,
and geographical location. Such extrapolations, therefore, are site-specific.
In cases where site-specific extrapolations from surrogate populations are not feasible, the TRW
recommends 1.7 - 2.2 |ig/dL as a plausible range, based on the results of Phase 1 of the NHANES
III as reported by Brody et al. (1994). Table A-2 summarizes the analysis of blood lead
concentrations from a sample of 2,083 women ages 20 - 49, and stratified into the three ethnic and
racial categories.
Table A-2. NHANES III Phase 1 Summary Statistics for Blood Lead
Concentration Among Different Populations of U.S. Women Ages 20 - 49
(Brody etal., 1994).
Population
Mexican American women
non-Hispanic black women
non-Hispanic white women
Total
No.
732
623
728
2,083
GM (95% CI)
2.0(1.7-2.5)
2.2(2.0-2.5)
1.7(1.6- 1.9)
The TRW recommends that the estimates from Table A-2 be used in combination with data on the
ethnic and racial demographics of the population of concern to select the most appropriate point
estimate from within the plausible range of 1.7 - 2.2 |ig/dL. For example, if the population at the
site was predominantly Mexican American, 2.0 |ig/dL might be selected as the point estimate. The
plausible range is based on surveys of large samples of the national population and may not
encompass central tendencies estimated from smaller regional or site-specific surveys, either
because of bias associated with the smaller sample or because of real differences between the
surveyed population and the national population. This needs to be evaluated in deciding whether
or not to use data from small surveys that yield point estimates for PbBadulto that fall outside of the
plausible range.
A-9
-------
5. Biokinetic Slope Factor (BKSF)
The BKSF parameter relates the blood lead concentration (|ig Pb/dL) to lead uptake (|ig
Pb/day). The TRW recommends a default value of 0.4 |ig Pb/dL blood per |ig Pb absorbed/day for
the BKSF parameter based on data reported by Pocock et al. (1983) on the relationship between tap
water lead concentrations and blood lead concentrations for a sample of adult males, and on
estimates of the bioavailability of lead in tap water (see Section 6 of the Appendix).
Pocock et al. (1983) analyzed data on lead concentrations in first draw tap water and blood
lead concentrations in a population of 910 adult males. A linear model imposed on the data yielded
a slope of 0.06 (|ig/dL per |ig/L first draw water) for water lead concentrations equal to or less than
100 |ig/L (a lower slope was applied to the data for higher water concentrations). Pocock et al.
(1983) also obtained data on lead concentrations in flushed water (and "random daytime") samples,
in addition to first draw samples. Given the following assumptions, it is possible to derive a slope
factor for ingested water lead (INGSF) from the Pocock et al. (1983) data:
The lead concentration of flushed water was 25% of the concentration of first draw water
(CPlst = 0.25) (U.S. EPA, 1995).
Daily water intake consisted of 30% first draw and 70% flushed (Flst = 0.3, Ff= 0.7) (U.S.
EPA, 1992).
Daily water ingestion (including tap water and beverages made with tap water) was 1 .4
L/day (IRW= 1.4) (U.S. EPA, 1989).
Based on the above assumptions, a INGSF of 0.09 |ig/dL per |ig intake/day is estimated as follows:
INGSF = _ 0.06 _ (Equation A-8)
INGSF =
1.4 -(0.3 + (0.25-0.7))
INGSF = 0.09
This suggests that the product of the BKSF, reflecting the slope for absorbed rather than ingested
lead, and the absorption factor for lead in drinking water (AFW) should be approximately 0.09 if it
is to match the estimate of INGSF based on the Pocock et al. (1983) study:
INGSF = BKSF AFw (Equation A-9)
A-10
-------
Values of AFW within the range 0.20 - 0.25 would correspond to a range for BKSF of 0.36 - 0.45,
or approximately 0.4 |ig/dL per |ig/day (rounded to one significant figure). A range of 0.20 - 0.25
for AFW is supported by data from numerous lead bioavailability studies (see Section 6 of the
Appendix for a more detailed discussion of these studies).
The above estimate of 0.4 |ig/dL per jig/day for the BKSF can be compared with the
approach described by Bowers et al. (1994), who used the same data set along with different
assumptions and arrived at essentially the same estimate of the BKSF, 0.375 or approximately 0.4
1-ig/dL per |ig/day. Bowers et al. (1994) assumed a daily tap water intake of 2 L/day and 8%
absorption of lead ingested in tap water; and did not make adjustments for a mixture of first draw
and flushed water intake in the Pocock et al. (1983) study.
Several uncertainties should be considered in applying the default value of 0.4 |ig/dL per
l-ig/day to any specific population. Since it is based on the Pocock et al. (1983) data, it represents
an extrapolation from adult men to women of child bearing age. Physiological changes associated
with pregnancy may affect the value of the BKSF (see Section 6 of the Appendix); therefore, some
uncertainty is associated with applying the default value to populations of pregnant women.
An additional uncertainty concerns the assumption of linearity of the relationship between
lead intake and blood lead concentration. The Pocock et al. (1983) study provides data on a large
sample population of adult men whose members were exposed to relatively low drinking water lead
levels; 898 subjects (97%) were exposed to first draw water lead concentrations less than 100 |ig/L
and 473 (52%) to 6 |ig/L or less. A smaller study of adult women exposed to higher concentrations
was reported by Sherlock et al. (1982, 1984); out of 114 subjects, 32 (28%) had flush drinking water
lead concentrations less than 100 |ig/L and only 13 (11%) less than 10 |ig/L. Sherlock et al. (1982,
1984) used a cube root regression model, rather than a linear model, to describe the relationship
between drinking water and blood lead concentration. Given the much larger sample size in the
Pocock et al. (1983) study, particularly towards the low end of the distribution for water lead
concentration, greater confidence can be placed in the estimated slope of the linear regression model
from the Pocock et al. (1983) study than in the cube root regression model of Sherlock et al. (1982,
1984). Nevertheless, it is useful to compare the output of the two models because they were applied
to the different sexes and because they differ so fundamentally in the treatment of the blood lead -
water lead slope; the slope is constant in the linear model and decreases in the cube root model as
water lead concentration increases. Figure A-l compares the output of the two models and shows
the output of a linear regression of the unweighted output of the Sherlock et al. (1984) model. Three
observations can be made from this comparison that are relevant to the BKSF:
1. Both the Pocock etal. (1983) and Sherlock etal. (1984) models predict higher blood
lead concentrations than would be expected in the average U.S. population today as
suggested from NHANES III. This is indicative of higher lead intakes in the study
populations which may have contributed to the apparent nonlinearities observed (e.g.
above 100 |ig/L in Pocock et al.(1983) and at lower concentrations in Sherlock et al.
(1984).
2. The cube root regression model of Sherlock et al. (1984) predicts lower blood lead
concentrations than the linear model of Pocock et al. (1983). This may reflect
A-ll
-------
greater lead intakes from sources other than drinking water in the Pocock et al.
(1983) population (see Section 6 of the Appendix for further discussion).
3. The linear approximation of the Sherlock et al. (1984) and the linear model from
Pocock et al. (1983) have similar slopes; 0.08 and 0.06 |ig/dL per |ig/L, respectively.
Thus, although the Sherlock et al. (1984) study casts some degree of uncertainty on
the assumption of linearity of the blood lead - drinking water lead relationship both
at low (<10 i-ig/L) and high (> 100 |ig/L) tap water lead concentrations, a linear
model with a constant slope of 0.06 |ig/dL per |ig/L appears to approximate the
output of the nonlinear model of Sherlock et al. (1984) reasonably well for water lead
concentrations less than 100 |ig/L.
A-12
-------
0
20
40 60
Pb in water (^g/L)
80
100
Figure A-l. Comparison of linear model of Pocock et al. (1983) with cube root model of Sherlock
et al. (1984) and a linear model imposed on the unweighted output of the Sherlock model over the
water lead range 0 -100 |ig/L (linear Sher84). The slope of the linear Sher84 model is 0.08 |ig/dL
per i-ig/L. The slope of the Pocock et al. (1983) model is 0.06 |ig/dL per |ig/L.
A-13
-------
Experimental data on the pharmacokinetics of lead in adult humans support the default value
of 0.4 (i-ig/dL per [ig/day absorbed lead) for BKSF estimated from Pocock et al. (1983). Several
distinct kinetic pools of lead are evident from observations of the rate of change of blood lead
isotope with time after a period of daily dosing in which lead is abruptly terminated (Rabinowitz et
al., 1976). A rapid exchange pool, denoted pool 1, includes the blood and a portion of the
extracellular fluid, and is the physiological pool from which urinary and hepatobiliary excretion of
blood lead occurs. Several estimates of the size of pool 1 (Vj) and the residence times for lead in
pool 1 (Tj) have been derived from experiments in which human subjects were administered tracer
doses of stable isotopes of lead from which pool 1 clearances (Cx) have been estimated; these
estimates are summarized in Table A-3.
