[Federal Register: May 20, 2008 (Volume 73, Number 98)]
[Proposed Rules]
[Page 29183-29291]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr20my08-24]
[[Page 29183]]
-----------------------------------------------------------------------
Part II
Environmental Protection Agency
-----------------------------------------------------------------------
40 CFR Parts 50, 51, 53 et al.
National Ambient Air Quality Standards for Lead; Proposed Rule
[[Page 29184]]
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 50, 51, 53 and 58
[EPA-HQ-OAR-2006-0735; FRL-8563-9]
RIN 2060-AN83
National Ambient Air Quality Standards for Lead
AGENCY: Environmental Protection Agency (EPA).
ACTION: Proposed rule.
-----------------------------------------------------------------------
SUMMARY: Based on its review of the air quality criteria and national
ambient air quality standards (NAAQS) for lead (Pb), EPA proposes to
make revisions to the primary and secondary NAAQS for Pb to provide
requisite protection of public health and welfare, respectively. EPA
proposes to revise various elements of the primary standard to provide
increased protection for children and other at-risk populations against
an array of adverse health effects, most notably including neurological
effects, particularly neurocognitive and neurobehavioral effects, in
children. With regard to the level and indicator of the standard, EPA
proposes to revise the level to within the range of 0.10 to 0.30 [mu]g/
m\3\ in conjunction with retaining the current indicator of Pb in total
suspended particles (Pb-TSP) but with allowance for the use of Pb-
PM10 data, and solicits comment on alternative levels up to
0.50 [mu]g/m\3\ and down below 0.10 [mu]g/m\3\. With regard to the
averaging time and form of the standard, EPA proposes two options: To
retain the current averaging time of a calendar quarter and the current
not-to-be-exceeded form, revised to apply across a 3-year span; and to
revise the averaging time to a calendar month and the form to the
second-highest monthly average across a 3-year span. EPA also solicits
comment on revising the indicator to Pb-PM10 and on the same
broad range of levels on which EPA is soliciting comment for the Pb-TSP
indicator (up to 0.50 [mu]g/m\3\). EPA also invites comment on when, if
ever, it would be appropriate to set a NAAQS for Pb at a level of zero.
EPA proposes to make the secondary standard identical in all respects
to the proposed primary standard.
EPA is also proposing corresponding changes to data handling
procedures, including the treatment of exceptional events, and to
ambient air monitoring and reporting requirements for Pb including
those related to sampling and analysis methods, network design,
sampling schedule, and data reporting. Finally, EPA is providing
guidance on its proposed approach for implementing the proposed revised
primary and secondary standards for Pb.
Consistent with the terms of a court order, by September 15, 2008
the Administrator will sign a notice of final rulemaking for
publication in the Federal Register.
DATES: Comments must be received by July 21, 2008. Under the Paperwork
Reduction Act, comments on the information collection provisions must
be received by OMB on or before June 19, 2008.
Public Hearings: EPA intends to hold public hearings on this
proposed rule in June 2008 in St. Louis, Missouri and Baltimore,
Maryland. These will be announced in a separate Federal Register notice
that provides details, including specific times and addresses, for
these hearings.
ADDRESSES: Submit your comments, identified by Docket ID No. EPA-HQ-
OAR-2006-0735 by one of the following methods:
http://www.regulations.gov: Follow the online instructions
for submitting comments.
E-mail: a-and-r-Docket@epa.gov.
Fax: 202-566-9744.
Mail: Docket No. EPA-HQ-OAR-2006-0735, Environmental
Protection Agency, Mail code 6102T, 1200 Pennsylvania Ave., NW.,
Washington, DC 20460. Please include a total of two copies.
Hand Delivery: Docket No. EPA-HQ-OAR-2006-0735,
Environmental Protection Agency, EPA West, Room 3334, 1301 Constitution
Ave., NW., Washington, DC. Such deliveries are only accepted during the
Docket's normal hours of operation, and special arrangements should be
made for deliveries of boxed information.
Instructions: Direct your comments to Docket ID No. EPA-HQ-OAR-
2006-0735. The EPA's policy is that all comments received will be
included in the public docket without change and may be made available
online at http://www.regulations.gov, including any personal
information provided, unless the comment includes information claimed
to be Confidential Business Information (CBI) or other information
whose disclosure is restricted by statute. Do not submit information
that you consider to be CBI or otherwise protected through http://
www.regulations.gov or e-mail. The http://www.regulations.gov Web site
is an ``anonymous access'' system, which means EPA will not know your
identity or contact information unless you provide it in the body of
your comment. If you send an e-mail comment directly to EPA without
going through http://www.regulations.gov, your e-mail address will be
automatically captured and included as part of the comment that is
placed in the public docket and made available on the Internet. If you
submit an electronic comment, EPA recommends that you include your name
and other contact information in the body of your comment and with any
disk or CD-ROM you submit. If EPA cannot read your comment due to
technical difficulties and cannot contact you for clarification, EPA
may not be able to consider your comment. Electronic files should avoid
the use of special characters, any form of encryption, and be free of
any defects or viruses. For additional information about EPA's public
docket, visit the EPA Docket Center homepage at http://www.epa.gov/
epahome/dockets.htm.
Docket: All documents in the docket are listed in the http://
www.regulations.gov index. Although listed in the index, some
information is not publicly available, e.g., CBI or other information
whose disclosure is restricted by statute. Certain other material, such
as copyrighted material, will be publicly available only in hard copy.
Publicly available docket materials are available either electronically
in http://www.regulations.gov or in hard copy at the Air and Radiation
Docket and Information Center, EPA/DC, EPA West, Room 3334, 1301
Constitution Ave., NW., Washington, DC. The Public Reading Room is open
from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal
holidays. The telephone number for the Public Reading Room is (202)
566-1744 and the telephone number for the Air and Radiation Docket and
Information Center is (202) 566-1742.
FOR FURTHER INFORMATION CONTACT: For further information in general or
specifically with regard to sections I through III or VII, contact Dr.
Deirdre Murphy, Health and Environmental Impacts Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Mail code C504-06, Research Triangle Park, NC 27711; telephone:
919-541-0729; fax: 919-541-0237; e-mail: Murphy.deirdre@epa.gov. With
regard to Section IV, contact Mr. Mark Schmidt, Air Quality Analysis
Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Mail code C304-04, Research Triangle
Park, NC 27711; telephone: 919-541-2416; fax: 919-541-1903; e-mail:
Schmidt.mark@epa.gov. With regard to Section V, contact Mr. Kevin
Cavender,
[[Page 29185]]
Air Quality Analysis Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Mail code C304-06,
Research Triangle Park, NC 27711; telephone: 919-541-2364; fax: 919-
541-1903; e-mail: Cavender.kevin@epa.gov. With regard to Section VI,
contact Mr. Larry Wallace, Ph.D., Air Quality Policy Division, Office
of Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Mail code C539-01, Research Triangle Park, NC 27711; telephone:
919-541-0906; fax: 919-541-0824; e-mail: Wallace.larry@epa.gov.
SUPPLEMENTARY INFORMATION:
General Information
What Should I Consider as I Prepare My Comments for EPA?
1. Submitting CBI. Do not submit this information to EPA through
http://www.regulations.gov or e-mail. Clearly mark the part or all of
the information that you claim to be CBI. For CBI information in a disk
or CD-ROM that you mail to EPA, mark the outside of the disk or CD-ROM
as CBI and then identify electronically within the disk or CD-ROM the
specific information that is claimed as CBI. In addition to one
complete version of the comment that includes information claimed as
CBI, a copy of the comment that does not contain the information
claimed as CBI must be submitted for inclusion in the public docket.
Information so marked will not be disclosed except in accordance with
procedures set forth in 40 CFR part 2.
2. Tips for Preparing Your Comments. When submitting comments,
remember to:
Identify the rulemaking by docket number and other
identifying information (subject heading, Federal Register date and
page number).
Follow directions--the agency may ask you to respond to
specific questions or organize comments by referencing a Code of
Federal Regulations (CFR) part or section number.
Explain why you agree or disagree, suggest alternatives,
and substitute language for your requested changes.
Describe any assumptions and provide any technical
information and/or data that you used.
If you estimate potential costs or burdens, explain how
you arrived at your estimate in sufficient detail to allow for it to be
reproduced.
Provide specific examples to illustrate your concerns, and
suggest alternatives.
Explain your views as clearly as possible, avoiding the
use of profanity or personal threats.
Make sure to submit your comments by the comment period
deadline identified.
Availability of Related Information
A number of documents relevant to this rulemaking, including the
advance notice of proposed rulemaking (72 FR 71488), the Air Quality
Criteria for Lead (Criteria Document) (USEPA, 2006a), the Staff Paper,
related risk assessment reports, and other related technical documents
are available on EPA's Office of Air Quality Planning and Standards
(OAQPS) Technology Transfer Network (TTN) Web site at http://
www.epa.gov/ttn/naaqs/standards/pb/s_pb_index.html. These and other
related documents are also available for inspection and copying in the
EPA docket identified above.
Table of Contents
The following topics are discussed in this preamble:
I. Background
A. Legislative Requirements
B. History of Lead NAAQS Reviews
C. Current Related Lead Control Programs
D. Current Lead NAAQS Review
II. Rationale for Proposed Decision on the Primary Standard
A. Multimedia, Multipathway Considerations and Background
1. Atmospheric Emissions and Distribution of Lead
2. Air-Related Human Exposure Pathways
3. Nonair-Related and Air-Related Background Human Exposure
Pathways
4. Contributions to Children's Lead Exposures
B. Health Effects Information
1. Blood Lead
a. Internal Disposition of Lead
b. Use of Blood Lead as Dose Metric
c. Air-to-Blood Relationships
2. Nature of Effects
a. Broad Array of Effects
b. Neurological Effects in Children
3. Lead-Related Impacts on Public Health
a. At-Risk Subpopulations
b. Potential Public Health Impacts
4. Key Observations
C. Human Exposure and Health Risk Assessments
1. Overview of Risk Assessment From Last Review
2. Design Aspects of Exposure and Risk Assessments
a. CASAC Advice
b. Health Endpoint, Risk Metric and Concentration-response
Functions
c. Case Study Approach
d. Air Quality Scenarios
e. Categorization of Policy-Relevant Exposure Pathways
f. Analytical Steps
g. Generating Multiple Sets of Risk Results
h. Key Limitations and Uncertainties
3. Summary of Estimates and Key Observations
a. Blood Pb Estimates
b. IQ Loss Estimates
D. Conclusions on Adequacy of the Current Primary Standard
1. Background
a. The Current Standard
b. Policy Options Considered in the Last Review
2. Considerations in the Current Review
a. Evidence-Based Considerations
b. Exposure- and Risk-Based Considerations
3. CASAC Advice and Recommendations
4. Administrator's Proposed Conclusions Concerning Adequacy
E. Conclusions on the Elements of the Standard
1. Indicator
2. Averaging Time and Form
3. Level for a Pb NAAQS With Pb-TSP Indicator
a. Evidence-Based Considerations
b. Exposure- and Risk-Based Considerations
c. CASAC Advice and Recommendations
d. Administrator's Proposed Conclusion Concerning Level
4. Level for a Pb NAAQS With Pb-PM10 Indicator
a. Considerations With Regard to Particles Not Captured by
PM10
b. CASAC Advice
c. Approaches for Levels for a PM10-Based Standard
F. Proposed Decision on the Primary Standard
III. Rationale for Proposed Decision on the Secondary Standard
A. Welfare Effects Information
B. Screening Level Ecological Risk Assessment
1. Design Aspects of the Assessment and Associated Uncertainties
2. Summary of Results
C. The Secondary Standard
1. Background on the Current Standard
2. Approach for Current Review
3. Conclusions on Adequacy of the Current Standard
a. Evidence-Based Considerations
b. Risk-Based Considerations
c. CASAC Advice and Recommendations
d. Administrator's Proposed Conclusions on Adequacy of Current
Standard
4. Conclusions and Proposed Decision on the Elements of the
Secondary Standard
IV. Proposed Appendix R on Interpretation of the NAAQS for Lead and
Proposed Revisions to the Exceptional Events Rule
A. Background
B. Interpretation of the NAAQS for Lead
1. Interpretation of a Standard Based on Pb-TSP
2. Interpretation of Alternative Elements
C. Exceptional Events Information Submission Schedule
V. Proposed Amendments to Ambient Monitoring Requirements
A. Sampling and Analysis Methods
1. Background
2. Proposed Changes
a. Pb-TSP Sampling Method
b. Pb-PM10 Sampling Method
c. Analysis Method
d. FEM Criteria
e. Quality Assurance
[[Page 29186]]
B. Network Design
1. Background
2. Proposed Changes
C. Sampling Schedule
1. Background
2. Proposed Changes
D. Monitoring for the Secondary NAAQS
1. Background
2. Proposed Changes
E. Other Monitoring Regulation Changes
1. Reporting of Average Pressure and Temperature
2. Special Purpose Monitoring Exemption
VI. Implementation Considerations
A. Designations for the Lead NAAQS
1. Potential Schedule for Designations of A Revised Lead NAAQS
B. Lead Nonattainment Area Boundaries
1. County-Based Boundaries
2. MSA-Based Boundaries
C. Classifications
D. Section 110(a)(2) Lead NAAQS Infrastructure Requirements
E. Attainment Dates
F. Attainment Planning Requirements
1. Schedule for Attaining a Revised Pb NAAQS
2. RACM for Lead Nonattainment Areas
3. Demonstration of Attainment for Lead Nonattainment Areas
4. Reasonable Further Progress (RFP)
5. Contingency Measures
6. Nonattainment New Source Review (NSR) and Prevention of
Significant Deterioration (PSD) Requirements
7. Emissions Inventories
8. Modeling
G. General Conformity
H. Transition From the Current NAAQS to a Revised NAAQS for Lead
VII. Statutory and Executive Order Reviews
References
I. Background
A. Legislative Requirements
Two sections of the Clean Air Act (Act) govern the establishment
and revision of the NAAQS. Section 108 (42 U.S.C. 7408) directs the
Administrator to identify and list each air pollutant that ``in his
judgment, cause or contribute to air pollution which may reasonably be
anticipated to endanger public health and welfare'' and whose
``presence * * * in the ambient air results from numerous or diverse
mobile or stationary sources'' and to issue air quality criteria for
those that are listed. Air quality criteria are to ``accurately reflect
the latest scientific knowledge useful in indicating the kind and
extent of all identifiable effects on public health or welfare which
may be expected from the presence of [a] pollutant in ambient air * *
*''. Section 109 (42 U.S.C. 7409) directs the Administrator to propose
and promulgate ``primary'' and ``secondary'' NAAQS for pollutants
listed under section 108. Section 109(b)(1) defines a primary standard
as one ``the attainment and maintenance of which in the judgment of the
Administrator, based on [air quality] criteria and allowing an adequate
margin of safety, are requisite to protect the public health.'' \1\ A
secondary standard, as defined in Section 109(b)(2), must ``specify a
level of air quality the attainment and maintenance of which, in the
judgment of the Administrator, based on criteria, is requisite to
protect the public welfare from any known or anticipated adverse
effects associated with the presence of [the] pollutant in the ambient
air.'' \2\
---------------------------------------------------------------------------
\1\ The legislative history of section 109 indicates that a
primary standard is to be set at ``the maximum permissible ambient
air level * * * which will protect the health of any [sensitive]
group of the population,'' and that for this purpose ``reference
should be made to a representative sample of persons comprising the
sensitive group rather than to a single person in such a group.'' S.
Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970).
\2\ Welfare effects as defined in section 302(h) (42 U.S.C.
7602(h)) include, but are not limited to, ``effects on soils, water,
crops, vegetation, man-made materials, animals, wildlife, weather,
visibility and climate, damage to and deterioration of property, and
hazards to transportation, as well as effects on economic values and
on personal comfort and well-being.''
---------------------------------------------------------------------------
The requirement that primary standards include an adequate margin
of safety was intended to address uncertainties associated with
inconclusive scientific and technical information available at the time
of standard setting. It was also intended to provide a reasonable
degree of protection against hazards that research has not yet
identified. Lead Industries Association v. EPA, 647 F.2d 1130, 1154
(D.C. Cir 1980), cert. denied, 449 U.S. 1042 (1980); American Petroleum
Institute v. Costle, 665 F.2d 1176, 1186 (D.C. Cir. 1981), cert.
denied, 455 U.S. 1034 (1982). Both kinds of uncertainties are
components of the risk associated with pollution at levels below those
at which human health effects can be said to occur with reasonable
scientific certainty. Thus, in selecting primary standards that include
an adequate margin of safety, the Administrator is seeking not only to
prevent pollution levels that have been demonstrated to be harmful but
also to prevent lower pollutant levels that may pose an unacceptable
risk of harm, even if the risk is not precisely identified as to nature
or degree. The CAA does not require the Administrator to establish a
primary NAAQS at a zero-risk level or at background concentration
levels, see Lead Industries Association v. EPA, 647 F.2d at 1156 n. 51,
but rather at a level that reduces risk sufficiently so as to protect
public health with an adequate margin of safety.
The selection of any particular approach to providing an adequate
margin of safety is a policy choice left specifically to the
Administrator's judgment. Lead Industries Association v. EPA, 647 F.2d
at 1161-62. In addressing the requirement for an adequate margin of
safety, EPA considers such factors as the nature and severity of the
health effects involved, the size of the population(s) at risk, and the
kind and degree of the uncertainties that must be addressed.
In setting standards that are ``requisite'' to protect public
health and welfare, as provided in section 109(b), EPA's task is to
establish standards that are neither more nor less stringent than
necessary for these purposes. Whitman v. American Trucking
Associations, 531 U.S. 457, 473. Further the Supreme Court ruled that
``[t]he text of Sec. 109(b), interpreted in its statutory and
historical context and with appreciation for its importance to the CAA
as a whole, unambiguously bars cost considerations from the NAAQS-
setting process * * *'' Id. at 472.\3\ Section 109(d)(1) of the Act
requires that ``[n]ot later than December 31, 1980, and at 5-year
intervals thereafter, the Administrator shall complete a thorough
review of the criteria published under section 108 and the national
ambient air quality standards promulgated under this section and shall
make such revisions in such criteria and standards and promulgate such
new standards as may be appropriate in accordance with section 108 and
subsection (b) of this section.'' Section 109(d)(2)(A) requires that
``The Administrator shall appoint an independent scientific review
committee composed of seven members including at least one member of
the National Academy of Sciences, one physician, and one person
representing State air pollution control agencies.'' Section
109(d)(2)(B) requires that, ``[n]ot later than January 1, 1980, and at
five-year intervals thereafter, the committee referred to in
subparagraph (A) shall complete a review of the criteria published
under section 108 and the national primary and secondary ambient air
quality standards promulgated under this section and shall recommend to
the Administrator any new national ambient air quality standards and
revisions of existing criteria and standards as may be appropriate
under section 108 and subsection (b) of this
[[Page 29187]]
section.'' Since the early 1980's, this independent review function has
been performed by the Clean Air Scientific Advisory Committee (CASAC)
of EPA's Science Advisory Board.
---------------------------------------------------------------------------
\3\ In considering whether the CAA allowed for economic
considerations to play a role in the promulgation of the NAAQS, the
Supreme Court rejected arguments that because many more factors than
air pollution might affect public health, EPA should consider
compliance costs that produce health losses in setting the NAAQS.
531 U.S. at 466. Thus, EPA may not take into account possible public
health impacts from the economic cost of implementation. Id.
---------------------------------------------------------------------------
B. History of Lead NAAQS Reviews
On October 5, 1978 EPA promulgated primary and secondary NAAQS for
Pb under section 109 of the Act (43 FR 46246). Both primary and
secondary standards were set at a level of 1.5 micrograms per cubic
meter ([mu]g/m\3\), measured as Pb in total suspended particulate
matter (Pb-TSP), not to be exceeded by the maximum arithmetic mean
concentration averaged over a calendar quarter. This standard was based
on the 1977 Air Quality Criteria for Lead (USEPA, 1977).
A review of the Pb standards was initiated in the mid-1980s. The
scientific assessment for that review is described in the 1986 Air
Quality Criteria for Lead (USEPA, 1986a), the associated Addendum
(USEPA, 1986b) and the 1990 Supplement (USEPA, 1990a). As part of the
review, the Agency designed and performed human exposure and health
risk analyses (USEPA, 1989), the results of which were presented in a
1990 Staff Paper (USEPA, 1990b). Based on the scientific assessment and
the human exposure and health risk analyses, the 1990 Staff Paper
presented options for the Pb NAAQS level in the range of 0.5 to 1.5
[mu]g/m3, and suggested the second highest monthly average
in three years for the form and averaging time of the standard (USEPA,
1990b). After consideration of the documents developed during the
review and the significantly changed circumstances since Pb was listed
in 1976, the Agency did not propose any revisions to the 1978 Pb NAAQS.
In a parallel effort, the Agency developed the broad, multi-program,
multimedia, integrated U.S. Strategy for Reducing Lead Exposure (USEPA,
1991). As part of implementing this strategy, the Agency focused
efforts primarily on regulatory and remedial clean-up actions aimed at
reducing Pb exposures from a variety of nonair sources judged to pose
more extensive public health risks to U.S. populations, as well as on
actions to reduce Pb emissions to air, such as bringing more areas into
compliance with the existing Pb NAAQS (USEPA, 1991).
C. Current Related Lead Control Programs
States are primarily responsible for ensuring attainment and
maintenance of national ambient air quality standards once EPA has
established them. Under section 110 of the Act (42 U.S.C. 7410) and
related provisions, States are to submit, for EPA approval, State
implementation plans (SIPs) that provide for the attainment and
maintenance of such standards through control programs directed to
sources of the pollutants involved. The States, in conjunction with
EPA, also administer the prevention of significant deterioration
program (42 U.S.C. 7470-7479) for these pollutants. In addition,
Federal programs provide for nationwide reductions in emissions of
these and other air pollutants through the Federal Motor Vehicle
Control Program under Title II of the Act (42 U.S.C. 7521-7574), which
involves controls for automobile, truck, bus, motorcycle, nonroad
engine, and aircraft emissions; the new source performance standards
under section 111 of the Act (42 U.S.C. 7411); and the national
emission standards for hazardous air pollutants under section 112 of
the Act (42 U.S.C. 7412).
As Pb is a multimedia pollutant, a broad range of Federal programs
beyond those that focus on air pollution control provide for nationwide
reductions in environmental releases and human exposures. In addition,
the Centers for Disease Control and Prevention (CDC) programs provide
for the tracking of children's blood Pb levels nationally and provide
guidance on levels at which medical and environmental case management
activities should be implemented (CDC, 2005a; ACCLPP, 2007).\4\ In
1991, the Secretary of the Health and Human Services (HHS)
characterized Pb poisoning as the ``number one environmental threat to
the health of children in the United States'' (Alliance to End
Childhood Lead Poisoning, 1991). In 1997, President Clinton created, by
Executive Order 13045, the President's Task Force on Environmental
Health Risks and Safety Risks to Children in response to increased
awareness that children face disproportionate risks from environmental
health and safety hazards (62 FR 19885).\5\ By Executive Orders issued
in October 2001 and April 2003, President Bush extended the work for
the Task Force for an additional three and a half years beyond its
original charter (66 FR 52013 and 68 FR 19931). The Task Force set a
Federal goal of eliminating childhood Pb poisoning by the year 2010 and
reducing Pb poisoning in children was the Task Force's top priority.
---------------------------------------------------------------------------
\4\ As described in Section III below the CDC stated in 2005
that no ``safe'' threshold for blood Pb levels in young children has
been identified (CDC, 2005a).
\5\ Co-chaired by the Secretary of the HHS and the Administrator
of the EPA, the Task Force consisted of representatives from 16
Federal departments and agencies.
---------------------------------------------------------------------------
Federal abatement programs provide for the reduction in human
exposures and environmental releases from in-place materials containing
Pb (e.g., Pb-based paint, urban soil and dust, and contaminated waste
sites). Federal regulations on disposal of Pb-based paint waste help
facilitate the removal of Pb-based paint from residences.\6\ Further,
in 1991, EPA lowered the maximum levels of Pb permitted in public water
systems from 50 parts per billion (ppb) to 15 ppb (56 FR 26460).
---------------------------------------------------------------------------
\6\ See ``Criteria for Classification of Solid Waste Disposal
Facilities and Practices and Criteria for Municipal Solid Waste
Landfills: Disposal of Residential Lead-Based Paint Waste; Final
Rule'' EPA-HQ-RCRA-2001-0017.
---------------------------------------------------------------------------
Federal programs to reduce exposure to Pb in paint, dust, and soil
are specified under the comprehensive federal regulatory framework
developed under the Residential Lead-Based Paint Hazard Reduction Act
(Title X). Under Title X and Title IV of the Toxic Substances Control
Act, EPA has established regulations and associated programs in the
following five categories: (1) Training and certification requirements
for persons engaged in lead-based paint activities; accreditation of
training providers; authorization of State and Tribal lead-based paint
programs; and work practice standards for the safe, reliable, and
effective identification and elimination of lead-based paint hazards;
(2) ensuring that, for most housing constructed before 1978, lead-based
paint information flows from sellers to purchasers, from landlords to
tenants, and from renovators to owners and occupants; (3) establishing
standards for identifying dangerous levels of Pb in paint, dust and
soil; (4) providing grant funding to establish and maintain State and
Tribal lead-based paint programs, and to address childhood lead
poisoning in the highest-risk communities; and (5) providing
information on Pb hazards to the public, including steps that people
can take to protect themselves and their families from lead-based paint
hazards.
Under Title IV of TSCA, EPA established standards identifying
hazardous levels of lead in residential paint, dust, and soil in 2001.
This regulation supports the implementation of other regulations which
deal with worker training and certification, Pb hazard disclosure in
real estate transactions, Pb hazard evaluation and control in
Federally-owned housing prior to sale and housing receiving Federal
assistance, and U.S. Department of Housing and Urban Development grants
to local jurisdictions to perform
[[Page 29188]]
Pb hazard control. The TSCA Title IV term ``lead-based paint hazard''
implemented through this regulation identifies lead-based paint and all
residential lead-containing dust and soil regardless of the source of
Pb, which, due to their condition and location, would result in adverse
human health effects. One of the underlying principles of Title X is to
move the focus of public and private decision makers away from the mere
presence of lead-based paint, to the presence of lead-based paint
hazards, for which more substantive action should be undertaken to
control exposures, especially to young children. In addition the
success of the program will rely on the voluntary participation of
states and tribes as well as counties and cities to implement the
programs and on property owners to follow the standards and EPA's
recommendations. If EPA were to set unreasonable standards (e.g.,
standards that would recommend removal of all Pb from paint, dust, and
soil), States and Tribes may choose to opt out of the Title X Pb
program and property owners may choose to ignore EPA's advice believing
it lacks credibility and practical value. Consequently, EPA needed to
develop standards that would not waste resources by chasing risks of
negligible importance and that would be accepted by States, Tribes,
local governments and property owners. In addition, a separate
regulation establishes, among other things, under authority of TSCA
section 402, residential Pb dust cleanup levels and amendments to dust
and soil sampling requirements (66 FR 1206).