Table A-3. Summary of Experimental Studies with Humans to Assess Clearance Rates of
Lead from Blood and Extracellular Fluid.
Subject
A
B
A
B
C
D
E
ACC
DN
PL
ACW
MJH
ANB
Mean ± SD
V/
(dL)
77
115
74
100
101
99
113
70e
94e
85e
94e
97e
95e
93 ±14
T~< b
(day)
34
50
34
40
37
40
27
29
39
40
48
41
40
38±6
T1/2C
(day)
24
35
24
28
26
28
19
20
27
28
33
28
28
27 ±4
Cid
(dL/day)
2.3
2.3
2.2
2.5
2.7
2.5
4.2
2.4
2.4
2.1
2.0
2.4
2.4
2.5 ±0.5
Reference
Rabinowitz et al., 1974
Rabinowitz et al., 1976
Chamberlain et al., 1978
aThe reported volume of pool 1, which refers to blood and rapidly exchangeable extracellular
fluid compartment.
b The reported residence time for lead in pool 1.
The half life of lead in pool 1; T,/2 = (Tj) x ln(2).
dClearance of lead from pool 1; Cl = V/IY
Estimated assuming Vl = Vblood x 1.7 (Rabinowitz et al., 1976).
A-14
-------
The above experiments support a value for Cj of 2.5 dL/day. At steady state, the clearance is
equivalent to the rate of uptake of lead into pool 1 per unit of blood lead concentration (|ig/day per
|ig/dL). Theoretically, this should correspond to a slope factor of 0.40 |ig/dL per [ig/day absorbed
lead (i.e., the reciprocal of the clearance estimate). Thus, the default value for the BKSF parameter
of 0.4 jig/dL per jig/day absorbed lead derived from the population survey data of Pocock et al.
(1983) is consistent with the clearance estimates from experimental studies.
6. Soil Lead Absorption Factor (AFS)
The AFS parameter is the fraction of lead in soil ingested daily that is absorbed from the
gastrointestinal tract. The TRW recommends a default value of 0. 12 based on the assumption that
the absorption factor for soluble lead (AFsoluble) is 0.2 and that the relative bioavailability of lead
in soil compared to soluble lead (RBFsoil/soluble) is 0.6:
(Equation A-10)
AFS = 0.2 0.6 = 0.12
The default value of 0.2 for AFsoluble in adults represents a weight of evidence determination based
on experimental estimates of the bioavailability of ingested lead in adult humans with consideration
of three major sources of variability that are likely to be present in populations, but are not always
represented in experimental studies; these are variability in food intake, lead intake, and lead form
and particle size.
Effect of food on lead bioavailability. The bioavailability of ingested soluble lead in adults
has been found to vary from less than 10% when ingested with a meal to 60 - 80% when ingested
after a fast (Blake, 1976; Blake etal., 1983; Blake and Mann, 1983;Grazianoetal., 1995;Heardand
Chamberlain, 1982; James etal., 1985; Rabinowitz etal., 1976, 1980). The general consensus is that
constituents of food in the gastrointestinal tract decrease absorption of ingested lead, although the
exact mechanisms by which this occurs are not entirely understood. Lead intake within a
population would be expected to occur at various times with respect to meals. Therefore, the central
tendency for lead absorption would be expected to reflect, in part, meal patterns within the
population and to have a value between the experimentally determined estimate for fasted and fed
subjects.
An estimate of a "meal -weighted" AFsoluble can be obtained from the data reported by James
et al. (1985) and certain simplifying assumptions. James et al. (1985) assessed the effects of food
on lead bioavailability by measuring the fraction retained in the whole body of adult subjects 7 days
after they ingested a dose of radioactive lead either after a fast or at various times before or after a
meal. The total lead dose was approximately 50 |ig (fasted) - 100 |ig (with food). Lead retention
was 61 ± 8.2 (SD)% when lead was ingested on the 12th hour of a 19-hour fast and decreased to 4%
- 1 6% when lead was ingested between 0 and 3 hours after a meal; retention was further reduced (3.5
± 2.9%) when lead was ingested with a meal (breakfast) (the bioavailability may have been more
than these retention estimates since some absorbed lead would have been excreted during the 7 day
A-15
-------
interval between dosing and measurement of whole-body lead). Since ingested material may be
retained in the human stomach or at least 1 hour (Hunt and Spurrel, 1951; Davenport, 1971), lead
bioavailability also may be reduced when lead is ingested 1 hour before a meal. The average "meal-
weighted" bioavailability can be estimated based on the average number of waking hours during the
day, the number of meals eaten, the bioavailability of lead ingested within 1 hour before a meal, the
bioavailability of lead ingested within 0 to 3 hours after a meal, and the bioavailability of lead at
other times during the day. For example, if it is assumed that people eat three meals each day and,
based on the James et al. (1985) study, the bioavailability of lead ingested within 1 hour before a
meal or 0 to 3 hours after a meal is approximately 0.1, and the bioavailability of lead ingested at all
other times in a 16 hour day is 0.6, then the average "meal-weighted" bioavailability during a 16
hour day is approximately 0.2:
(0.1 12 hrs) + (0.6 4 hrs) = Q 23
16 hrs
This example suggests that the use of 0.2 as a default value for AFsoluble is plausible for
populations in which soil lead intake occurs throughout the day, interspersed with meals. This may
not apply to all members of a population. For example, the average bioavailability would be higher
if less than three meals were consumed each day (e.g., using a similar calculation it can be shown
that the average bioavailability for one meal each day would be 0.5). Average bioavailability also
may be greater than 0.2 if lead intake was to occur predominantly in the early morning, before the
first meal of the day.
Although lead bioavailability may be lower in individuals whose soil lead ingestion
coincides with meals, the TRW cautions against the use of a value less than 0.2 for several reasons.
Iron and calcium deficiency associated with poor nutritional status may enhance absorption (U.S.
EPA, 1990). In addition, numerous factors may affect the absorption, distribution, excretion, and
mobilization of lead during pregnancy: increased plasma volume (i.e., hemodilution); decreased
hematocrit; previous exposure history of the mother (i.e., bone lead sequestration); changes in
nutritional status; significant loss of body weight or depletion of fat stores; hormonal modulation;
age; race; administration of drugs; and illness (Silbergeld, 1991). There is likely to be significant
inter-individual variability in these factors, and studies of women at different stages of pregnancy
have not shown clear trends in effects on blood lead concentration (Gershanik et al., 1974;
Alexander and Delves, 1981; Baghurst et al., 1987; Silbergeld, 1991). While there is evidence to
support 0.2 as a reasonable estimate of AFsoluble for women of child-bearing age, there is still some
basis for concern regarding potentially elevated absorption during pregnancy. However, a potential
increase in lead absorption during pregnancy would be expected to occur dynamically with changes
in bone mobilization, blood volume and glomerular filtration rate. Thus, the TRW cautions against
adjusting the value for AFsoluble (or BKSF) based on assumptions regarding the effects of pregnancy
on blood lead concentration.
Nonlinearity in blood lead concentration. Another reason for caution in adopting values
for AFsoluble less than 0.2 derives from uncertainty about the relationship between blood lead
concentration, lead intake, and lead absorption. Several studies have shown that the relationship
between environmental lead levels (e.g., drinking water lead concentration) and blood lead
A-16
-------
concentration is nonlinear and suggest the possibility that fractional absorption of ingested lead is
dose-dependent, and decreases as lead intake (and blood lead concentration) increases. Pocock et
al. (1983) reported a nonlinear relationship between blood lead concentration and water lead that
could be approximated by two linear equations: a slope of 0.06 |ig/dL per |ig/L was estimated for
water lead concentrations equal to or less than 100 |ig/L and a slope of 0.01 was estimated for water
lead concentrations above 100|ig/L. Sherlock etal. (1982,1984) used a cube root regression model
to relate blood and water lead concentrations; however, over the range of water lead concentrations
of 100 i-ig/L or less, the slope of 0.06 |ig/dL per |ig/L water lead from Pocock et al. (1983)
approximates the relationship observed in the Sherlock et al. (1982, 1984) study (Figure A-l). The
linear relationship between water lead and blood lead in the Pocock et al. (1983) study extends from
a blood lead concentration range of 14 to 20 |ig/dL. Based on these data, the value of AFsoluble of 0.2
may be considered a reasonable default estimate if applied to exposure scenarios in which the
estimates of blood lead concentration do not exceed 20 |ig/dL. At blood lead concentrations greater
than this, absorption of soluble lead may be less than the default value.
An appropriate value of AFsoluble also can be supported by estimating the range of daily lead
intake that is likely to result in a linear relationship between intake and blood lead concentration.
Data represented in Figure A-l suggest that if water lead concentrations are less than 100 |ig/L, the
blood lead - water lead relationship is approximately linear. If assumptions regarding the magnitude
of first draw and flushed water intakes and lead concentrations are applied (see Equations A-8 and
A-9 and discussion of BKSF), a first draw water lead concentration of 100 |ig/L in the Pocock et al.