On March 31, 2008, the Agency issued a new rule (Lead: Renovation,
Repair and Painting [RRP] Program) to protect children from lead-based
paint hazards. This rule applies to renovators and maintenance
professionals who perform renovation, repair, or painting in housing,
child-care facilities, and schools built prior to 1978. It requires
that contractors and maintenance professionals be certified; that their
employees be trained; and that they follow protective work practice
standards. These standards prohibit certain dangerous practices, such
as open flame burning or torching of lead-based paint. The required
work practices also include posting warning signs, restricting
occupants from work areas, containing work areas to prevent dust and
debris from spreading, conducting a thorough cleanup, and verifying
that cleanup was effective. The rule will be fully effective by April
2010. States and tribes may become authorized to implement this rule,
and the rule contains procedures for the authorization of states,
territories, and tribes to administer and enforce these standards and
regulations in lieu of a federal program. In announcing this rule, EPA
noted that almost 38 million homes in the United States contain some
lead-based paint, and that this rule's requirements were key components
of a comprehensive effort to eliminate childhood Pb poisoning. To
foster adoption of the rule's measures, EPA also intends to conduct an
extensive education and outreach campaign to promote awareness of these
new requirements.
Programs associated with the Comprehensive Environmental Response,
Compensation, and Liability Act (CERCLA or Superfund) and Resource
Conservation Recovery Act (RCRA) also implement abatement programs,
reducing exposures to Pb and other pollutants. For example, EPA
determines and implements protective levels for Pb in soil at Superfund
sites and RCRA corrective action facilities. Federal programs,
including those implementing RCRA, provide for management of hazardous
substances in hazardous and municipal solid waste.\7\ For example,
Federal regulations concerning batteries in municipal solid waste
facilitate the collection and recycling or proper disposal of batteries
containing Pb.\8\ Similarly, Federal programs provide for the reduction
in environmental releases of hazardous substances such as Pb in the
management of wastewater (http://www.epa.gov/owm/).
---------------------------------------------------------------------------
\7\ See, e.g., ``Hazardous Waste Management System;
Identification and Listing of Hazardous Waste: Inorganic Chemical
Manufacturing Wastes; Land Disposal Restrictions for Newly
Identified Wastes and CERCLA Hazardous Substance Designation and
Reportable Quantities; Final Rule'', http://www.epa.gov/epaoswer/
hazwaste/state/revision/frs/fr195.pdf and http://www.epa.gov/
epaoswer/hazwaste/ldr/basic.htm.
\8\ See, e.g., ``Implementation of the Mercury-Containing and
Rechargeable Battery Management Act'' http://www.epa.gov/epaoswer/
hazwaste/recycle/battery.pdf and ``Municipal Solid Waste Generation,
Recycling, and Disposal in the United States: Facts and Figures for
2005'' http://www.epa.gov/epaoswer/osw/conserve/resources/msw-
2005.pdf.
---------------------------------------------------------------------------
A variety of federal nonregulatory programs also provide for
reduced environmental release of Pb containing materials through more
general encouragement of pollution prevention, promotion of reuse and
recycling, reduction of priority and toxic chemicals in products and
waste, and conservation of energy and materials. These include the
Resource Conservation Challenge (http://www.epa.gov/epaoswer/osw/
conserve/index.htm), the National Waste Minimization Program (http://
www.epa.gov/epaoswer/hazwaste/minimize/leadtire.htm), ``Plug in to
eCycling'' (a partnership between EPA and consumer electronics
manufacturers and retailers; http://www.epa.gov/epaoswer/hazwaste/
recycle/electron/crt.htm#crts), and activities to reduce the practice
of backyard trash burning (http://www.epa.gov/msw/backyard/pubs.htm).
Efforts such as those programs described above have been successful
in that blood Pb levels in all segments of the population have dropped
significantly from levels observed around 1990. In particular, blood Pb
levels for the general population of children 1 to 5 years of age have
dropped to a median level of 1.6 [mu]g/dL and a level of 3.9 [mu]g/dL
for the 90th percentile child in the 2003-2004 National Health and
Nutrition Examination Survey (NHANES) as compared to median and 90th
percentile levels in 1988-1991 of 3.5 [mu]g/dL and 9.4 [mu]g/dL,
respectively (http://www.epa.gov/envirohealth/children/body_burdens/
b1-table.htm). These levels (median and 90th percentile) for the
general population of young children \9\ are at the low end of the
historic range of blood Pb levels for general population of children
aged 1-5 years. However, as discussed in Section II.B.1.b, levels have
been found to vary among children of different socioeconomic status and
other demographic characteristics (CD, p. 4-21) and racial/ethnic and
income disparities in blood Pb levels in children persist. The decline
in blood Pb levels in the United States has resulted from coordinated,
intensive efforts at the national, state, and local levels. The Agency
has continued to grapple with soil and dust Pb levels from the
historical use of Pb in paint and gasoline and other sources.
---------------------------------------------------------------------------
\9\ The 95th percentile value for the 2003-2004 NHANES is 5.1
[mu]g/dL (Axelrad, 2008).
---------------------------------------------------------------------------
EPA's research program, with other Federal agencies, defines,
encourages and conducts research needed to locate and assess serious
risks and to develop methods and tools to characterize and help reduce
risks. For example, EPA's Integrated Exposure Uptake Biokinetic Model
for Lead in Children (IEUBK model) for Pb in children and the Adult
Lead Methodology are widely used and accepted as tools that provide
guidance in evaluating site specific data. More recently, in
recognition of the need for a single model that predicts Pb
concentrations in tissues for children and adults, EPA is developing
the All Ages Lead Model (AALM) to provide researchers and risk
assessors with a
[[Page 29189]]
pharmacokinetic model capable of estimating blood, tissue, and bone
concentrations of Pb based on estimates of exposure over the lifetime
of the individual. EPA research activities on substances including Pb
focus on better characterizing aspects of health and environmental
effects, exposure, and control or management of environmental releases
(see http://www.epa.gov/ord/researchaccomplishments/index.html).
D. Current Lead NAAQS Review
EPA initiated the current review of the air quality criteria for Pb
on November 9, 2004, with a general call for information (69 FR 64926).
A project work plan (USEPA, 2005a) for the preparation of the Criteria
Document was released in January 2005 for CASAC and public review. EPA
held a series of workshops in August 2005, inviting recognized
scientific experts to discuss initial draft materials that dealt with
various lead-related issues being addressed in the Pb air quality
criteria document. The first draft of the Criteria Document (USEPA,
2005b) was released for CASAC and public review in December 2005 and
discussed at a CASAC meeting held on February 28-March 1, 2006.
A second draft Criteria Document (USEPA, 2006b) was released for
CASAC and public review in May 2006, and discussed at the CASAC meeting
on June 28, 2006. A subsequent draft of Chapter 7--Integrative
Synthesis (Chapter 8 in the final Criteria Document), released on July
31, 2006, was discussed at an August 15, 2006, CASAC teleconference.
The final Criteria Document was released on September 30, 2006 (USEPA,
2006a; cited throughout this preamble as CD). While the Criteria
Document focuses on new scientific information available since the last
review, it integrates that information with scientific criteria from
previous reviews.
In February 2006, EPA released the Plan for Review of the National
Ambient Air Quality Standards for Lead (USEPA, 2006c) that described
Agency plans and a timeline for reviewing the air quality criteria,
developing human exposure and risk assessments and an ecological risk
assessment, preparing a policy assessment, and developing the proposed
and final rulemakings.
In May 2006, EPA released for CASAC and public review a draft
Analysis Plan for Human Health and Ecological Risk Assessment for the
Review of the Lead National Ambient Air Quality Standards (USEPA,
2006d), which was discussed at a June 29, 2006, CASAC meeting
(Henderson, 2006). The May 2006 assessment plan discussed two
assessment phases: A pilot phase and a full-scale phase. The pilot
phase of both the human health and ecological risk assessments was
presented in the draft Lead Human Exposure and Health Risk Assessments
and Ecological Risk Assessment for Selected Areas (ICF, 2006;
henceforth referred to as the first draft Risk Assessment Report) which
was released for CASAC and public review in December 2006. The first
draft Staff Paper, also released in December 2006, discussed the pilot
assessments and the most policy-relevant science from the Criteria
Document. These documents were reviewed by CASAC and the public at a
public meeting on February 6-7, 2007 (Henderson, 2007a).
Subsequent to that meeting, EPA conducted full-scale human exposure
and health risk assessments, although no further work was done on the
ecological assessment due to resource limitations. A second draft Risk
Assessment Report (USEPA, 2007a), containing the full-scale human
exposure and health risk assessments, was released in July 2007 for
review by CASAC at a meeting held on August 28-29, 2007. Taking into
consideration CASAC comments (Henderson, 2007b) and public comments on
that document, we conducted additional human exposure and health risk
assessments. A final Risk Assessment Report (USEPA, 2007b) and final
Staff Paper (USEPA, 2007c) were released on November 1, 2007.
The final Staff Paper presents OAQPS staff's evaluation of the
public health and welfare policy implications of the key studies and
scientific information contained in the Criteria Document and presents
and interprets results from the quantitative risk/exposure analyses
conducted for this review. Further, the Staff Paper presents OAQPS
staff recommendations on a range of policy options for the
Administrator to consider concerning whether, and if so how, to revise
the primary and secondary Pb NAAQS. Such an evaluation of policy
implications is intended to help ``bridge the gap'' between the
scientific assessment contained in the Criteria Document and the
judgments required of the EPA Administrator in determining whether it
is appropriate to retain or revise the NAAQS for Pb. In evaluating the
adequacy of the current standard and a range of alternatives, the Staff
Paper considered the available scientific evidence and quantitative
risk-based analyses, together with related limitations and
uncertainties, and focused on the information that is most pertinent to
evaluating the basic elements of national ambient air quality
standards: indicator,\10\ averaging time, form,\11\ and level. These
elements, which together serve to define each standard, must be
considered collectively in evaluating the public health and welfare
protection afforded by the Pb standards. The information, conclusions,
and OAQPS staff recommendations presented in the Staff Paper were
informed by comments and advice received from CASAC in its reviews of
the earlier draft Staff Paper and drafts of related risk/exposure
assessment reports, as well as comments on these earlier draft
documents submitted by public commenters.
---------------------------------------------------------------------------
\10\ The ``indicator'' of a standard defines the chemical
species or mixture that is to be measured in determining whether an
area attains the standard.
\11\ The ``form'' of a standard defines the air quality
statistic that is to be compared to the level of the standard in
determining whether an area attains the standard.
---------------------------------------------------------------------------
Subsequent to completion of the Staff Paper, EPA issued an advance
notice of proposed rulemaking (ANPR) that was signed by the
Administrator on December 5, 2007 (72 FR 71488-71544). The ANPR is one
of the key features of the new NAAQS review process that EPA has
instituted over the past two years to help to improve the efficiency of
the process the Agency uses in reviewing the NAAQS while ensuring that
the Agency's decisions are informed by the best available science and
broad participation among experts in the scientific community and the
public. The ANPR provided the public an opportunity to comment on a
wide range of policy options that could be considered by the
Administrator. The substantial number of comments we received on the Pb
NAAQS ANPR helped inform the narrower range of options we are proposing
and taking comment on today. The new process (described at http://
www.epa.gov/ttn/naaqs/.) is being incorporated into the various ongoing
NAAQS reviews being conducted by the Agency, including the current
review of the Pb NAAQS.
A public meeting of the CASAC was held on December 12-13, 2007 to
provide advice and recommendations to the Administrator based on its
review of the ANPR and the previously released final Staff Paper and
Risk Assessment Report. Information about this meeting was published in
the Federal Register on November 20, 2007 (72 FR 65335-65336),
transcripts of the meeting are in the Docket for this review and
CASAC's letter to the Administrator (Henderson, 2008) is also available
on the EPA Web site (http://www.epa.gov/sab).
[[Page 29190]]
A public comment period for the ANPR extended from December 17,
2007 through January 16, 2008 and comments received are in the Docket
for this review. Comments were received from nearly 9000 private
citizens (roughly 200 of them were not part of one of several mass
comment campaign), 13 state and local agencies, one federal agency,
three regional or national associations of government agencies or
officials, 15 nongovernmental environmental or public health
organizations (including one submission on behalf of a coalition of 23
organizations) and five industries or industry organizations. Although
the Agency has not developed formal responses to comments received on
the ANPR, these comments have been considered in the development of
this notice and are generally described in subsequent sections on
proposed conclusions with regard to the adequacy of the standards and
with regard to the Administrator's proposed decisions on revisions to
the standards.
The schedule for completion of this review is governed by a
judicial order in Missouri Coalition for the Environment, v. EPA (No.
4:04CV00660 ERW, Sept. 14, 2005). The order governing this review,
entered by the court on September 14, 2005 and amended on April 29,
2008, specifies that EPA sign, for publication, notices of proposed and
final rulemaking concerning its review of the Pb NAAQS no later than
May 1, 2008 and September 15, 2008, respectively. In light of the
compressed schedule ordered by the court for issuing the final rule,
EPA may be able to respond only to those comments submitted during the
public comment period on this proposal. EPA has considered all of the
comments submitted to date in preparing this proposal, but if
commenters believe that comments submitted on the ANPR are fully
applicable to the proposal and wish to ensure that those comments are
addressed by EPA as part of the final rulemaking, the earlier comments
should be resubmitted during the comment period on this proposal.
This action presents the Administrator's proposed decisions on the
review of the current primary and secondary Pb standards. Throughout
this preamble a number of judgments, conclusions, findings, and
determinations proposed by the Administrator are noted. While they
identify the reasoning that supports this proposal, they are not
intended to be final or conclusive in nature. The EPA invites general,
specific, and/or technical comments on all issues involved with this
proposal, including all such proposed judgments, conclusions, findings,
and determinations.
II. Rationale for Proposed Decision on the Primary Standard
This section presents the rationale for the Administrator's
proposed decision that the current primary standard is not requisite to
protect public health with an adequate margin of safety, and that the
existing Pb primary standard should be revised. With regard to the
primary standard for Pb, EPA is proposing options for the revision of
the various elements of the standard to provide increased protection
for children and other at-risk populations against an array of adverse
health effects, most notably including neurological effects in
children, particularly neurocognitive and neurobehavioral effects. With
regard to the level and indicator of the standard, EPA proposes to
revise the level of the standard to a level within the range of 0.10 to
0.30 [mu]g/m\3\ in conjunction with retaining the current indicator of
Pb in total suspended particles (Pb-TSP) but with allowance for the use
of Pb-PM10 data. With regard to the form and averaging time of the
standard, EPA proposes the following options: (1) To retain the current
averaging time of a calendar quarter and the current not-to-be-exceeded
form, revised so as to apply across a 3-year span, and (2) to revise
the averaging time to a calendar month and the form to be the second-
highest monthly average across a 3-year span. EPA also solicits comment
on revising the indicator to Pb-PM10.
As discussed more fully below, this proposal is based on a thorough
review, in the Criteria Document, of the latest scientific information
on human health effects associated with the presence of Pb in the
ambient air. This proposal also takes into account: (1) Staff
assessments of the most policy-relevant information in the Criteria
Document and staff analyses of air quality, human exposure, and health
risks presented in the Staff Paper, upon which staff recommendations
for revisions to the primary Pb standard are based; (2) CASAC advice
and recommendations, as reflected in discussions of the ANPR and drafts
of the Criteria Document and Staff Paper at public meetings, in
separate written comments, and in CASAC's letters to the Administrator;
and (3) public comments received during the development of these
documents, either in connection with CASAC meetings or separately.
In developing this proposal, EPA has drawn upon an integrative
synthesis of the entire body of evidence, published through late 2006,
on human health effects associated with Pb exposure. Some 6000 newly
available studies were considered in this review. As discussed below in
section II.B, this body of evidence addresses a broad range of health
endpoints associated with exposure to Pb (EPA, 2006a, chapter 8), and
includes hundreds of epidemiologic studies conducted in the U.S.,
Canada, and many countries around the world since the time of the last
review (EPA, 2006a, chapter 6). This proposal also draws upon the
results of the quantitative exposure and risk assessments, discussed
below in section II.C. Evidence- and exposure/risk-based considerations
that form the basis for the Administrator's proposed decisions on the
adequacy of the current standard and on the elements of the proposed
alternative standards are discussed below in section II.D.2 and II.D.3,
respectively.
A. Multimedia, Multipathway Considerations and Background
1. Atmospheric Emissions and Distribution of Lead
Lead is emitted into the air from many sources encompassing a wide
variety of source types (Staff Paper, Section 2.2). Further, once
deposited out of the air, Pb can subsequently be resuspended into the
air (CD, pp. 2-62 to 2-66). There are over 100 categories of sources of
Pb emissions included in the EPA's 2002 National Emissions Inventory
(NEI),\12 \ the top five of which include: Mobile sources (leaded
aviation gas) \13\; industrial, commercial, institutional and process
boilers; utility boilers; iron and steel foundries; and primary Pb
smelting (Staff Paper Section 2.2). Further, there are some 13,000
industrial, commercial or institutional point sources in the 2002 NEI,
each with one or more processes that emit Pb to the atmosphere. In
addition to these 13,000 sources, there are approximately 3,000
airports at which leaded gasoline is used (Staff Paper, p. 2-8). Among
these sources, more than one thousand are estimated to emit at least a
tenth of a ton of Pb per year (Staff Paper, Section 2.2.3). Because of
its persistence, Pb emissions contribute to media
[[Page 29191]]
concentrations for some time into the future.
---------------------------------------------------------------------------
\12\ As noted in the Staff Paper, quantitative estimates of
emissions associated with resuspension of soil-bound Pb particles
and contaminated road dust are not included in the 2002 NEI.
\13\ The emissions estimates identified as mobile sources in the
current NEI are currently limited to combustion of leaded aviation
gas in piston-engine aircraft. Lead emissions estimates for other
mobile source emissions of Pb (e.g., brake wear, tire wear, loss of
Pb wheel weights and others) are not included in the current NEI.
---------------------------------------------------------------------------
Lead emitted to the air is predominantly in particulate form, with
the particles occurring in many sizes. Once emitted, Pb particles can
be transported long or short distances depending on their size, which
influences the amount of time spent in aerosol phase. In general,
larger particles tend to deposit more quickly, within shorter distances
from emissions points, while smaller particles will remain in aerosol
phase and travel longer distances before depositing. Additionally, once
deposited, Pb particles can be resuspended back into the air and
undergo a second dispersal. Thus, the atmospheric transport processes
of Pb contribute to its broad dispersal, with larger particles
generally occurring as a greater contribution to total airborne Pb at
locations closer to the point of emission than at more distant
locations where the relative contribution from smaller particles is
greater (CD, Section 2.3.1 and p. 3-3).
Airborne concentrations of Pb in total suspended particulate matter
(Pb-TSP) in the United States have fallen substantially since the
current Pb NAAQS was set in 1978.\14\ Despite this decline, there have
still been a small number of areas, associated with large stationary
sources of Pb, that have not met the NAAQS over the past few years. The
average maximum quarterly mean concentration for the time period 2003-
2005 among source-oriented monitoring sites in the U.S. is 0.48 [mu]g/
m3, while the corresponding average for non-source-oriented
sites is 0.03 [mu]g/m3.\15\ The average and median among all
monitoring-site-specific maximum quarterly mean concentrations for this
time period are 0.17 [mu]g/m3 and 0.03 [mu]g/m3,
respectively. Coincident with the historical trend in reduction in Pb
levels, however, there has also been a substantial reduction in number
of Pb-TSP monitoring sites. As described below in section II.B.3.b,
many of the highest Pb emitting sources in the 2002 NEI do not have
nearby Pb-TSP monitors, which may lead to underestimates of the extent
of occurrences of relatively higher Pb concentrations (as recognized in
the Staff Paper, Section 2.3.2 and, with regard to more recent
analysis, in section II.B.3.b below).
---------------------------------------------------------------------------
\14\ Air Pb concentrations nationally are estimated to have
declined more than 90% since the early 1980s, in locations not known
to be directly influenced by stationary sources (Staff Paper, pp. 2-
22 to 2-23).
\15\ The data set included data for 189 monitor sites meeting
the data analysis screening criteria. Details with regard to the
data set and analyses supporting the values provided here are
presented in Section 2.3.2 of the Staff Paper.
---------------------------------------------------------------------------
2. Air-Related Human Exposure Pathways
As when the standard was set in 1978, we recognize that exposure to
air Pb can occur directly by inhalation, or indirectly by ingestion of
Pb-contaminated food, water or nonfood materials including dust and
soil (43 FR 46247). This occurs as Pb emitted into the ambient air is
distributed to other environmental media and can contribute to human
exposures via indoor and outdoor dusts, outdoor soil, food and drinking
water, as well as inhalation of air (CD, pp. 3-1 to 3-2). Accordingly,
people are exposed to Pb emitted into ambient air by both inhalation
and ingestion pathways. In general, air-related pathways include those
pathways where Pb passes through ambient air on its path from a source
to human exposure. EPA considers risks to public health from exposure
to Pb that was emitted into the air as relevant to our consideration of
the primary standard. Therefore , we consider these air-related
pathways to be policy-relevant in this review. Air-related Pb exposure
pathways include: Inhalation of airborne Pb (that may include Pb
emitted into the air and deposited and then resuspended); and ingestion
of Pb that, once airborne, has made its way into indoor dust, outdoor
dust or soil, dietary items (e.g., crops and livestock), and drinking
water (e.g., CD, Figure 3-1).
Ambient air Pb contributes to Pb in indoor dust through transport
of Pb suspended in ambient air that is then deposited indoors and
through transport of Pb that has deposited outdoors from ambient air
and is transported indoors in ways other than through ambient air (CD,
Section 3.2.3; Adgate et al., 1998). For example, infiltration of
ambient air into buildings brings airborne Pb indoors where deposition
of particles contributes to Pb in dust on indoor surfaces (CD, p. 3-28;
Caravanos et al., 2006a). Indoor dust may be ingested (e.g., via hand-
to-mouth activity by children; CD, p. 8-12) or may be resuspended
through household activities and inhaled (CD, p. 8-12). Ambient air Pb
can also deposit onto outdoor surfaces (including surface soil) with
which humans may come into contact (CD, Section 2.3.2; Farfel et al.,
2003; Caravanos et al., 2006a, b). Human contact with this deposited Pb
may result in incidental ingestion from this exposure pathway and may
also result in some of this Pb being carried indoors (e.g., on clothes
and shoes) adding to indoor dust Pb (CD, p. 3-28; von Lindern et al.,
2003a, b). Additionally, Pb from ambient air that deposits on outdoor
surfaces may also be resuspended and carried indoors in the air where
it can be inhaled. Thus, indoor dust receives air-related Pb directly
from ambient air coming indoors and also more indirectly, after
deposition from ambient air onto outdoor surfaces.
As mentioned above, humans may contact Pb in dust on outdoor
surfaces, including surface soil and other materials, that has
deposited from ambient air (CD, Section 3.2; Caravanos et al., 2006a;
Mielke et al., 1991; Roels et al., 1980). Human exposure to this
deposited Pb can occur through incidental ingestion, and, when the
deposited Pb is resuspended, by inhalation. Atmospheric deposition of
Pb also contributes to Pb in vegetation, both as a result of contact
with above ground portions of the plant and through contributions to
soil and transport of Pb into roots (CD, pp. 7-9 and AXZ7-39; USEPA,
1986a, Sections 6.5.3 and 7.2.2.2.1). Livestock may subsequently be
exposed to Pb in vegetation (e.g., grasses and silage) and in surface
soils via incidental ingestion of soil while grazing (USEPA 1986a,
Section 7.2.2.2.2). Atmospheric deposition is estimated to comprise a
significant proportion of Pb in food (CD, p. 3-48; Flegel et al., 1990;
Juberg et al., 1997; Dudka and Miller, 1999). Atmospheric deposition
outdoors also contributes to Pb in surface waters, although given the
widespread use of settling or filtration in drinking water treatment,
air-related Pb is generally a small component of Pb in treated drinking
water (CD, Section 2.3.2 and p. 3-33).
Air-related exposure pathways are affected by changes to air
quality, including changes in concentrations of Pb in air and/or
changes in atmospheric deposition of Pb. Further, because of its
persistence in the environment, Pb deposited from the air may
contribute to human and ecological exposures for years into the future
(CD, pp. 3-18 to 3-19, pp. 3-23 to 2-24). Thus, because of the roles in
human exposure pathways of both air concentration and air deposition,
and of the persistence of Pb, once deposited, some pathways respond
more quickly to changes in air quality than others. Pathways most
directly involving Pb in ambient air and exchanges of ambient air with
indoor air respond more quickly while pathways involving exposure to Pb
deposited from ambient air into the environment generally respond more
slowly (CD, pp. 3-18 to 3-19).
[[Page 29192]]
Exposure pathways tied most directly to ambient air, and that
consequently have the potential to respond relatively more quickly to
changes in air Pb, include inhalation of ambient air, and ingestion of
Pb in indoor dust directly contaminated with Pb from ambient air.\16\
Lead from ambient air contaminates indoor dust directly when outdoor
air comes inside (through open doors or windows, for example) and Pb in
that air deposits to indoor surfaces (Caravanos et al., 2006a; CD, p.
8-22). This includes Pb that was previously deposited outdoors and is
then resuspended and transported indoors. Lead in dust on outdoor
surfaces also responds to air deposition (Caravanos et al., 2006).
Pathways in which the air quality impact is reflected over a somewhat
longer time frame generally are associated with outdoor atmospheric
deposition, and include ingestion pathways such as the following: (1)
Ingestion of Pb in outdoor soil; (2) ingestion of Pb in indoor dust
indirectly contaminated with Pb from the outdoor air (e.g, ``tracking
in'' of Pb deposited to outdoor surface soil, as compared to ambient
air transport of resuspended outdoor soil); (3) ingestion of Pb in diet
that is attributable to deposited air Pb, and; (4) ingestion of Pb in
drinking water that is attributable to deposited air Pb (e.g., Pb
entering water bodies used for drinking supply).
---------------------------------------------------------------------------
\16\ We note that in the risk assessment, we only assessed
alternate standard impacts on the subset of air-related pathways
that respond relatively quickly to changes in air Pb.
---------------------------------------------------------------------------
3. Nonair-Related and Air-Related Background Human Exposure Pathways
As when the standard was set in 1978, there continue to be multiple
sources of exposure, both air-related and others (nonair-related).
Human exposure pathways that are not air-related are those in which Pb
does not pass through ambient air. These pathways as well as air-
related human exposure pathways that involve natural sources of Pb to
air are considered policy-relevant background in this review. In the
context of NAAQS for other criteria pollutants which are not multimedia
in nature, such as ozone, the term policy-relevant background is used
to distinguish anthropogenic air emissions from naturally occurring
non-anthropogenic emissions to separate pollution levels that can be
controlled by U.S. regulations from levels that are generally
uncontrollable by the United States (USEPA, 2007d). In the case of Pb,
however, due to the multimedia, multipathway nature of human exposures
to Pb, policy-relevant background is defined more broadly to include
not only the ``quite low'' levels of naturally occurring Pb emissions
into the air from non-anthropogenic sources such as volcanoes, sea
salt, and windborne soil particles from areas free of anthropogenic
activity (see below), but also Pb from nonair sources. These are
collectively referred to as ``policy-relevant background.''
The pathways of human exposure to Pb that are not air-related
include ingestion of Pb from indoor Pb paint \17\, Pb in diet as a
result of inadvertent additions during food processing, and Pb in
drinking water attributable to Pb in distribution systems (CD, Chapter
3). Other less prevalent, potential pathways of Pb exposure that are
not air-related include ingestion of some calcium supplements or of
food contaminated during storage in some Pb glazed glassware, and hand-
to-mouth contact with some imported vinyl miniblinds or with some hair
dyes containing Pb acetate, as well as some cosmetics and folk remedies
(CD, pp. 3-50 to 3-51).