(1983) study represents a water lead intake of approximately 70 jig/day:
100-1.4-(0.3 + (0.25-0.7)) « 70
We do not know with certainty the total lead intake in the Pocock et al. (1983) population,
although we can be certain that it exceeded the above estimated intake from drinking water since
intake from diet and other sources, including occupational, would have occurred; this is consistent
with the higher blood lead concentrations that were observed in the male population. Sherlock et
al. (1982) estimated that, in their study population of adult women, the dietary contribution to total
lead intake was equal to that from drinking water when the water lead concentration was 100 |ig/L,
and that the contribution of lead from sources other than diet and water was very small. If the same
assumption is applied to the Pocock et al. (1983) study, it is likely that total lead intake in the male
population was at least 140 [ig/day (70 [ig/day from drinking water and 70 [ig/day from diet; the
Pocock et al., 1983 study included 40 households from the Sherlock et al., 1982 study site), and may
have been higher because of occupational exposure in the male population. A crude estimate of the
relative magnitudes of the non-water lead intakes in the two studies can be obtained by comparing
the predicted water lead concentration required to achieve the same blood lead concentration in the
two populations. For example, a water lead concentration of 100 |ig/L corresponded to a predicted
blood lead concentration of approximately 18 |ig/dL in the female population (Sherlock etal., 1984);
the same blood lead concentration corresponded to a water lead concentration of 50 |ig/L in the
male population (Pocock et al., 1983). Therefore, the non-water lead intakes in the male population
may have been twice that in the female population. If it is assumed that drinking water and diet
contributed equally to lead intake in both studies, then a drinking water lead concentration of 100
|ig/L in the Pocock et al. (1983) study translates to a total lead intake of approximately 300 jig/day:
A-17
-------
1 total = Avoter + Idiet + 1'other (Equation A-11)
Itotal = 70 + 70 + 140 * 300 \iglday
Thus, the departure from linearity observed in the Pocock et al. (1983) study may have occurred at
lead intakes at or above 300 jig/day. In the various experimental assessments of lead bioavailability,
subjects ingested lead in amounts that varied among the studies but were all within the range 100 -
300 jig (Blake, 1976; Blake et al., 1983; Blake and Mann, 1983; Graziano et al., 1995; Heard and
Chamberlain, 1982; James et al., 1985; Rabinowitz et al., 1976, 1980), which is within the
approximate linear range, if the extrapolation from the Pocock et al. (1983) and Sherlock et al.
(1982) studies is reasonable. Based on these considerations, the value of AFsoluble of 0.2 is
considered to be a reasonable default value if applied to exposure scenarios in which lead intakes
are less than 300 [ig/day. At intakes greater than this, absorption of soluble lead may be less than
the default value; however, it can be similarly argued that, based on the Sherlock et al. (1984)
regression model, the default AFsoluble may underestimate absorption by some degree at low
exposures.
Effect of lead form and particle size on lead bioavailability. The default value of 0.2 for
AFsoluble applies to soluble forms of lead in drinking water and food and would be expected to
overestimate absorption of less soluble forms of lead in soil. Experimental studies have shown that
the bioavailability of lead in soil tends to be less than that of soluble lead. Weis et al. (1994)
assessed the relative bioavailability of lead in soil compared to water soluble lead (acetate) in
immature swine and estimated that the relative bioavailability of lead in soil from Leadville, CO was
0.6 to 0.8. Ruby et al. (1996) reported estimates of the relative bioavailability of lead in a variety
of soils from mining sites and smelters as assessed in the Sprague-Dawley rat; the estimates ranged
from 0.09 to 0.4. Maddaloni et al. (1996) reported preliminary data from a study in which 6 fasted
human subjects were administered a single dose of lead-contaminated soil. The dose was 250 |ig
lead normalized to a 70 kg body weight; the concentration of lead in the soil was 2850 |ig/g and the
amount of soil administered to each subject was generally a little less than 100 mg. The average
estimate of lead absorption in the six subjects was 26%. If the absorption factor for soluble lead in
fasted adults is assumed to be 0.6 (James et al., 1985), then the Maddaloni et al. (1996) estimate
suggests a relative bioavailability of 0.5 (i.e., 0.3/0.6) for lead in soil.
Based on the above evidence, the TRW considers 0.6 to be a plausible default point estimate
for the relative bioavailability of lead in soil compared to soluble lead (RBFsoil/soluble) when site-
specific data are not available. Such data are highly desirable as variation in relative bioavailability
is expected for different species of lead and different particle sizes (Barltrop and Meek, 1975,1979),
both of which may vary from site to site. For example, the bioavailability of metallic lead has been
shown to decrease with increasing particle size (Barltrop and Meek, 1979), therefore, the default
value forRBFsoil/soluble may overestimate absorption of lead if applied to soils contaminated with large
lead particles such as firing range debris or mine tailings. Here again, the TRW cautions against the
use of a lower value for the RBFsoil/soluble, unless it can be supported by experimental assessments
of relative bioavailability.
A-18
-------
The default value of 0.6 for RBFsoil/soluble, coupled with the default value of 0.2 for AFsoluble,
yields a default value of 0.12 for AFS (0.6 0.2). The TRW considers 0.12 to be a plausible point
estimate for the absorbed fraction of ingested soil lead for use in assessments in which site-specific
data on lead bioavailability are not available. The default value of 0.12 takes into account
uncertainties regarding the possible nonlinearity in the relationship between lead intake and
absorption and should be adequately protective in scenarios in which predicted blood lead
concentrations are less than 20 |ig/dL. The use of the default value for populations that have
substantially higher blood lead concentrations may result in an overestimate of lead uptake, and
conversely, lead uptake may be underestimated at lower exposures.
7. Daily Soil Ingestion Rate (IRS)
The TRW recommends a default value of 0.05 g/day as a plausible point estimate of the
central tendency for daily soil intake from all occupational sources, including soil in indoor dust,
resulting from non-contact intensive activities. This would include exposures that are predominantly
indoors. More intensive soil contact would be expected for predominantly outdoor activities such
as construction, excavation, yard work, and gardening (Hawley, 1985). Site-specific data on soil
contact intensity, including potential seasonal variations, should be considered in evaluating whether
or not the default value is applicable to the population of concern and, if not, activity-weighted
estimates of IRS that more accurately reflect the site can be developed.
In adopting the single IRS parameter to describe all sources of ingested soil, the methodology
remains consistent with recommendations of the Superfund program and their implementation for
risk assessment; specifically, the 0.05 g/day value used for adult soil ingestion addresses all
occupational soil intake by the individual, whether directly from soil or indirectly through contact
with dust (U.S. EPA, 1993). This value specifically applies to the assessment of soil lead risk, and
not risks associated with non-soil sources of lead in dust. In making soil ingestion exposure
estimates under the Risk Assessment Guidelines for Superfund (RAGS) framework, no specific
assumptions are needed about the fraction of soil intake that occurs through dust.
An alternative approach was needed in the IEUBK Model because childhood lead exposures
are often strongly influenced by indoor sources of lead in dust (e.g., indoor paint) (U.S. EPA,
1994b). In a situation where indoor sources of dust contamination are important, an exposure
estimate that addresses only soil exposures (including the soil component of dust) would be
incomplete. The IEUBK Model assigns separate values to outdoor soil and total indoor dust
ingestion and partitions the indoor dust into soil-derived and non-soil-derived sources. At a
minimum, paired soil and indoor dust samples should be collected to adequately characterize
exposure to lead where indoor sources of dust lead may be significant.
Alternate method for calculating soil and dust ingestion as separate exposure pathways.
In this alternate approach, separate estimates are made of lead intake from the direct ingestion of
outdoor soil and from the ingestion of indoor dust (which may contain lead from soil and as well as
from indoor sources such as deteriorated lead based paint). Exposure to lead from soil (outdoor
A-19
-------
contact) can be calculated using Equation A-12, while exposure to lead from indoor dust can be
calculated using Equation A-13.
INTAKE,
PbS
S,outdoors
EF
Site
AT
(Equation A-12)
INTAKED.ndoors
AT
(Equation A-13)
INTAKE
S, outdoors
INTAKE
D, indoors
PbS
PbD
TR
1JX
S, outdoors
, indoors
EFSlte
AT
Daily average intake (ingestion) of lead from soil ingested
outdoors (|ag/day).
Daily average intake (ingestion) of lead from dust ingested indoors
Soil lead concentration (i-ig/g) (average concentration in assessed
individual exposure area).
Indoor dust lead concentration (i-ig/g).
Intake rate (ingestion) of outdoor soil (g/day).
Intake rate (ingestion) of indoor dust (g/day).
Exposure frequency at site (days of exposure during the averaging
period); may be taken as days per year for continuing, long term
exposures.
Averaging time, the total period during which the assessed
exposures (from all sources) occur (days). May be taken as 365
days per year for continuing, long term exposures.