---------------------------------------------------------------------------
\17\ Weathering of outdoor Pb paint may also contribute to soil
Pb levels adjacent to the house.
---------------------------------------------------------------------------
Some amount of Pb in the air derives from background sources, such
as volcanoes, sea salt, and windborne soil particles from areas free of
anthropogenic activity (CD, Section 2.2.1). The impact of these sources
on current air concentrations is expected to be quite low (relative to
current concentrations) and has been estimated to fall within the range
from 0.00002 [mu]g/m3 and 0.00007 [mu]g/m3 based
on mass balance calculations for global emissions (CD, Section 3.1 and
USEPA 1986, Section 7.2.1.1.3). The midpoint in this range, 0.00005
[mu]g/m3, has been used in the past to represent the
contribution of naturally occurring air Pb to total human exposure
(USEPA 1986, Section 7.2.1.1.3). The data available to derive such an
estimate are limited and such a value might be expected to vary
geographically with the natural distribution of Pb. Comparing this to
reported air Pb measurements is complicated by limitations of the
common analytical methods and by inconsistent reporting practices. This
value is one half the lowest reported nonzero value in AQS. Little
information is available regarding anthropogenic sources of airborne Pb
located outside of North America, which would also be considered
policy-relevant background. In considering contributions from policy-
relevant background to human exposures and associated health effects,
however, any credible estimate of policy-relevant background in air is
likely insignificant in comparison to the contributions from exposures
to nonair media.
4. Contributions to Children's Lead Exposures
As when the standard was set in 1978, EPA recognizes that there
remain today contributions to blood Pb levels from nonair sources. The
relative contribution of Pb in different exposure media to human
exposure varies, particularly for different age groups. For example,
some studies have found that dietary intake of Pb may be a predominant
source of Pb exposure among adults, greater than consumption of water
and beverages or inhalation (CD, p. 3-43).\18\ For young children,
however, ingestion of indoor dust can be a significant Pb exposure
pathway, such that dust ingested via hand-to-mouth activity can be a
more important source of Pb exposure than inhalation, although indoor
dust can also be resuspended through household activities and pose an
inhalation risk as well (CD, p. 3-27 to 3-28; Melnyk et al. 2000).\19\
---------------------------------------------------------------------------
\18\ ``Some recent exposure studies have evaluated the relative
importance of diet to other routes of Pb exposure. In reports from
the NHEXAS, Pb concentrations measured in households throughout the
Midwest were significantly higher in solid food compared to
beverages and tap water (Clayton et al., 1999; Thomas et al., 1999).
However, beverages appeared to be the dominant dietary pathway for
Pb according to the statistical analysis (Clayton et al., 1999),
possibly indicating greater bodily absorption of Pb from liquid
sources (Thomas et al., 1999). Dietary intakes of Pb were greater
than those calculated for intake from home tap water or inhalation
on a [mu]g/day basis (Thomas et al., 1999). The NHEXAS study in
Arizona showed that, for adults, ingestion was a more important Pb
exposure route than inhalation (O'Rourke et al., 1999).'' (CD, p. 3-
43)
\19\ For example, the Criteria Document states the following:
``Given the large amount of time people spend indoors, exposure to
Pb in dusts and indoor air can be significant. For children, dust
ingested via hand-to-mouth activity is often a more important source
of Pb exposure than inhalation. Dust can be resuspended through
household activities, thereby posing an inhalation risk as well.
House dust Pb can derive both from Pb-based paint and from other
sources outside the home. The latter include Pb-contaminated
airborne particles from currently operating industrial facilities or
resuspended soil particles contaminated by deposition of airborne Pb
from past emissions.'' (CD, p. E-6)
---------------------------------------------------------------------------
Estimating contributions from nonair sources is complicated by the
existence of multiple and varied air-related pathways (as described in
section II.A.2 above), as well as the persistent nature of Pb. For
example, Pb that is a soil or dust contaminant today may have been
airborne yesterday or many years ago. The studies currently available
and reviewed in the Criteria Document that evaluate the multiple
pathways of Pb exposure, when considering exposure contributions from
outdoor dust/soil, do
[[Page 29193]]
not usually distinguish between outdoor soil/dust Pb resulting from
historical emissions and outdoor soil/dust Pb resulting from recent
emissions. Further, while indoor dust Pb has been identified as being a
predominant contributor to children's blood Pb, available studies do
not generally distinguish the different pathways (air-related and
other) contributing to indoor dust Pb. The exposure assessment for
children performed for this review has employed available data and
methods to develop estimates intended to inform a characterization of
these pathways (as described in section II.C below).
Relative contributions to a child's total Pb exposure from air-
related exposure pathways (such as those identified in the sections
above) compared to other (nonair-related) Pb exposures depends on many
factors including ambient air concentrations and air deposition in the
area where the child resides (as well as in the area from which the
child's food derives), access to other sources of Pb exposure such as
Pb paint, tap water affected by plumbing containing Pb and access to
Pb-tainted products. Studies indicate that in the absence of paint-
related exposures, Pb from other sources such as stationary sources of
Pb emissions may dominate a child's Pb exposures (CD, section 3.2). In
other cases, such as children living in older housing with peeling
paint or where renovations have occurred, the dominant source may be
lead paint used in the house in the past (CD, pp. 3-50 and 3-51).
Depending on Pb levels in a home's tap water, drinking water can
sometimes be a significant source (CD, section 3.3). And in still other
cases, there may be more of a mixture of contributions from multiple
sources, with no one source dominating (CD, Chapter 3).
As recognized in sections B.1.1 and II.B.3.a, blood Pb levels are
the commonly used index of exposure for Pb and they reflect external
sources of exposure, behavioral characteristics and physiological
factors. Lead derived from differing sources or taken into the body as
a result of differing exposure pathways (e.g., air- as compared to
nonair-related), is not easily distinguished. As mentioned above,
complications to consideration of estimates of air-related or
conversely, nonair, blood Pb levels are the roles of air Pb in human
exposure pathways and the persistence of Pb in the environment. As
described in section II.A.2, air-related pathways (those in which Pb
passes through the air on its path from source to human exposure) are
varied, including inhalation and ingestion, indoor dust, outdoor dust/
soil and diet, Pb suspended in and deposited from air, and encompassing
a range of time frames from more immediate to less so. Estimates of
blood Pb levels associated with air-related exposure pathways or only
with nonair exposure pathways will vary depending on how completely the
air-related pathways are characterized.
Consistent with reductions in air Pb concentrations (as described
in section II.A.1 above) which contribute to blood Pb, nonair
contributions have also been reduced. For example, the use of Pb paint
in new houses has declined substantially over the 20th century, such
that according to the National Survey of Lead and Allergens in Housing
(USHUD, 2002) an estimated 24% of U.S. housing constructed between 1960
and 1978; 69% of the housing constructed between 1940 and 1959; and 87%
of the pre-1940 housing contains lead-based paint. Additionally, Pb
contributions to diet have been reported to have declined significantly
since 1978, perhaps as much as 70% or more between then and 1990 (WHO,
1995) and the 2006 Criteria Document identifies a drop in dietary Pb
intake by 2 to 5 year olds of 96% between the early 1980s and mid 1990s
(CD, Section 3.4 and p. 8-14).\20\ These reductions are generally
attributed to reductions in gasoline-related airborne Pb as well as the
reduction in use of Pb solder in canning food products (CD, Section
3.4).\21\ There have also been reductions in tap water Pb levels (CD,
section 3.3 and pp. 8-13 to 8-14). Contamination from the distribution/
plumbing system appears to remain the predominant source of Pb in the
drinking water (CD, section 3.3 and pp. 8-013 to 8-14).
---------------------------------------------------------------------------
\20\ Additionally, the 1977 Criteria Document included a dietary
Pb intake estimate for the general population of 100 to 350 [mu]g
Pb/day, with estimates near and just below 100 [mu]g/day for young
children (USEPA 1977, pp. 1-2 and 12-32) and the 2006 Criteria
Document cites recent studies (for the mid-1990s) indicating a
dietary intake ranging from 2 to 10 [mu]g Pb/day for children (CD,
Section 3.4 and p. 8-14).
\21\ Sources of Pb in food were identified in the 1986 Criteria
Document as including air-related sources, metals used in processing
raw foodstuffs, solder used in packaging and water used in cooking
(1986a, section 3.1.2).
---------------------------------------------------------------------------
The availability of estimates of blood Pb levels resulting only
from air-related sources and exposures or only from those unrelated to
air is limited and, given the discussion above, would be expected to
vary for different populations. In addition to potential differences in
air-related and nonair-related blood Pb levels among populations with
different exposure circumstances (e.g., relatively more or lesser
exposure to air-related Pb), the absolute levels may also vary among
different age groups. As described in section II.B.1.b, average total
blood Pb levels in the U.S. differ among age groups, with levels being
highest in children aged one to five years old. We also note that
behavioral characteristics that influence Pb exposures vary among age
groups. For example as noted above, the predominant Pb exposure
pathways may differ between adults and children. The extent of any
quantitative impact of these differences on estimates of nonair blood
Pb levels is unknown.\22\
---------------------------------------------------------------------------
\22\ As noted earlier in this section, for children, dust
ingestion by hand-to-mouth activity can be an important source of Pb
exposure, while for adults, dietary Pb can be predominant.
---------------------------------------------------------------------------
In their advice to the Agency on levels for the standard, the CASAC
Pb Panel explored several approaches to deriving a level, one of which
required an estimate of the nonair component of blood Pb for the
average child. They recommended consideration of 1.0 to 1.4 [mu]g/dL or
lower for such an estimate for the average nonair blood Pb level for
young children (Henderson, 2007a, p. D-1). This range was developed
with consideration of simulations of the integrated exposure and uptake
biokinetic (IEUBK) model for lead for which the exposure concentration
inputs included zero air concentration and concentrations for soil and
dust of 50 ppm and 35 ppm, respectively (Henderson, 2007a, p. F-
60).\23\ \24\ \25\
---------------------------------------------------------------------------
\23\ The soil and dust levels are described as ``typical
geochemical non-air input levels for dust and soil'' (Henderson,
2007a, p. F-60). The values used for these levels in this simulation
fall within the range of 1 to 200 ppm described in the Criteria
Document for soil not influenced by sources (CD, p. 3-18).
\24\ The other IEUBK inputs (e.g., exposure and biokinetic
factors) were those used in the IEUBK modeling for the risk
assessment in this review (Henderson, 2007a, p. F-60).
\25\ Individual CASAC member comments describing the IEUBK
simulations stated that the modeling produced a nonair blood Pb
level of ``1.4 [mu]g/dL as a geometric mean'' (Henderson, 2007a, p.
F-61).
---------------------------------------------------------------------------
As is evident from the prior discussion, the many different
exposure pathways contributing to children's blood Pb levels, and other
factors, complicate our consideration of the available data with regard
to characterization of levels particular to specific pathways, air-
related or otherwise.
B. Health Effects Information
The following summary focuses on health endpoints associated with
the range of exposures considered to be most relevant to current
exposure levels and makes note of several key aspects of the health
evidence for Pb. First (as
[[Page 29194]]
described in Section II.A, above), because exposure to atmospheric Pb
particles occurs not only via direct inhalation of airborne particles,
but also via ingestion of deposited ambient Pb, the exposure considered
is multimedia and multipathway in nature, occurring via both the
inhalation and ingestion routes. Second, the exposure index or dose
metric most commonly used and associated with health effects
information is an internal biomarker (i.e., blood Pb). Additionally,
the exposure duration of interest (i.e., that influencing internal dose
pertinent to health effects of interest) may span months to potentially
years, as does the time scale of the environmental processes
influencing Pb deposition and fate. Lastly, the nature of the evidence
for the health effects of greatest interest for this review,
neurological effects, particularly neurocognitive and neurobehavioral
effects, in young children, are epidemiological data substantiated by
toxicological data that provide biological plausibility and insights on
mechanisms of action (CD, sections 5.3, 6.2 and 8.4.2).
In recognition of the multi-pathway aspects of Pb, and the use of
an internal exposure metric in health risk assessment, the next section
describes the internal disposition or distribution of Pb, and the use
of blood Pb as an internal exposure or dose metric. This is followed by
a discussion of the nature of Pb-induced health effects that emphasizes
those with the strongest evidence. Potential impacts of Pb exposures on
public health, including recognition of potentially susceptible or
vulnerable subpopulations, are then discussed. Finally, key
observations about Pb-related health effects are summarized.
1. Blood Lead
The health effects of Pb are remote from the portals of entry to
the body (i.e., the respiratory system and gastrointestinal tract).
Consequently, the internal disposition and distribution of Pb in the
blood is an integral aspect of the relationship between exposure and
effect. Additionally, the focus on blood Pb as the dose metric in
consideration of the Pb health effects evidence, while reducing our
uncertainty with regard to causality, leads to an additional
consideration with regard to contribution of air-related sources and
exposure pathways to blood Pb.
a. Internal Disposition of Lead
This section briefly summarizes the current state of knowledge of
Pb disposition pertaining to both inhalation and ingestion routes of
exposure as described in the Criteria Document.
Inhaled Pb particles deposit in the different regions of the
respiratory tract as a function of particle size (CD, pp. 4-3 to 4-4).
Lead associated with smaller particles, which are predominantly
deposited in the pulmonary region, may, depending on solubility, be
absorbed into the general circulation or transported to the
gastrointestinal tract (CD, pp. 4-3). Lead associated with larger
particles, which are predominantly deposited in the head and conducting
airways (e.g., nasal pharyngeal and tracheobronchial regions of
respiratory tract), may be transported into the esophagus and
swallowed, thus making its way to the gastrointestinal tract (CD, pp.
4-3 to 4-4), where it may be absorbed into the blood stream. Thus, Pb
can reach the gastrointestinal tract either directly through the
ingestion route or indirectly following inhalation.
Once in the blood stream, where approximately 99% of the Pb
associates with red blood cells, the Pb is quickly distributed
throughout the body (e.g., within days) with the bone serving as a
large, long-term storage compartment, and soft tissues (e.g., kidney,
liver, brain, etc.) serving as smaller compartments, in which Pb may be
more mobile (CD, sections 4.3.1.4 and 8.3.1.). Additionally, the
epidemiologic evidence indicates that Pb freely crosses the placenta
resulting in continued fetal exposure throughout pregnancy, and that
exposure increases during the later half of pregnancy (CD, section
6.6.2).
During childhood development, bone represents approximately 70% of
a child's body burden of Pb, and this accumulation continues through
adulthood, when more than 90% of the total Pb body burden is stored in
the bone (CD, section 4.2.2). Accordingly, levels of Pb in bone are
indicative of a person's long-term, cumulative exposure to Pb. In
contrast, blood Pb levels are usually indicative of recent exposures.
Depending on exposure dynamics, however, blood Pb may--through its
interaction with bone--be indicative of past exposure or of cumulative
body burden (CD, section 4.3.1.5).
Throughout life, Pb in the body is exchanged between blood and
bone, and between blood and soft tissues (CD, section 4.3.2), with
variation in these exchanges reflecting ``duration and intensity of the
exposure, age and various physiological variables'' (CD, p. 4-1). Past
exposures that contribute Pb to the bone, consequently, may influence
current levels of Pb in blood. Where past exposures were elevated in
comparison to recent exposures, this influence may complicate
interpretations with regard to recent exposure (CD, sections 4.3.1.4 to
4.3.1.6). That is, higher blood Pb concentrations may be indicative of
higher cumulative exposures or of a recent elevation in exposure (CD,
pp. 4-34 and 4-133).
In several studies investigating the relationship between Pb
exposure and blood Pb in children (e.g., Lanphear and Roghmann 1997;
Lanphear et al., 1998), blood Pb levels have been shown to reflect Pb
exposures, with particular influence associated with exposures to Pb in
surface dust. Further, as stated in the Criteria Document ``these and
other studies of populations near active sources of air emissions
(e.g., smelters, etc.) substantiate the effect of airborne Pb and
resuspended soil Pb on interior dust and blood Pb'' (CD, p. 8-22).
b. Use of Blood Lead as Dose Metric
Blood Pb levels are extensively used as an index or biomarker of
exposure by national and international health agencies, as well as in
epidemiological (CD, sections 4.3.1.3 and 8.3.2) and toxicological
studies of Pb health effects and dose-response relationships (CD,
Chapter 5). The prevalence of the use of blood Pb as an exposure index
or biomarker is related to both the ease of blood sample collection
(CD, p. 4-19; Section 4.3.1) and by findings of association with a
variety of health effects (CD, Section 8.3.2). For example, the U.S.
Centers for Disease Control and Prevention (CDC), and its predecessor
agencies, have for many years used blood Pb level as a metric for
identifying children at risk of adverse health effects and for
specifying particular public health recommendations (CDC, 1991; CDC,
2005a). In 1978, when the current Pb NAAQS was established, the CDC
recognized a blood Pb level of 30 [mu]g/dL as a level warranting
individual intervention (CDC, 1991). In 1985, the CDC recognized a
level of 25 [mu]g/dL for individual child intervention, and in 1991,
they recognized a level of 15 [mu]g/dL for individual intervention and
a level of 10 [mu]g/dL for implementing community-wide prevention
activities (CDC, 1991; CDCa, 2005). In 2005, with consideration of a
review of the evidence by their advisory committee, CDC revised their
statement on Preventing Lead Poisoning in Young Children, specifically
recognizing the evidence of adverse health effects in children with
blood Pb levels below 10 [mu]g/dL \26\ and the data demonstrating that
[[Page 29195]]
no ``safe'' threshold for blood Pb had been identified, and emphasizing
the importance of preventative measures (CDC, 2005a, ACCLPP, 2007).\27\
---------------------------------------------------------------------------
\26\ As described by the Advisory Committee on Childhood Lead
Poisoning Prevention, ``In 1991, CDC defined the blood lead level
(BLL) that should prompt public health actions as 10 [mu]g/dL.
Concurrently, CDC also recognized that a BLL of 10 [mu]g/dL did not
define a threshold for the harmful effects of lead. Research
conducted since 1991 has strengthened the evidence that children's
physical and mental development can be affected at BLLS <10 [mu]g/
dL'' (ACCLPP, 2007).
\27\ With the 2005 statement, CDC did not lower the 1991 level
of concern and identified a variety of reasons, reflecting both
scientific and practical considerations, for not doing so, including
a lack of effective clinical or public health interventions to
reliably and consistently reduce blood Pb levels that are already
below 10 [mu]g/dL, the lack of a demonstrated threshold for adverse
effects, and concerns for deflecting resources from children with
higher blood Pb levels (CDC, 2005a). CDC's Advisory Committee on
Childhood Lead Poisoning Prevention recently provided
recommendations regarding interpreting and managing blood Pb levels
below 10 [mu]g/dL in children and reducing childhood exposures to Pb
(ACCLPP, 2007).
---------------------------------------------------------------------------
Since 1976, the CDC has been monitoring blood Pb levels nationally
through the National Health and Nutrition Examination Survey (NHANES).
This survey monitors blood Pb levels in multiple age groups in the U.S.
This information indicates variation in mean blood Pb levels across the
various age groups monitored. For example, mean values in 2001-2002 for
ages 1-5, 6-11, 12-19 and greater than or equal to 20 years of age, are
1.70, 1.25, 0.94, and 1.56, respectively (CD, p. 4-22).
The NHANES information has documented the dramatic decline in mean
blood Pb levels in the U.S. population that has occurred since the
1970s and that coincides with regulations regarding leaded fuels,
leaded paint, and Pb-containing plumbing materials that have reduced Pb
exposure among the general population (CD, Sections 4.3.1.3 and 8.3.3;
Schwemberger et al., 2005). The Criteria Document summarizes related
information as follows (CD, p. E-6).
In the United States, decreases in mobile sources of Pb,
resulting from the phasedown of Pb additives created a 98% decline
in emissions from 1970 to 2003. NHANES data show a consequent
parallel decline in blood-Pb levels in children aged 1 to 5 years
from a geometric mean of ~15 [mu]g/dL in 1976-1980 to ~1-2 [mu]g/dL
in the 2000-2004 period.
While levels in the U.S. general population, including geometric mean
levels in children aged 1-5, have declined significantly, levels have
been found to vary among children of different socioeconomic status
(SES) and other demographic characteristics (CD, p. 4-21). For example,
while the 2001-2004 median blood level for children aged 1-5 of all
races and ethnic groups is 1.6 [mu]g/dL, the median for the subset
living below the poverty level is 2.3 [mu]g/dL and 90th percentile
values for these two groups are 4.0 [mu]g/dL and 5.4 [mu]g/dL,
respectively. Similarly, the 2001-2004 median blood level for black,
non-Hispanic children aged 1-5 is 2.5 [mu]g/dL, while the median level
for the subset of that group living below the poverty level is 2.9
[mu]g/dL and the median level for the subset living in more well-off
households (i.e., with income more than 200% of the poverty level) is
1.9 [mu]g/dL. Associated 90th percentile values for 2001-2004 are 6.4
[mu]g/dL (for black, non-Hispanic children aged 1-5), 7.7 [mu]g/dL (for
the subset of that group living below the poverty level) and 4.1 [mu]g/
dL (for the subset living in a household with income more than 200% of
the poverty level).\28\ The recently released RRP rule (discussed above
in section I.C) is expected to contribute to further reductions in BLL
for children living in houses with Pb paint.
---------------------------------------------------------------------------
\28\ This information is available at: http://www.epa.gov/
envirohealth/children/body_burdens/b1-table.htm (click on
``Download a universal spreadsheet file of the Body Burdens data
tables'').
Bone measurements, as a result of the generally slower Pb turnover
in bone, are recognized as providing a better measure of cumulative Pb
exposure (CD, Section 8.3.2). The bone pool of Pb in children, however,
is thought to be much more labile than that in adults due to the more
rapid turnover of bone mineral as a result of growth (CD, p. 4-27). As
a result, changes in blood Pb concentration in children more closely
parallel changes in total body burden (CD, pp. 4-20 and 4-27). This is
in contrast to adults, whose bone has accumulated decades of Pb
exposures (with past exposures often greater than current ones), and
for whom the bone may be a significant source long after exposure has
ended (CD, Section 4.3.2.5).
c. Air-to-Blood Relationships
As described in Section II.A, Pb in ambient air contributes to Pb
in blood by multiple pathways, with the pertinent exposure routes
including both inhalation and ingestion (CD, Sections 3.1.3.2, 4.2 and
4.4; Hilts, 2003). The quantitative relationship between ambient air Pb
and blood Pb, which is often termed a slope or ratio, describes the
increase in blood Pb (in [mu]g/dL) per unit of air Pb (in [mu]g/m
\3\).\29\
---------------------------------------------------------------------------
\29\ Ratios are presented in the form of 1:x, with the 1
representing air Pb (in [mu]g/m\3\) and x representing blood Pb (in
[mu]g/dL). Description of ratios as higher or lower refers to the
values for x (i.e., the change in blood Pb per unit of air Pb).
Slopes are presented as simply the value of x.
---------------------------------------------------------------------------
The evidence on this quantitative relationship is now, as in the
past, limited by the circumstances in which the data are collected.
These estimates are generally developed from studies of populations in
various Pb exposure circumstances. The 1986 Criteria Document discussed
the studies available at that time that addressed the relationship
between air Pb and blood Pb,\30\ recognizing that there is significant
variability in air-to-blood ratios for different populations exposed to
Pb through different air-related exposure pathways and at different
exposure levels.
---------------------------------------------------------------------------
\30\ We note that the 2006 Criteria Document did not include a
discussion of more recent studies on air-to-blood ratios.
---------------------------------------------------------------------------
In discussing the available evidence, the 1986 Criteria Document
observed that estimates of air-to-blood ratios that included air-
related ingestion pathways in addition to the inhalation pathway are
``necessarily higher'' (in terms of blood Pb response) than those
estimates based on inhalation alone (USEPA 1986a, p. 11-106). Thus, the
extent to which studies account for the full set of air-related
exposure pathways affects the magnitude of the resultant air-to-blood
estimates, such that fewer pathways included as ``air-related'' yield
lower ratios. The 1986 Criteria Document also observed that ratios
derived from studies focused only on inhalation pathways (e.g., chamber
studies, occupational studies) have generally been on the order of 1:2
or lower, while ratios derived from studies including more air-related
pathways were generally higher (USEPA, 1986a, p. 11-106). Further, the
current evidence appears to indicate higher ratios for children as
compared to those for adults (USEPA, 1986a), perhaps due to behavioral
differences between the age groups.
Reflecting these considerations, the 1986 Criteria Document
identified a range of air-to-blood ratios for children that reflected
both inhalation and ingestion-related air Pb contributions as generally
ranging from 1:3 to 1:5 based on the information available at that time
(USEPA 1986a, p. 11-106). Table 11-36 (p. 11-100) in the 1986 Criteria
Document (drawn from Table 1 in Brunekreef, 1984) presents air-to-blood
ratios from a number of studies in children (i.e., those with
identified air monitoring methods and reliable blood Pb data). For
example, air-to-blood ratios from the subset of those studies that used
quality control protocols and presented adjusted slopes \31\ include
[[Page 29196]]
adjusted ratios of 3.6 (Zielhuis et al., 1979), 5.2 (Billick et al.,
1979, 1980), 2.9 (Billick et al., 1983), and 8.5 (Brunekreef et al,
1983).
---------------------------------------------------------------------------
\31\ Brunekreef et al. (1984) discusses potential confounders to
the relationship between air Pb and blood Pb, recognizing that
ideally all possible confounders should be taken into account in
deriving an adjusted air-to-blood relationship from a community
study. The studies cited here adjusted for parental education
(Zielhuis et al., 1979), age and race (Billick et al., 1979, 1980)
and additionally measuring height of air Pb (Billick et al., 1983);
Brunekreef et al. (1984) used multiple regression to control for
several confounders. The authors conclude that ``presentation of
both unadjusted and (stepwise) adjusted relationships is advisable,
to allow insight in the range of possible values for the
relationship'' (p. 83). Unadjusted ratios were presented for two of
these studies, including ratios of 4.0 (Zielhuis et al., 1979) and
18.5 (Brunekreef et al., 1983). Note, that the Brunekreef et al.,
1983 study is subject to a number of sources of uncertainty that
could result in air-to-blood Pb ratios that are biased high,
including the potential for underestimating ambient air Pb levels
due to the use of low volume British Smoke air monitors and the
potential for ongoing (higher historical) ambient air Pb levels to
have influenced blood Pb levels (see Section V.B.2 of the 1989 Pb
Staff Report for the Pb NAAQS review, EPA, 1989). In addition, the
1989 Staff Report notes that the higher air-to-blood ratios obtained
from this study could reflect the relatively lower blood Pb levels
seen across the study population (compared with blood Pb levels
reported in other studies from that period).
---------------------------------------------------------------------------
Additionally, the 1986 Criteria Document noted that ratios derived
from studies involving higher blood and air Pb levels are generally
smaller than ratios from studies involving lower blood and air Pb
levels (USEPA, 1986a. p. 11-99). In consideration of this factor, we
note that the range of 1:3 to 1:5 in air-to-blood ratios for children
noted in the 1986 Criteria Document generally reflected study
populations with blood Pb levels in the range of approximately 10-30
[mu]g/dL (USEPA 1986a, pp. 11-100; Brunekreef, 1984), much higher than
those common in today's population. This observation suggests that air-
to-blood ratios relevant for today's population of children would
likely extend higher than the 1:3 to 1:5 range identified in the 1986
Criteria Document.