Note that, in Equations A-12 and A-13, exposure frequency refers to the number of days that an
individual is present at the site and does not partition between periods of indoor and outdoor
exposures. The intake rate is a long term average value appropriate for that media and is influenced
by both the duration of outdoor (or indoor) exposures and the intensity of those exposures.
Calculation of IRS outdoors and IRD ^00^ from total intake of soil and dust (IRS+D).
Intermediary calculations may be needed to generate estimates of the parameters in the intake
equations. An estimate of the total intake of soil and dust materials (IRS+D) serves as a starting point.
Note that IRS+D differs from IRS which was discussed above, because IRS+D includes not only the
total mass of soil ingested (both directly and as a component of indoor dust), but also the ingested
A-20
-------
mass of non-soil derived dust components including various materials of indoor origin. Since a
substantial fraction of the mass of indoor dust comes from sources other than outdoor soils, an
estimate of IRS+D will be higher than the corresponding estimate of IRS Secondly, an estimate of
the fraction the total soil and dust intake that is ingested directly as soil is needed (Weightingsoil).
This estimate needs to take into account the intensity and duration of the outdoor soil intake and the
indoor dust intake. Equations A-14 and A-15 can be used to derive media-specific ingestion rates
from IRS+D and Weightingsoil.
IRs,outdoorS = Weightmgsoil-IRs+D (Equation A-14)
IRD,indoors = d ~ Weighting-soil) IRS+ D (Equation A-15)
Weightingsoil = Fraction of total soil and dust intake that is directly ingested as soil
(dimensionless).
IRS+D = Total daily average intake of outdoor soil and indoor dust (all dust
components) (g/day).
Data are needed to generate separate estimates of the concentrations of lead in outdoor soil and in
indoor dust. A site assessment using this alternate methodology would generally be based on direct
measurement data for both soil and dust at the facilities of concern. For comparison with exposure
estimates based on total soil ingestion (the primary approach presented in this paper), Equation A-16
may be utilized to estimate the ratio of dust lead concentration to soil lead concentration.
PbD = PbS-KSD (Equation A-16)
KSD = Ratio of indoor dust lead concentration to soil lead concentration (dimensionless).
Assuming that the same absorption fraction is applicable to both soil and dust, Equation A-17 may
be used to estimate the uptake of lead from these two sources.
UPTAKE = AFS>D (INTAKEs>outdoors + INTAKED>indoors) (Equation A-17)
UPTAKE = Daily average uptake of lead from the gastrotintestinal tract into the systemic
circulation; soil and dust sources (|ig/day).
A-21
-------
AFS D = Absolute gastrointestinal absorption fraction for ingested lead in soil and dust
(dimensionless).
Comparison of lead intake estimated from principal and alternate approaches. It is
helpful to compare exposure estimates derived using our principal approach based on total soil
intake (including soil present in ingested dust) with the results of the disaggregated pathway analysis
for soil and dust. We will consider the case in which there are not important indoor sources of lead
in dust. We can then compare the total lead intake estimates from the two approaches.
Under the model based on total soil ingestion (which we re-label as IRSi,otal for clarity):
PbS-IR~
INTAKE = - - (Equation A- 18)
By contrast, using the disaggregated soil and dust model, Equations A-14, A-15, A-16, and A-18
may be combined to give Equation A-19:
mT,rF PbS-IRs+D-(Weightingsoil+KSD-(1- WeightingsJ)-EFSite
INTAKE = (Equation A-19)
AL
When applied to the same exposure assessment problem, the two approaches should give equivalent
estimates of lead intake. The estimates will be equivalent when:
IRS+D - (Weightingsoil + KSD (1 - WeightingSJ) = IRs>total
8. Exposure Frequency (EFS)
The TRW recommends a default value of 219 days/year. This is the same as the central
tendency occupational exposure frequency recommended by U. S. EPA (1993) Superfund guidance,
which is based on 1991 data from the Bureau of Labor Statistics. This estimate corresponds to the
average time spent at work by both full-time and part-time workers engaged in non-contact
intensive activities (U.S. EPA, 1993). Site-specific data on exposure frequency should be
considered in evaluating whether or not the default value is applicable to the population of concern.
In evaluating site-specific data, it should be kept in mind that exposure frequency and daily soil
ingestion rate (IRS) may be interdependent variables, particularly in contact-intensive scenarios;
therefore, the assignment of a site-specific value to EFS should prompt an evaluation of the
applicability of the default value for IRS to the population of concern (see Section 7 of the Appendix
for further discussion).
A-22
-------
Nonresidential exposure scenarios in which exposure frequency would be substantially less
than 219 days/year are frequently encountered. Examples include trespassing and recreational use
of a site. Important methodology constraints on exposure frequency and duration must be
considered in assigning values to EFS that would represent infrequent contact with the site; these
constraints relate to the steady state assumptions that underlie the BKSF. The BKSF derived from
the Pocock et al. (1983) data applies to exposures that result in a quasi-steady state for blood lead
concentration; that is, an intake over a sufficient duration for the blood lead concentration to become
nearly constant over time. Based on estimates of the first order elimination half-time for lead in
blood of approximately 30 daysfor adults (Rabinowitz, et al., 1974,1976; Chamberlain et al., 1978),
a constant lead intake rate over a duration of 90 days would be expected to achieve a blood lead
concentration that is sufficiently close the quasi-steady state. This is the minimum exposure
duration to which this methodology should be applied.
Infrequent exposures (i.e., less than 1 day per week) over a minimum duration of 90 days
would be expected to produce oscillations in blood lead concentrations associated with the
absorption and subsequent clearance of lead from the blood between each exposure event. Based
on the above assumptions about the elimination half-time lead in blood, the TRW recommends that
this methodology should not be applied to scenarios in which EFS is less than 1 day/week.
9. Applying Monte Carlo Analysis to the Adult Lead Methodology
Recent EPA guidance (Browner, 1995) recommends that risk assessments include a clear and
transparent discussion of variability and uncertainty. The lead risk assessment methodology
presented here develops explicit estimates of the variability of blood lead levels among adults who
are exposed to specified concentrations of environmental lead. This analysis relies on data from a
large number of studies (baseline blood lead levels, variability of blood lead levels, contact rates
with environmental media, lead bioavailability, and lead biokinetics) to support a predictive
probabilistic (lognormal) model for adult and fetal blood lead concentrations. Important issues
regarding the uncertainty in parameter inputs and the mathematical form of the model are discussed
in the sections of this Appendix. The TRW recognizes that there is considerable scientific interest
in the different analytical approaches that may be applied to aid in the analysis of variability and
uncertainty in risk assessments. In particular, under appropriate circumstances, Monte Carlo
methods may provide a useful approach for developing quantitative estimates of the variability,
uncertainty (or both) in risk predictions.
The TRW chose not to pursue application of Monte Carlo or other stochastic simulation
methods in this effort addressing adult lead risk assessment. Several factors went into this decision.
First, the TRW understood the needs of EPA Regions for a risk model that could be developed
relatively rapidly and which Regional lead risk assessors could apply easily with limited need for
additional study or training. These considerations made it advantageous to focus on models that are
conceptually similar to the IEUBK model for children in terms of applying a parametric lognormal
modeling approach to address distributions for blood lead levels. Secondly, the TRW recognized
that there would be substantial scientific issues associated with developing widely applicable
stochastic simulation models for adult lead risk assessment. These difficulties primarily relate to
the absence of reliable distributional data for a variety of important variables in the assessment. As
A-23
-------
one example, very limited data are available on soil ingestion rates in adults and a distributional
choice for this key parameter would depend heavily on individual judgement with little Agency
precedent for support. Additionally, in a stochastic assessment, a greater complexity would arise
due to likely correlations among the variables in the adult lead risk assessment. Stochastic analyses
need to explicitly account for important correlations among variables if the simulations are to
provide realistic distributions of risk. As an example, dependence is likely to exist between the
starting (non-site related) blood lead concentrations for individuals and their site-related increases
in blood lead. This dependence may result from individual patterns of behavior and from biological
factors associated with lead pharmacokinetics. However, data on this dependence are sparse or
absent, and the necessary statistical estimates of the correlation strength would depend heavily on
personal judgement.
The TRW does encourage further efforts to better define the distributional data on which
stochastic simulations of lead risks might rest. Further attention to these data can provide useful
insights for lead risk assessment. The TRW also recognizes that Regions may be presented with
lead risk assessments based on Monte Carlo modeling. In order to facilitate review of Monte Carlo
analyses, some EPA Regions have found it important to establish requirements for the orderly
development and review of these assessments. Borrowing on this approach, the TRW recommends
that:
A plan for the use of Monte Carlo analysis in a lead risk assessment should be submitted
to responsible Regional personnel and accepted by them before the Monte Carlo analysis
is undertaken.
In general, it is expected that site-specific exposure related parameters that are supported
with site-specific information will provide the basis for proposed Monte Carlo
simulations.