More recently, a study of changes in children's blood Pb levels
associated with reduced Pb emissions and associated air concentrations
near a Pb smelter in Canada (for children through six years of age)
reports a ratio of 1:6 and additional analysis of the data by EPA for
the initial time period of the study resulted in a ratio of 1:7 (CD,
pp. 3-23 to 3-24; Hilts, 2003).\32\ Ambient air and blood Pb levels
associated with the Hilts (2003) study range from 1.1 to 0.03 [mu]g/
m\3\, and associated population mean blood Pb levels range from 11.5 to
4.7 [mu]g/dL, which are lower than levels associated with the older
studies cited in the 1986 Criteria Document (USEPA, 1986).
---------------------------------------------------------------------------
\32\ This study considered changes in ambient air Pb levels and
associated blood Pb levels over a five-year period which included
closure of an older Pb smelter and subsequent opening of a newer
facility in 1997 and a temporary (3 month) shutdown of all smelting
activity in the summer of 2001. The author observed that the air-to-
blood ratio for children in the area over the full period was
approximately 1:6. The author noted limitations in the dataset
associated with exposures in the second time period, after the
temporary shutdown of the facility in 2001, including sampling of a
different age group at that time and a shorter time period (3
months) at these lower ambient air Pb levels prior to collection of
blood Pb levels. Consequently, EPA calculated an alternate air-to-
blood Pb ratio based on consideration for ambient air Pb and blood
Pb reductions in the first time period (after opening of the new
facility in 1997).
---------------------------------------------------------------------------
Sources of uncertainty related to air-to-blood ratios obtained from
Hilts (2003) study have been identified. One such area of uncertainty
relates to the pattern of changes in indoor Pb dustfall (presented in
Table 3 in the article) which suggests a potentially significant
decrease in Pb impacts to indoor dust prior to closure of an older Pb
smelter and start-up of a newer facility in 1997. Some have suggested
that this earlier reduction in indoor dustfall suggests that a
significant portion of the reduction in Pb exposure (and therefore, the
blood Pb reduction reflected in air-to-blood ratios) may have resulted
from efforts to increase public awareness of the Pb contamination issue
(e.g., through increased cleaning to reduce indoor dust levels) rather
than reductions in ambient air Pb and associated indoor dust Pb
contamination. In addition, notable fluctuations in blood Pb levels
observed prior to 1997 (as seen in Figure 2 of the article) have raised
questions as to whether factors other than ambient air Pb reduction
could be influencing decreases in blood Pb.\33\
---------------------------------------------------------------------------
\33\ In the publication, the author acknowledges that remedial
programs (e.g., community and home-based dust control and education)
may have been responsible for some of the blood Pb reduction seen
during the study period (1997 to 2001). However, the author points
out that these programs were in place in 1992 and he suggests that
it is unlikely that they contributed to the sudden drop in blood Pb
levels occurring after 1997. In addition, the author describes a
number of aspects of the analysis, which could have implications for
air-to-blood ratios including a tendency over time for children with
lower blood Pb levels to not return for testing, and inclusion of
children aged 6 to 36 months in Pb screening in 2001 (in contrast to
the wider age range up to 60 months as was done in previous years).
---------------------------------------------------------------------------
In addition to the study by Hilts (2003), we are aware of two other
studies published since the 1986 Criteria Document that report air-to-
blood ratios for children (Tripathi et al., 2001 and Hayes et al.,
1994). These studies were not cited in the 2006 Criteria Document, but
were referenced in public comments received by EPA during this
review.\34\ The study by Tripathi et al. (2001) reports an air-to-blood
ratio of approximately 1:3.6 for an analysis of children aged six
through ten in India. The ambient air and blood Pb levels in this study
(geometric mean blood Pb levels generally ranged from 10 to 15 [mu]g/
dL) are similar to levels reported in older studies reviewed in the
1986 Criteria Document and are much higher than current conditions in
the U.S. The study by Hayes (1994) compared patterns of ambient air Pb
reductions and blood Pb reductions for large numbers of children in
Chicago between 1971 and 1988, a period when significant reductions
occurred in both measures. The study reports an air-to-blood ratio of
1:5.6 associated with ambient air Pb levels near 1 [mu]g/m\3\ and a
ratio of 1:16 for ambient air Pb levels in the range of 0.25 [mu]g/
m\3\, indicating a pattern of higher ratios with lower ambient air Pb
and blood Pb levels consistent with conclusions in the 1986 Criteria
Document.\35\
---------------------------------------------------------------------------
\34\ EPA is not basing its proposed decisions on these two
studies, but notes that these estimates are consistent with other
studies that were included in the 1986 and 2006 Criteria Documents
and accordingly considered by CASAC and the public.
\35\ As with all studies, we note that there are strengths and
limitations for these two studies which may affect the specific
magnitudes of the reported ratios, but that the studies' findings
and trends are generally consistent with the conclusions from the
1986 Criteria Document.
---------------------------------------------------------------------------
In their advice to the Agency, CASAC identified air-to-blood ratios
of 1:5, as used by the World Health Organization (2000), and 1:10, as
supported by an empirical analysis of changes in air Pb and changes in
blood Pb between 1976 and the time when the phase-out of Pb from
gasoline was completed (Henderson, 2007a).\36\
---------------------------------------------------------------------------
\36\ The CASAC Panel stated ``The Schwartz and Pitcher analysis
showed that in 1978, the midpoint of the National Health and
Nutrition Examination Survey (NHANES) II, gasoline Pb was
responsible for 9.1 [mu]g/dL of blood Pb in children. Their estimate
is based on their coefficient of 2.14 [mu]g/dL per 100 metric tons
(MT) per day of gasoline use, and usage of 426 MT/day in 1976.
Between 1976 and when the phase-out of Pb from gasoline was
completed, air Pb concentrations in U.S. cities fell a little less
than 1 [mu]g/m\3\ (24). These two facts imply a ratio of 9-10 [mu]g/
dL per [mu]g/m\3\ reduction in air Pb, taking all pathways into
account.'' (Henderson, 2007a, pp. D-2 to D-3).
---------------------------------------------------------------------------
Beyond considering the evidence presented in the published
literature and that reviewed in Pb Criteria Documents, we have also
considered air-to-blood ratios derived from the exposure assessment for
this review (discussed below in section II.C). In that assessment,
current modeling tools and information on children's activity patterns,
behavior and physiology (e.g., CD, Section 4.4) were used to estimate
blood Pb levels associated with
[[Page 29197]]
multimedia and multipathway Pb exposure. The results from the various
case studies included in this assessment, with consideration of the
context in which they were derived (e.g., the extent to which the range
of air-related pathways were simulated), are also informative to our
understanding of air-to-blood ratios.
For the general urban case study, air-to-blood ratios ranged from
1:2 to 1:9 across the alternative standard levels assessed, which
ranged from the current standard of 1.5 [mu]g/m\3\ down to a level of
0.02 [mu]g/m\3\. This pattern of model-derived ratios generally
supports the range of ratios obtained from the literature and also
supports the observation that lower ambient air Pb levels are
associated with higher air-to-blood ratios. There are a number of
sources of uncertainty associated with these model-derived ratios. The
hybrid indoor dust Pb model, which is used in estimating indoor dust Pb
levels for the urban case studies, uses a HUD dataset reflecting
housing constructed before 1980 in establishing the relationship
between dust loading and concentration, which is a key component in the
hybrid dust model (see Section Attachment G-1 of the Risk Assessment,
Volume II). Given this application of the HUD dataset, there is the
potential that the non-linear relationship between indoor dust Pb
loading and concentration (which is reflected in the structure of the
hybrid dust model) could be driven more by the presence of indoor Pb
paint than contributions from outdoor ambient air Pb. We also note that
only recent air pathways were adjusted in modeling the impact of
ambient air Pb reductions on blood Pb levels in the urban case studies,
which could have implications for the air-to-blood ratios.
For the primary Pb smelter (subarea) case study, air-to-blood
ratios ranged from 1:10 to 1:19 across the same range of alternative
standard levels, from 1.5 down to 0.02 [mu]g/m\3\.\37\ Because these
ratios are based on regression modeling developed using empirical data,
there is the potential for these ratios to capture more fully the
impact of ambient air on indoor dust Pb (and ultimately blood Pb),
including longer timeframe impacts resulting from changes in outdoor
deposition. Therefore, given that these ratios are higher than ratios
developed for the general urban case study using the hybrid indoor dust
Pb model (which only considers reductions in recent air), the ratios
estimated for the primary Pb smelter (subarea) support the evidence-
based observation discussed above that consideration of more of the
exposure pathways relating ambient air Pb to blood Pb, may result in
higher air-to-blood Pb ratios. In considering this case study, some
have suggested, however, that the regression modeling fails to
accurately reflect the temporal relationship between reductions in
ambient air Pb and indoor dust Pb, which could result in an over-
estimate of the degree of dust Pb reduction associated with a specified
degree of ambient air Pb reduction, which in turn could produce air-to-
blood Pb ratios that are biased high.
---------------------------------------------------------------------------
\37\ As noted below in section II.C.3.a, air-to-blood ratios for
the primary Pb smelter (full study area) range from 1:3 to 1:7
across the same range of alternative standard levels (from 1.5 down
to 0.02 [mu]g/m\3\).
---------------------------------------------------------------------------
In summary, in EPA's view, the current evidence in conjunction with
the results and observations drawn from the exposure assessment,
including related uncertainties, supports consideration of a range of
air-to-blood ratios for children ranging from 1:3 to 1:7, reflecting
multiple air-related pathways beyond simply inhalation and the lower
air and blood Pb levels pertinent to this review. In light of the
uncertainties that remain in the available information on air-to-blood
ratios, EPA requests comment on this range and on the appropriate
weight to place on specific ratios within this range.
2. Nature of Effects
a. Broad Array of Effects
Lead has been demonstrated to exert ``a broad array of deleterious
effects on multiple organ systems via widely diverse mechanisms of
action'' (CD, p. 8-24 and Section 8.4.1). This array of health effects
includes effects on heme biosynthesis and related functions;
neurological development and function; reproduction and physical
development; kidney function; cardiovascular function; and immune
function. The weight of evidence varies across this array of effects
and is comprehensively described in the Criteria Document. There is
also some evidence of Pb carcinogenicity, primarily from animal
studies, together with limited human evidence of suggestive
associations (CD, Sections 5.6.2, 6.7, and 8.4.10).\38\
---------------------------------------------------------------------------
\38\ Lead has been classified as a probable human carcinogen by
the International Agency for Research on Cancer, based mainly on
sufficient animal evidence, and as reasonably anticipated to be a
human carcinogen by the U.S. National Toxicology Program (CD,
Section 6.7.2). U.S. EPA considers Pb a probable carcinogen (http://
www.epa.gov/iris/subst/0277.htm; CD, p. 6-195).
---------------------------------------------------------------------------
This review is focused on those effects most pertinent to ambient
exposures, which given the reductions in ambient Pb levels over the
past 30 years, are generally those associated with individual blood Pb
levels in children and adults in the range of 10 [mu]g/dL and lower.
Tables 8-5 and 8-6 in the Criteria Document highlight the key such
effects observed in children and adults, respectively (CD, pp. 8-60 to
8-62). The effects include neurological, hematological and immune
effects for children, and hematological, cardiovascular and renal
effects for adults. As evident from the discussions in Chapters 5, 6
and 8 of the Criteria Document, ``neurotoxic effects in children and
cardiovascular effects in adults are among those best substantiated as
occurring at blood Pb concentrations as low as 5 to 10 [mu]g/dL (or
possibly lower); and these categories are currently clearly of greatest
public health concern'' (CD, p. 8-60).\39\ The toxicological and
epidemiological information available since the time of the last review
``includes assessment of new evidence substantiating risks of
deleterious effects on certain health endpoints being induced by
distinctly lower than previously demonstrated Pb exposures indexed by
blood Pb levels extending well below 10 [mu]g/dL in children and/or
adults'' (CD, p. 8-25). Some health effects associated with individual
blood Pb levels extend below 5 [mu]g/dL, and some studies have observed
these effects at the lowest blood levels considered.
---------------------------------------------------------------------------
\39\ With regard to blood Pb levels in individual children
associated with particular neurological effects, the Criteria
Document states ``Collectively, the prospective cohort and cross-
sectional studies offer evidence that exposure to Pb affects the
intellectual attainment of preschool and school age children at
blood Pb levels <10 [mu]g/dL (most clearly in the 5 to 10 [mu]g/dL
range, but, less definitively, possibly lower).'' (p. 6-269)
---------------------------------------------------------------------------
With regard to population mean levels, the Criteria Document points
to studies reporting ``Pb effects on the intellectual attainment of
preschool and school age children at population mean concurrent blood-
Pb levels ranging down to as low as 2 to 8 [mu]g/dL'' (CD, p. E-9).
We note that many studies over the past decade have, in
investigating effects at lower blood Pb levels, utilized the CDC
advisory level for individual children (10 [mu]g/dL) as a benchmark for
assessment, and this is reflected in the numerous references in the
Criteria Document to 10 [mu]g/dL. Individual study conclusions stated
with regard to effects observed below 10 [mu]g/dL are usually referring
to individual blood Pb levels. In fact, many such study groups have
been restricted to individual blood Pb levels below 10 [mu]g/dL or
below levels lower than 10 [mu]g/dL. We note that the
[[Page 29198]]
mean blood Pb level for these groups will necessarily be lower than the
blood Pb level they are restricted below.