Scientific review is needed to determine that the risk assessment conformed to the plan
and to evaluate the reliability of the results.
These recommendations are designed to ensure that assessments can provide meaningful results that
can be understood and evaluated. If analyses are submitted in a format that is difficult to understand,
the utility of the analysis will be diminished. We recommend that Regional staff seek advice from
the TRW as a resource in this process.
A-24
-------
10. References
Alexander, F.W. and H.T. Delves. 1981. Blood lead levels during pregnancy. Int. Arch. Occup.
Environ. Health. 48: 35-39.
Baghurst, P.A., AJ. McMichael, G.V. Vimpani, E.F. Robertson, P.O. Clark, andN.R. Wigg. 1987.
Determinants of blood lead concentrations of pregnant women living in Port Pirie and surrounding
areas. Medical J. of Australia. 146: 69-73.
Balbus-Kornfeld, J. 1994. Comments and Recommendations on the Draft Interim Guidance for
Screening Levels of Lead in Soil for Non-Residential Sites. Letter from John Balbus-Kornfeld to
Bruce Means. November 17, 1994.
Barltrop, D. and F. Meek. 1975. Absorption of different lead compounds. Postgrad. Med. J. 51:
805-809.
Barltrop, D. and F. Meek. 1979. Effect of particle size on lead absorption from the gut. Arch.
Environ. Health. 34: 280-285.
Blake, K.C.H. 1976. Absorption of 203Pb from gastrointestinal tract of man. Environ. Res. 11: 1-4.
Blake, K.C.H. and M. Mann. 1983. Effect of calcium and phosphorus on the gastrointestinal
absorption of 203Pb in man. Environ. Res. 30: 188-194.
Blake, K.C.H., G.O. Barbezat and M. Mann. 1983. Effect of dietary constituents on the
gastrointestinal absorption of 203Pb in man. Environ. Res. 30: 182-187.
Bowers, T.S., B.D. Beck and H.S. Karam. 1994. Assessing the relationship between environmental
lead concentrations and adult blood lead levels. Risk Analysis. 14(2): 183-189.
Brody, DJ. Personal communication on October 24, 1996 and October 29, 1996.
Brody, D.J., J.L. Pirkle, R.A. Kramer, K.M. Flegal, T.D. Matte, E.W. Gunter and D.C. Paschal.
1994. Blood lead levels in the U.S. population. Phase 1 of the third National Health and Nutrition
Examination Survey (NHANES III, 1988 to 1991). JAMA. 272(4): 277-283.
Browner, C.M. 1995. Policy for Risk Characterization at the U.S. EPA. Memorandum from U.S.
EPA Administrator dated March 21, 1995.
Chamberlain, A.C., M.J. Heard, P. Little, D. Newton, A.C. Wells and R.D. Wiffen. 1978.
Investigations into lead from motor vehicles. Harwell, United Kingdom: United Kingdom Atomic
Energy Authority, Report No. AERE-R9198.
Davenport, H.W. 1971. Gastric digestion and emptying; absorption. In: Physiology of the
Digestive Tract, 3rd ed. Year Book Medical Publishers Inc., Chicago, pp. 165-168.
A-25
-------
Franklin, C.A., MJ. Inskip, C.L. Baccanale, EJ. O'Flaherty, W.I. Manton, D.L. Schanzer, J.
Blenkinsop and C.M. Edwards. 1995. Transplacental transfer of lead in non-human primates
(Macacafascicularis): use of serially administered stable isotope tracers of lead to elicit contribution
of maternal bone lead to blood lead and the fetus. Poster presented at the 1995 meeting of the
Society of Toxicology, Baltimore, MD. The Toxicologist. 15:194.
Gershanik, J. J., G.G. Brooks, and J. A. Little. 1974. Blood lead values in pregnant women and their
offspring. Amer. J. Obstet. Gynecol. 4: 508-511.
Goyer, R.A. 1990. Transplacental transport of lead. Environ. Health Perspect. 89: 101-105.
Graziano, J.H., D. Popovac, P. Factor-Litvak, P. Shrout, J. Kline, MJ. Murphy, Y. Zhao, A.
Mehmeti, X. Ahmedi, B. Rajovic, Z. Zvicer, D. Nenezic, N. Lolacono and Z. Stein. 1990.
Determinants of elevated blood lead during pregnancy in a population surrounding a lead smelter
in Kosovo, Yugoslavia. Environ. Health Perspect. 89:95-100.
Graziano, J.H., W.I. Manton, C.B Blum and N. J. Lolacono. 1995. Bioavailability of lead in wine,
by stable isotope dilution. Poster presented at the 1995 meeting of the Society of Toxicology,
Baltimore, MD. The Toxicologist. 15: 135 (abst).
Hawley, J.D. 1985. Assessment of health risk from exposure to contaminated soil. Risk Analysis.
5: 289-302.
Heard, MJ. and A.C. Chamberlain. 1982. Effect of minerals and food on uptake of lead from the
gastrointestinal tract in humans. Human Toxicol. 1: 411-415.
Hunt, J.N. and W.R. Spurrell. 1951. The pattern of emptying of the human stomach. J. Physiol.
113: 157-168.
James, H.M., M.E. Milburn and J. A. Blair. 1985. Effects of meals and meal times on uptake of lead
from the gastrointestinal tract of humans. Human Toxicol. 4: 401-407.
Maddaloni, M., W. Manton, C. Blum, N. Lolacono and J. Graziano. 1996. Bioavailability of soil-
borne lead in adults, by stable isotope dilution. The Toxicologist. 30: 15 (abst.)
NRC. 1993. Measuring Lead Exposure in Infants, Children and Other Sensitive Populations.
National Academy Press. Washington, DC. ISBN 0-309-04927-X.
Pocock, S J., A.G. Shaper, M. Walker, C J. Wale, B. Clayton, T. Delves, R.F. Lacey, R.F. Packham
and P. Powell. 1983. Effects of tap water lead, water hardness, alcohol, and cigarettes on blood lead
concentrations. J. Epidemiol. Commun. Health. 37: 1-7.
Rabinowitz, M.B., G.W. Wetherill and J.D. Koppel. 1974. Studies of human lead metabolism by
use of stable isotope tracers. Environ. Health Perspect. 7: 145-153.
A-26
-------
Rabinowitz, M.B., G.W. Wetherill and J.D. Koppel. 1976. Kinetic analysis of lead metabolism in
health humans. J. Clin. Invest. 58: 260-270.
Rabinowitz, M.B., J.D. Koppel and G.W. Wetherill. 1980. Effect of food intake on fasting
gastrointestinal lead absorption in humans. Am. J. Clin. Nutr. 33: 1784-1788.
Ruby, M.V., A . Davis, R. Schoof, S. Eberle and C. M. Sellstone. 1996. Estimation of lead and
arsenic bioavailability using a physiologically based extract!on test. Environ. Sci. Technol. 30:422-
430.
Sherlock, J., G. Smart, G.I. Forbes, M.R. Moore, WJ. Patterson, W.N. Richards and T.S. Wilson.
1982. Assessment of lead intakes and dose-response for a population in Ayr exposed to a
plumbosolvent water supply. Human Toxicol. 1: 115-122.
Sherlock, J.C., D. Ashby, H.T. Delves, G.I. Forbes, M.R. Moore, WJ. Patterson, S.J. Pocock, MJ.
Quinn, W.N. Richards and T.S. Wilson. 1984. Reduction in exposure to lead from drinking water
and its effect on blood lead concentrations. Human Toxicol. 3: 383-392.
Silbergeld, E.K. 1991. Lead in bone: Implications for toxicology during pregnancy and lactation.
Environ. Health Perspect. 91: 63-70.
U.S. EPA. 1986. Air Quality Criteria for Lead Volumes I - IV. Environmental Criteria and
Assessment Office, Office of Research and Development, RTF, NC. EPA 600/8-83-028 a-d.
U.S. EPA. 1989. Exposure Factors Handbook. Office of Health and Environmental Assessment,
Washington, DC. EPA/600/8-89/043.
U.S. EPA. 1990. Supplement to the 1986 EPA Air Quality Criteria Document for Lead - Volume
1 Addendum. Office of Research and Development, Office of Health and Environmental
Assessment, Washington, DC. EPA-600/8-89/049A.
U.S. EPA. 1992. A TRW Report: Review of the EPA Uptake Biokinetic Model for Lead at the
Butte NPL Site. Technical Review Workgroup for Lead, October, 1992.
U.S. EPA. 1993. Superfund's Standard Default Exposure Factors for the Central Tendency and
RME-Draft. Working Draft, November 1993.
U.S. EPA. 1994a. Revised Interim Soil Lead Guidance for CERCLA Sites and RCRA Corrective
Action Facilities. OSWERDirectiveNo. 9355.4-12. Office of Emergency and Remedial Response,
Washington, D.C. EPA/540/F-94/043, PB94-963282.
U.S. EPA. 1994b. Technical Support Docuement: Parameters and Equations Used in the Inegrated
Exposure Uptake Biokinetic Model for Lead in Children (v. 0.99d). Office of Emergency and
Remedial Response, Washington, D.C. EPA/540/R-94/040, PB94-963505.
A-27
-------
U.S. EPA. 1994c. Guidance Manual for the Integrated Exposure Uptake Biokinetic Model for Lead
in Children. Office of Emergency and Remedial Response, Washington, D.C. EPA/540/R-93/081,
PB93-963510.
U.S. EPA. 1995. A TRW Report: Review of a Methodology for Establishing Risk-Based Soil
Remediation Goals for the Commercial Areas of the California Gulch Site. Technical Review
Workgroup for Lead, October, 1995.
Weis, C.P., G.M. Henningsen, R.L. Poppenga, BJ. Thacker, A. Curtis, R. Jolly and T. Harpstead.
1994. Use of an immature swine model to sensitively differentiate lead absorption from soluble and
mineralogical matrices. Presented at the Society for Environmental Geochemistry and Health, Salt
Lake City, UT, July 18-19, 1994.
A-28
-------
APPENDIX B
Calculations of Preliminary Remediation Goals (PRGs)
-------
-------
Calculations of Preliminary Remediation Goals (PRGs)
U.S. EPA Technical Review Workgroup for Lead, Adult Lead Committee
Version date 8/14/01
Exposure
Variable*
PbBfetill 0.95
Rfetal/maternal
BKSF
GSD;
PbB0
IRS
IRS+D
ws
KSD
AFS,D
EFS,D
Als,D
PRG
^^HiG' ;';v'
', Equation1
- 1*<
X
X
X
X
X
X
X
X
X
2**
X
X
X
X
X
X
X
X
X
X
X
Description of Exposure Variable
95th percentile PbB in fetus
Fetal/maternal PbB ratio
Biokinetic Slope Factor
Geometric standard deviation PbB
Baseline PbB
Soil ingestion rate (including soil-derived indoor dust)
Total ingestion rate of outdoor soil and indoor dust
Weighting factor; fraction of IRS+D ingested as outdoor soil
Mass fraction of soil in dust
Absorption fraction (same for soil and dust)
Exposure frequency (same for soil and dust)
Averaging time (same for soil and dust)
Preliminary Remediation Goal
'Unit*
ug/dL
--
ug/dL per
ug/day
--
ug/dL
g/day
g/day
--
--
--
days/yr
days/yr
ppm
Values for Non-Residential Exposure Scenario
Using Equation 1
GSD1 = 1J
10
0.9
0.4
1.9
1.4
0.050
--
--
--
0.12
219
365
1,712
GSBi=2J,:
10
0.9
0.4
2.3
1.8
0.050
--
--
--
0.12
219
365
, 710 -
Using Equation 2
GSDi = lJ
10
0.9
0.4
1.9
1.4
--
0.050
1.0
0.7
0.12
219
365
1,712
GSDT-2.3' '
10
0.9
0.4
2.3
1.8
--
0.050
1.0
0.7
0.12
219
365
711)
1 Equation 1 does not apportion exposure between soil and dust ingestion (excludes Ws, KSD). When IRS = IRS+D and Ws = 1.0, the equations yield the same PRG.
'Equation 1, based on Eq. 4 in U.S. EPA (1996)
([PbB95 fetal/(R*(GSDi1 645)])-PbB0)*ATSD
BKSF*(IRS+D*AFS-D*EFS-D)
"Equation 2, alternate approach based on Eq. 4 and Eq. A-19 in U.S. EPA (1996)
([PbBfet
BKSF*([(IRS+D)*AF
il,0.95/(R
*£F *
"(GSD;1
WS]+[K
645)])-PbB0)*AT
SD*(IRS+D)*(1-W
S,D
S
)*
AFD*EFD])
Source: U.S. EPA (1996). Recommendations of the Technical Review Workgroup for Lead for an Interim Approach to Assessing Risks Associated with Adult Exposures to Lead in Soil
B-l
-------
-------
APPENDIX C
Memorandum April 7,1999
-------
-------
MEMORANDUM
DATE: April 7, 1999
SUBJECT: Use of the TRW Interim Adult Lead Methodology in Risk Assessment
TO: Mark Maddaloni, Chair
TRW Adult Lead Subgroup
FROM: Pat Van Leeuwen
Region 5 Superfund Program
Paul White
ORD/NCEA
The December 1996 TRW report "Recommendations of the Technical Review
Workgroup for Lead for an Interim Approach to Assessing Risks Associated with Adult
Exposures to Lead in Soil" presents tools that can be used in a risk assessment to provide an
evaluation of risk relevant to the adult population. This report, referred to as the interim Adult
Lead Methodology (ALM), focuses on the estimation of the blood lead concentrations in fetuses
carried by women exposed to lead contaminated soils. However, the presentation in that
document emphasizes the calculation of cleanup goals for soil and it has become apparent that
there is some confusion among risk assessors regarding how to apply this methodology in a
"forward" manner to predict baseline risks resulting from measured soil concentrations. This
memorandum presents equations for calculation of fetal risks from adult exposures to specified
levels of soil lead contamination. This approach will support EPA's goal of limiting the risk of
elevated fetal blood lead concentrations due to lead exposures to women of child-bearing age.
We have prepared this memorandum so that it may be included as an Appendix to the interim
ALM to provide the needed clarification to risk assessors and others who use the ALM to
support an evaluation of risk.
The risk assessment methodology in the ALM is based on a lognormal probability model
for blood levels in adult women exposed to lead contaminated soils, coupled with an estimated
constant of proportionality between fetal and maternal blood lead levels. These relationships
specify that the distribution of fetal blood lead levels also follows a lognormal distribution:
PbBfetal ~ Lognormal(GM,GSD)
C-l
-------
Estimation of the probability that fetal blood lead levels will exceed the EPA blood lead level of
concern of 10 ug/dL is a two step process:
(1) Calculate the geometric mean (central) fetal blood lead concentration (PbBfetalGM).
Equation A-3 in the ALM Appendix provides an estimate of the central tendency adult blood
lead level which is used to provide a plausible estimate of the geometric mean in the lognormal
model for blood lead. When the expressions for lead UPTAKE (ALM Equations A-l and A-2)
are substituted into ALM Equation A-3 the following relationship is obtained:
PbS-BKSF-IR.-AF.-EF,
PbBadult
-------
IRS = Intake rate of soil, including both outdoor soil and the soil-derived
component of indoor dust (g/day).
AFS = Absolute gastrointestinal absorption fraction for ingested lead in soil and
lead in dust derived from soil (dimensionless).
EFS = Exposure frequency for contact with assessed soils and/or dust derived in
part from these soils (days of exposure during the averaging period); may
be taken as days per year for continuing, long term exposures.
AT = Averaging time; the total period during which soil contact may occur; 365
days/year for continuing long term exposures.
(2) Determine the probability that the blood lead level for a fetus carried by a woman
exposed to lead at a site exceeds 10 ug/dL. This calculation uses the fetal geometric mean (GM)
blood lead from Equation 1 and the geometric standard deviation (GSD) value appropriate for
the risk assessment. Note that because of the assumption of proportionality between fetal and
maternal blood lead levels, the adult GSD and the fetal GSD are equal. If the assessor is using a
spreadsheet or statistical program that provides a function to calculate lognormal probabilities,
the GM and GSD values may directly used to calculate the exceedence probabilities. (Care
must be taken to determine the exact form of the inputs needed by the statistical function, e. g.,
whether log scale inputs are required.) Alternatively, the following formula and table provide
the needed tools for the probability calculation.
Recall that the logarithm of a lognormal variable follows a normal probability
distribution. Exceedence probabilities for the lognormal model can be determined from standard
normal model statistical tables after the GM, GSD, and exceedence criterion are converted to log
scale values and a "standard normal deviate" or "z-value" is calculated:
ln(10) - ln(GA<)
Z= (Equation 2)
In this equation, ln( ) represents the natural logarithm function (log base e) which is
applied in the definition of the lognormal distribution. Note, however, that calculations using
base 10 logarithms would also yield the same numerical result.
A statistical program or a normal probability table can then be used to determine the
exceedence probability. The attached standard normal table displays both positive and negative
values of z for ease of reference. The table gives the probability, p, that a standard normal
variable has a value less than z. The probability that the fetal blood lead level exceeds 10 ug/dL
is obtained by from the expression 1-p.
EXAMPLE:
Assume that the risk calculation (Equation 1) gives a GM fetal blood lead level of 7.0
ug/dL, and the appropriate GSD is 1.8.
C-2
-------
Then:
, - ln(7.0) = 2.303 - 1.946 =
ln(1.8) 0.588
Under the normal distribution, the probability that z is less than 0.607 is p = 0.728 (obtained
from a statistical program). From the attached normal table the probability may be adequately
approximated by rounding 0.607 to 0.61 to get a probability of 0.729, or approximately 0.73.1
The probability that the fetal blood lead level exceeds 10 ug/dL is estimated as 1 - p = 1 - 0.73
0.27, or approximately 27%.
JThe precision of values obtained from the table can be increased, when necessary, by using linear
interpolation between the table entries. For this example, interpolation using the values of z = 0.60 and z = 0.61 can
be applied to calculate a probability, p = 0.728, as shown in the following equation:
0.72575+ (0.72907- 0.72575). (2^07^0.600) = Q ?2g general to interpoiate the probability, p, associated with
(0.610-0.600)
z, find the z-values bracketing z in the normal table; i. e., find the z\ and z2 in adjoining rows of the table so that z\ <
z < z2. Next find the values p{ and p2 in the table corresponding to z\ and z2, respectively. The linearly interpolated
value for p is then: p=p, +(p2-p^)1 .
C-4
-------
Normal Probability Table
z-value p (prob. of /-value p
p (prob.
lesser value)
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
-3.50 0.00023
-3.49 0.00024
-3.48 0.00025
-3.47 0.00026
-3.46 0.00027
-3.45 0.00028
-3.44 0.00029
-3.43 0.00030
-3.42 0.00031
-3.41 0.00032
-3.40 0.00034
-3.39 0.00035
-3.38 0.00036
-3.37 0.00038
-3.36 0.00039
-3.35 0.00040
-3.34 0.00042
-3.33 0.00043
-3.32 0.00045
-3.31 0.00047
-3.30 0.00048
-3.29 0.00050
-3.28 0.00052
-3.27 0.00054
-3.26 0.00056
-3.25 0.00058
-3.24 0.00060
-3.23 0.00062
-3.22 0.00064
-3.21 0.00066
-3.20 0.00069
-3.19 0.00071
-3.18 0.00074
-3.17 0.00076
-3.16 0.00079
-3.15 0.00082
-3.14 0.00084
-3.13 0.00087
-3.12 0.00090
-3.11 0.00094
-3.10 0.00097
-3.09 0.00100
-3.08 0.00104
-3.07 0.00107
-3.06 0.00111
-3.05 0.00114
-3.04 0.00118
-3.03 0.00122
-3.02 0.00126
-3.01 0.00131
-3.00 .00135
-2.99 0.00139
-2.98 0.00144
-2.97 0.00149
-2.96 0.00154
-2.95 0.00159
-2.94 0.00164
-2.93 0.00169
-2.92 0.00175
-2.91 0.00181
-2.90 0.00187
-2.89 0.00193
-2.88 0.00199
-2.87 0.00205
-2.86 0.00212
-2.85 0.00219
-2.84 0.00226
-2.83 0.00233
-2.82 0.00240
-2.81 0.00248
-2.80 0.00256
-2.79 0.00264
-2.78 0.00272
-2.77 0.00280
-2.76 0.00289
-2.75 0.00298
-2.74 0.00307
-2.73 0.00317
-2.72 0.00326
-2.71 0.00336
-2.70 0.00347
-2.69 0.00357
-2.68 0.00368
-2.67 0.00379
-2.66 0.00391
-2.65 0.00402
-2.64 0.00415
-2.63 0.00427
-2.62 0.00440
-2.61 0.00453
-2.60 0.00466
-2.59 0.00480
-2.58 0.00494
-2.57 0.00508
-2.56 0.00523
-2.55 0.00539
-2.54 0.00554
-2.53 0.00570
-2.52 0.00587
-2.51 0.00604
-2.50 0.00621
-2.49 0.00639
-2.48 0.00657
-2.47 0.00676
-2.46 0.00695
-2.45 0.00714
-2.44 0.00734
-2.43 0.00755
-2.42 0.00776
-2.41 0.00798
-2.40 0.00820
-2.39 0.00842
-2.38 0.00866
-2.37 0.00889
-2.36 0.00914
-2.35 0.00939
-2.34 0.00964
-2.33 0.00990
-2.32 0.01017
-2.31 0.01044
-2.30 0.01072
-2.29 0.01101
-2.28 0.01130
-2.27 0.01160
-2.26 0.01191
-2.25 0.01222
-2.24 0.01255
-2.23 0.01287
-2.22 0.01321
-2.21 0.01355
-2.20 0.01390
-2.19 0.01426
-2.18 0.01463
-2.17 0.01500
-2.16 0.01539
-2.15 0.01578
-2.14 0.01618
-2.13 0.01659
-2.12 0.01700
-2.11 0.01743
-2.10 0.01786
-2.09 0.01831
-2.08 0.01876
-2.07 0.01923
-2.06 0.01970
-2.05 0.02018
-2.04 0.02068
-2.03 0.02118
-2.02 0.02169
-2.01 0.02222
-2.00 0.02275
-1.99 0.02330
-1.98 0.02385
-1.97 0.02442
-1.96 0.02500
-1.95 0.02559
-1.94 0.02619
-1.93 0.02680
-1.92 0.02743
-1.91 0.02807
-1.90 0.02872
-1.89 0.02938
-1.88 0.03005
-1.87 0.03074
-1.86 0.03144
-1.85 0.03216
-1.84 0.03288
-1.83 0.03362
-1.82 0.03438
-1.81 0.03515
-1.80 0.03593
-1.79 0.03673
-1.78 0.03754
-1.77 0.03836
-1.76 0.03920
-1.75 0.04006
-1.74 0.04093
-1.73 0.04182
-1.72 0.04272
-1.71 0.04363
-1.70 0.04457
-1.69 0.04551
-1.68 0.04648
-1.67 0.04746
-1.66 0.04846
-1.65 0.04947
-1.64 0.05050
-1.63 0.05155
-1.62 0.05262
-1.61 0.05370
-1.60 0.05480
-1.59 0.05592
-1.58 0.05705
-1.57 0.05821
-1.56 0.05938
-1.55 0.06057
-1.54 0.06178
-1.53 0.06301
-1.52 0.06426
-1.51 0.06552
-1.50 0.06681
-1.49 0.06811
-1.48 0.06944
-1.47 0.07078
-1.46 0.07215
-1.45 0.07353
-1.44 0.07493
-1.43 0.07636
-1.42 0.07780
-1.41 0.07927
-1.40 0.08076
-1.39 0.08226
-1.38 0.08379
-1.37 0.08534
-1.36 0.08691
-1.35 0.08851
-1.34 0.09012
-1.33 0.09176
-1.32 0.09342
-1.31 0.09510
-1.30 0.09680
-1.29 0.09853
-1.28 0.10027
-1.27 0.10204
-1.26 0.10383
-1.25 0.10565
-1.24 0.10749
-1.23 0.10935
-1.22 0.11123
-1.21 0.11314
-1.20 0.11507
-1.19 0.11702
-1.18 0.11900
-1.17 0.12100
-1.16 0.12302
-1.15 0.12507
-1.14 0.12714
-1.13 0.12924
-1.12 0.13136
-1.11 0.13350
-1.10 0.13567
-1.09 0.13786
-1.08 0.14007
-1.07 0.14231
-1.06 0.14457
-1.05 0.14686
-1.04 0.14917
-1.03 0.15151
-1.02 0.15386
-1.01 0.15625
C-5
-------
Normal Probability Table
z-value p (prob. of /-value p
p (prob.
lesser value)
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
-1.00 0.15866
-0.99 0.16109
-0.98 0.16354
-0.97 0.16602
-0.96 0.16853
-0.95 0.17106
-0.94 0.17361
-0.93 0.17619
-0.92 0.17879
-0.91 0.18141
-0.90 0.18406
-0.89 0.18673
-0.88 0.18943
-0.87 0.19215
-0.86 0.19489
-0.85 0.19766
-0.84 0.20045
-0.83 0.20327
-0.82 0.20611
-0.81 0.20897
-0.80 0.21186
-0.79 0.21476
-0.78 0.21770
-0.77 0.22065
-0.76 0.22363
-0.75 0.22663
-0.74 0.22965
-0.73 0.23270
-0.72 0.23576
-0.71 0.23885
-0.70 0.24196
-0.69 0.24510
-0.68 0.24825
-0.67 0.25143
-0.66 0.25463
-0.65 0.25785
-0.64 0.26109
-0.63 0.26435
-0.62 0.26763
-0.61 0.27093
-0.60 0.27425
-0.59 0.27760
-0.58 0.28096
-0.57 0.28434
-0.56 0.28774
-0.55 0.29116
-0.54 0.29460
-0.53 0.29806
-0.52 0.30153
-0.51 0.30503
-0.50 0.30854
-0.49 0.31207
-0.48 0.31561
-0.47 0.31918
-0.46 0.32276
-0.45 0.32636
-0.44 0.32997
-0.43 0.33360
-0.42 0.33724
-0.41 0.34090
-0.40 0.34458
-0.39 0.34827
-0.38 0.35197
-0.37 0.35569
-0.36 0.35942
-0.35 0.36317
-0.34 0.36693
-0.33 0.37070
-0.32 0.37448
-0.31 0.37828
-0.30 0.38209
-0.29 0.38591
-0.28 0.38974
-0.27 0.39358
-0.26 0.39743
-0.25 0.40129
-0.24 0.40517
-0.23 0.40905
-0.22 0.41294
-0.21 0.41683
-0.20 0.42074
-0.19 0.42465
-0.18 0.42858
-0.17 0.43251
-0.16 0.43644
-0.15 0.44038
-0.14 0.44433
-0.13 0.44828
-0.12 0.45224
-0.11 0.45620
-0.10 0.46017
-0.09 0.46414
-0.08 0.46812
-0.07 0.47210
-0.06 0.47608
-0.05 0.48006
-0.04 0.48405
-0.03 0.48803
-0.02 0.49202
-0.01 0.49601
0.00 0.50000
0.01 0.50399
0.02 0.50798
0.03 0.51197
0.04 0.51595
0.05 0.51994
0.06 0.52392
0.07 0.52790
0.08 0.53188
0.09 0.53586
0.10 0.53983
0.11 0.54380
0.12 0.54776
0.13 0.55172
0.14 0.55567
0.15 0.55962
0.16 0.56356
0.17 0.56749
0.18 0.57142
0.19 0.57535
0.20 0.57926
0.21 0.58317
0.22 0.58706
0.23 0.59095
0.24 0.59483
0.25 0.59871
0.26 0.60257
0.27 0.60642
0.28 0.61026
0.29 0.61409
0.30 0.61791
0.31 0.62172
0.32 0.62552
0.33 0.62930
0.34 0.63307
0.35 0.63683
0.36 0.64058
0.37 0.64431
0.38 0.64803
0.39 0.65173
0.40 0.65542
0.41 0.65910
0.42 0.66276
0.43 0.66640
0.44 0.67003
0.45 0.67364
0.46 0.67724
0.47 0.68082
0.48 0.68439
0.49 0.68793
0.50 0.69146
0.51 0.69497
0.52 0.69847
0.53 0.70194
0.54 0.70540
0.55 0.70884
0.56 0.71226
0.57 0.71566
0.58 0.71904
0.59 0.72240
0.60 0.72575
0.61 0.72907
0.62 0.73237
0.63 0.73565
0.64 0.73891
0.65 0.74215
0.66 0.74537
0.67 0.74857
0.68 0.75175
0.69 0.75490
0.70 0.75804
0.71 0.76115
0.72 0.76424
0.73 0.76730
0.74 0.77035
0.75 0.77337
0.76 0.77637
0.77 0.77935
0.78 0.78230
0.79 0.78524
0.80 0.78814
0.81 0.79103
0.82 0.79389
0.83 0.79673
0.84 0.79955
0.85 0.80234
0.86 0.80511
0.87 0.80785
0.88 0.81057
0.89 0.81327
0.90 0.81594
0.91 0.81859
0.92 0.82121
0.93 0.82381
0.94 0.82639
0.95 0.82894
0.96 0.83147
0.97 0.83398
0.98 0.83646
0.99 0.83891
1.00 0.84134
1.01 0.84375
1.02 0.84614
1.03 0.84849
1.04 0.85083
1.05 0.85314
1.06 0.85543
1.07 0.85769
1.08 0.85993
1.09 0.86214
1.10 0.86433
1.11 0.86650
1.12 0.86864
1.13 0.87076
1.14 0.87286
1.15 0.87493
1.16 0.87698
1.17 0.87900
1.18 0.88100
1.19 0.88298
1.20 0.88493
1.21 0.88686
1.22 0.88877
1.23 0.89065
1.24 0.89251
1.25 0.89435
1.26 0.89617
1.27 0.89796
1.28 0.89973
1.29 0.90147
1.30 0.90320
1.31 0.90490
1.32 0.90658
1.33 0.90824
1.34 0.90988
1.35 0.91149
1.36 0.91309
1.37 0.91466
1.38 0.91621
1.39 0.91774
1.40 0.91924
1.41 0.92073
1.42 0.92220
1.43 0.92364
1.44 0.92507
1.45 0.92647
1.46 0.92785
1.47 0.92922
1.48 0.93056
1.49 0.93189
C-6
-------
Normal Probability Table
z-value p (prob. of /-value p
p (prob.
lesser value)
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
z-value
p (prob. of
lesser value)
1.50 0.93319
1.51 0.93448
1.52 0.93574
1.53 0.93699
1.54 0.93822
1.55 0.93943
1.56 0.94062
1.57 0.94179
1.58 0.94295
1.59 0.94408
1.60 0.94520
1.61 0.94630
1.62 0.94738
1.63 0.94845
1.64 0.94950
1.65 0.95053
1.66 0.95154
1.67 0.95254
1.68 0.95352
1.69 0.95449
1.70 0.95543
1.71 0.95637
1.72 0.95728
1.73 0.95818
1.74 0.95907
1.75 0.95994
1.76 0.96080
1.77 0.96164
1.78 0.96246
1.79 0.96327
1.80 0.96407
1.81 0.96485
1.82 0.96562
1.83 0.96638
1.84 0.96712
1.85 0.96784
1.86 0.96856
1.87 0.96926
1.88 0.96995
1.89 0.97062
1.90 0.97128
1.91 0.97193
1.92 0.97257
1.93 0.97320
1.94 0.97381
1.95 0.97441
1.96 0.97500
1.97 0.97558
1.98 0.97615
1.99 0.97670
2.00 0.97725
2.01 0.97778
2.02 0.97831
2.03 0.97882
2.04 0.97932
2.05 0.97982
2.06 0.98030
2.07 0.98077
2.08 0.98124
2.09 0.98169
2.10 0.98214
2.11 0.98257
2.12 0.98300
2.13 0.98341
2.14 0.98382
2.15 0.98422
2.16 0.98461
2.17 0.98500
2.18 0.98537
2.19 0.98574
2.20 0.98610
2.21 0.98645
2.22 0.98679
2.23 0.98713
2.24 0.98745
2.25 0.98778
2.26 0.98809
2.27 0.98840
2.28 0.98870
2.29 0.98899
2.30 0.98928
2.31 0.98956
2.32 0.98983
2.33 0.99010
2.34 0.99036
2.35 0.99061
2.36 0.99086
2.37 0.99111
2.38 0.99134
2.39 0.99158
2.40 0.99180
2.41 0.99202
2.42 0.99224
2.43 0.99245
2.44 0.99266
2.45 0.99286
2.46 0.99305
2.47 0.99324
2.48 0.99343
2.49 0.99361
2.50 0.99379
2.51 0.99396
2.52 0.99413
2.53 0.99430
2.54 0.99446
2.55 0.99461
2.56 0.99477
2.57 0.99492
2.58 0.99506
2.59 0.99520
2.60 0.99534
2.61 0.99547
2.62 0.99560
2.63 0.99573
2.64 0.99585
2.65 0.99598
2.66 0.99609
2.67 0.99621
2.68 0.99632
2.69 0.99643
2.70 0.99653
2.71 0.99664
2.72 0.99674
2.73 0.99683
2.74 0.99693
2.75 0.99702
2.76 0.99711
2.77 0.99720
2.78 0.99728
2.79 0.99736
2.80 0.99744
2.81 0.99752
2.82 0.99760
2.83 0.99767
2.84 0.99774
2.85 0.99781
2.86 0.99788
2.87 0.99795
2.88 0.99801
2.89 0.99807
2.90 0.99813
2.91 0.99819
2.92 0.99825
2.93 0.99831
2.94 0.99836
2.95 0.99841
2.96 0.99846
2.97 0.99851
2.98 0.99856
2.99 0.99861
3.00 0.99865
3.01 0.99869
3.02 0.99874
3.03 0.99878
3.04 0.99882
3.05 0.99886
3.06 0.99889
3.07 0.99893
3.08 0.99896
3.09 0.99900
3.10 0.99903
3.11 0.99906
3.12 0.99910
3.13 0.99913
3.14 0.99916
3.15 0.99918
3.16 0.99921
3.17 0.99924
3.18 0.99926
3.19 0.99929
3.20 0.99931
3.21 0.99934
3.22 0.99936
3.23 0.99938
3.24 0.99940
3.25 0.99942
3.26 0.99944
3.27 0.99946
3.28 0.99948
3.29 0.99950
3.30 0.99952
3.31 0.99953
3.32 0.99955
3.33 0.99957
3.34 0.99958
3.35 0.99960
3.36 0.99961
3.37 0.99962
3.38 0.99964
3.39 0.99965
3.40 0.99966
3.41 0.99968
3.42 0.99969
3.43 0.99970
3.44 0.99971
3.45 0.99972
3.46 0.99973
3.47 0.99974
3.48 0.99975
3.49 0.99976
C-7
-------
------- |