Threshold levels, in terms of blood Pb levels in individual
children, for neurological effects cannot be discerned from the
currently available studies (CD, pp. 8-60 to 8-63). The Criteria
Document states ``There is no level of Pb exposure that can yet be
identified, with confidence, as clearly not being associated with some
risk of deleterious health effects'' (CD, p. 8-63). As discussed in the
Criteria Document, ``a threshold for Pb neurotoxic effects may exist at
levels distinctly lower than the lowest exposures examined in these
epidemiologic studies'' (CD, p. 8-67).\40\
---------------------------------------------------------------------------
\40\ In consideration of the evidence from experimental animal
studies with regard to the issue of threshold for neurotoxic
effects, the CD notes that there is little evidence that allows for
clear delineation of a threshold, and that ``blood-Pb levels
associated with neurobehavioral effects appear to be reasonably
parallel between humans and animals at reasonably comparable blood-
Pb concentrations; and such effects appear likely to occur in humans
ranging down at least to 5-10 [mu]g/dL, or possibly lower (although
the possibility of a threshold for such neurotoxic effects cannot be
ruled out at lower blood-Pb concentrations)'' (CD, p. 8-38).
---------------------------------------------------------------------------
In summary, the Agency has identified neurological, hematological
and immune effects in children and neurological, hematological,
cardiovascular and renal effects in adults as the effects observed at
blood Pb levels near or below 10 [mu]g/dL and further considers
neurological effects in children and cardiovascular effects in adults
to be categories of effects that ``are currently clearly of greatest
public health concern'' (CD, pp. 8-60 to 8-62). Neurological effects in
children are discussed further below.
b. Neurological Effects in Children
Among the wide variety of health endpoints associated with Pb
exposures, there is general consensus that the developing nervous
system in young children is among, if not, the most sensitive. As
described in the Criteria Document, neurotoxic effects in children and
cardiovascular effects in adults are categories of effects that are
``currently clearly of greatest public health concern'' (CD, p. 8-
60).\41\ While also recognizing the occurrence of adult cardiovascular
effects at somewhat similarly low blood Pb levels \42\, neurological
effects in children are considered to be the sentinel effects in this
review and are the focus of the quantitative risk assessment conducted
for this review (discussed below in section III.C).
---------------------------------------------------------------------------
\41\ The Criteria Document states ``neurotoxic effects in
children and cardiovascular effects in adults are among those best
substantiated as occurring at blood-Pb concentrations as low as 5 to
10 [mu]g/dL (or possibly lower); and these categories of effects are
currently clearly of greatest public health concern (CD, p. 8-60).''
\42\ For example, the Criteria Document describes associations
of blood Pb in adults with blood pressure in studies with population
mean blood Pb levels ranging from approximately 2 to 6 [mu]g/dL (CD,
section 6.5.2 and Table 6-2).
---------------------------------------------------------------------------
The nervous system has long been recognized as a target of Pb
toxicity, with the developing nervous system affected at lower
exposures than the mature system (CD, Sections 5.3, 6.2.1, 6.2.2, and
8.4). While blood Pb levels in U.S. children ages one to five years
have decreased notably since the late 1970s, newer studies have
investigated and reported associations of effects on the
neurodevelopment of children with these more recent blood Pb levels
(CD, Chapter 6). Functional manifestations of Pb neurotoxicity during
childhood include sensory, motor, cognitive and behavioral impacts.
Numerous epidemiological studies have reported neurocognitive,
neurobehavioral, sensory, and motor function effects in children with
blood Pb levels below 10 [mu]g/dL (CD, Sections 6.2 and 8.4). \43\ As
discussed in the Criteria Document, ``extensive experimental laboratory
animal evidence has been generated that (a) substantiates well the
plausibility of the epidemiologic findings observed in human children
and adults and (b) expands our understanding of likely mechanisms
underlying the neurotoxic effects'' (CD, p. 8-25; Section 5.3).
---------------------------------------------------------------------------
\43\ Further, neurological effects in general include behavioral
effects, such as delinquent behavior (CD, sections 6.2.6 and
8.4.2.2), sensory effects, such as those related to hearing and
vision (CD, sections 6.2.7 and 8.4.2.3), and deficits in neuromotor
function (CD, p. 8-36).
---------------------------------------------------------------------------
The evidence for neurotoxic effects in children is a robust
combination of epidemiological and toxicological evidence (CD, Sections
5.3, 6.2 and 8.5). The epidemiological evidence is supported by animal
studies that substantiate the biological plausibility of the
associations, and contributes to our understanding of mechanisms of
action for the effects (CD, Section 8.4.2).
Cognitive effects associated with Pb exposures that have been
observed in epidemiological studies have included decrements in
intelligence test results, such as the widely used IQ score, and in
academic achievement as assessed by various standardized tests as well
as by class ranking and graduation rates (CD, Section 6.2.16 and pp 8-
29 to 8-30). As noted in the Criteria Document with regard to the
latter, ``Associations between Pb exposure and academic achievement
observed in the above-noted studies were significant even after
adjusting for IQ, suggesting that Pb-sensitive neuropsychological
processing and learning factors not reflected by global intelligence
indices might contribute to reduced performance on academic tasks''
(CD, pp 8-29 to 8-30).
Other cognitive effects observed in studies of children have
included effects on attention, executive functions, language, memory,
learning and visuospatial processing (CD, Sections 5.3.5, 6.2.5 and
8.4.2.1), with attention and executive function effects associated with
Pb exposures indexed by blood Pb levels below 10 [mu]g/dL (CD, Section
6.2.5 and pp. 8-30 to 8-31). The evidence for the role of Pb in this
suite of effects includes experimental animal findings (discussed in
CD, Section 8.4.2.1; p. 8-31), which provide strong biological
plausibility of Pb effects on learning ability, memory and attention
(CD, Section 5.3.5), as well as associated mechanistic findings. With
regard to persistence of effects the Criteria Document states the
following (CD, p. 8-67):
Persistence or apparent ``irreversibility'' of effects can
result from two different scenarios: (1) Organic damage has occurred
without adequate repair or compensatory offsets, or (2) exposure
somehow persists. As Pb exposure can also derive from endogenous
sources (e.g., bone), a performance deficit that remains detectable
after external exposure has ended, rather than indicating
irreversibility, could reflect ongoing toxicity due to Pb remaining
at the critical target organ or Pb deposited at the organ post-
exposure as the result of redistribution of Pb among body pools. The
persistence of effect appears to depend on the duration of exposure
as well as other factors that may affect an individual's ability to
recover from an insult. The likelihood of reversibility also seems
to be related, at least for the adverse effects observed in certain
organ systems, to both the age-at-exposure and the age-at-
assessment.
The evidence with regard to persistence of Pb-induced deficits observed
in animal and epidemiological studies is described in discussion of
those studies in the Criteria Document (CD, Sections 5.3.5, 6.2.11, and
8.5.2). It is additionally important to note that there may be long-
term consequences of such deficits over a lifetime. Poor academic
skills and achievement can have ``enduring and important effects on
objective parameters of success in real life,'' as well as increased
risk of antisocial and delinquent behavior (CD, Section 6.2.16).
As discussed in the Criteria Document, while there is no direct
animal test parallel to human IQ tests, ``in animals a wide variety of
tests that assess attention, learning, and memory suggest that Pb
exposure {of animals{time} results in a global deficit in functioning,
[[Page 29199]]
just as it is indicated by decrements in IQ scores in children'' (CD,
p. 8-27). The animal and epidemiological evidence for this endpoint are
consistent and complementary (CD, p. 8-44). As stated in the Criteria
Document (p. 8-44):
Findings from numerous experimental studies of rats and of
nonhuman primates, as discussed in Chapter 5, parallel the observed
human neurocognitive deficits and the processes responsible for
them. Learning and other higher order cognitive processes show the
greatest similarities in Pb-induced deficits between humans and
experimental animals. Deficits in cognition are due to the combined
and overlapping effects of Pb-induced perseveration, inability to
inhibit responding, inability to adapt to changing behavioral
requirements, aversion to delays, and distractibility. Higher level
neurocognitive functions are affected in both animals and humans at
very low exposure levels (<10 [mu]g/dL), more so than simple
cognitive functions.
Epidemiologic studies of Pb and child development have demonstrated
inverse associations between blood Pb concentrations and children's IQ
and other cognitive-related outcomes at successively lower Pb exposure
levels over the past 30 years (CD, p. 6-64). This is supported by
multiple studies performed over the past 15 years (as discussed in the
CD, Section 6.2.13). For example, the overall weight of the available
evidence, described in the Criteria Document, provides clear
substantiation of neurocognitive decrements being associated in
children with mean blood Pb levels in the range of 5 to 10 [mu]g/dL,
and some analyses indicate Pb effects on intellectual attainment of
children for which population mean blood Pb levels in the analysis
ranged from 2 to 8 [mu]g/dL (CD, Sections 6.2, 8.4.2 and 8.4.2.6).\44\
That is, while blood Pb levels in U.S. children have decreased notably
since the late 1970s, newer studies have investigated and reported
associations of effects on the neurodevelopment of children with blood
Pb levels similar to the more recent blood Pb levels (CD, Chapter 6).
---------------------------------------------------------------------------
\44\ ``The overall weight of the available evidence provides
clear substantiation of neurocognitive decrements being associated
in young children with blood-Pb concentrations in the range of 5-10
[mu]g/dL, and possibly somewhat lower. Some newly available analyses
appear to show Pb effects on the intellectual attainment of
preschool and school age children at population mean concurrent
blood-Pb levels ranging down to as low as 2 to 8 [mu]g/dL.'' (CD, p.
E-9)
---------------------------------------------------------------------------
The evidence described in the Criteria Document with regard to the
effect on children's cognitive function of blood Pb levels at the lower
concentration range includes the international pooled analysis by
Lanphear and others (2005), studies of individual cohorts such as the
Rochester, Boston, and Mexico City cohorts (Canfield et al., 2003a;
Canfield et al., 2003b; Bellinger and Needleman, 2003; Tellez-Rojo et
al., 2006), the study of African-American inner-city children from
Detroit (Chiodo et al., 2004), the cross-sectional study of young
children in three German cities (Walkowiak et al., 1998) and the cross-
sectional analysis of a nationally representative sample from the
NHANES III \45\ (Lanphear et al., 2000). These studies included
differing adjustments for different important potential confounders
(e.g., parental IQ or HOME score) or surrogates of these measures
(e.g., parental education and SES factors) through multivariate
analyses.46 47 Each of these studies has individual
strengths and limitations, however, a pattern of positive findings is
demonstrated across the studies. In these studies, statistically
significant associations of neurocognitive decrement \48\ with blood Pb
were found in the full study cohorts, as well as in some subgroups
restricted to children with lower blood Pb levels for which mean blood
Pb levels extended below 5 [mu]g/dL. More specifically, a statistically
significant association was reported for full-scale IQ with blood Pb at
age five in a subset analysis (n=71) of the Rochester cohort for which
the population mean blood Pb level was 3.32 [mu]g/dL, as well as in the
full study group (mean=5.8 [mu]g/dL, n=171) (Canfield et al., 2003a;
Canfield, 2008). Full-scale IQ was also significantly associated with
blood Pb at age seven and a half in a subset analysis (n=200) in the
Detroit inner-city study for which the population mean blood Pb level
was 4.1 [mu]g/dL, as well as the other subgroup with higher blood Pb
levels (mean=4.6 [mu]g/dL, n=224) and in the full study group (mean=5.4
[mu]g/dL, n=246); additionally, performance IQ was significantly
associated with blood Pb in those analyses as well as in the subset
analysis (n=120) for which the population mean blood Pb level was 3
[mu]g/dL (although full-scale IQ was not significantly associated with
blood Pb in this lowest blood Pb subgroup) (Chiodo et al., 2004,
Chiodo, 2008). Vocabulary, one of ten subtests of the full-scale IQ,
was significantly associated with blood
[[Page 29200]]
Pb at age six in the German three-city study (n=384) in which the mean
blood Pb level was 4.2 [mu]g/dL (Walkowiak et al., 1998). In a Mexico
City cohort of infants age two, the mental development index (MDI) and
psychomotor development index (PDI) were significantly associated with
blood Pb in the full study group (mean=4.28 [mu]g/dL, n=294); further,
the MDI (but not the PDI) was significantly associated with blood Pb in
the subset analysis (n=193) for which the population mean blood Pb
level was 2.9 [mu]g/dL, and PDI (but not the MDI) was significantly
associated with blood Pb in the subset analysis (n=101) for which the
population mean blood Pb was 6.9 [mu]g/dL (Tellez-Rojo et al., 2006;
Tellez-Rojo, 2008). Scores on academic achievement tests for reading
and math were significantly associated with blood Pb at age six through
sixteen in a subgroup analysis (n=4043) of the NHANES III data for
which the population mean blood Pb level was 1.7 [mu]g/dL, as discussed
below (Lanphear et al. 2000; Auinger, 2008).
---------------------------------------------------------------------------
\45\ The NHANES III survey was conducted in 1988-1994.
\46\ Some studies also employed exclusion criteria which limited
variation in socioeconomic status across the study population.
Further, with regard to adjustment for potential confounders in the
large pooled international analysis (Lanphear et al. 2005),
discussed below, the authors adjusted for HOME score, birth weight,
maternal IQ and maternal education. Canfield et al. (2003) adjusted
for maternal IQ, maternal education, HOME score, birth weight, race,
tobacco use during pregnancy, household income, gender, and iron
status. Bellinger and Needleman (2003) adjusted for maternal IQ,
HOME score, SES, child stress, maternal age, race, gender, birth
order, marital status. Chiodo et al. (2004) adjusted for primary
care-giver education and vocabulary, HOME score, family environment
scale, SES, gender, number of children under 18, birth order.
Tellez-Rojo et al. (2006) adjusted for maternal IQ, birth weight and
gender; the authors also state that other potentially confounding
variables that were not found to be significant at p<.10 were not
adjusted for. Walkoviak et al. (1998) adjusted for parental
education, breastfeeding, nationality and gender. In Lanphear et al.
(2000), the authors adjusted for race/ethnicity and poverty index
ratio, as surrogates for HOME score/SES status, and adjusted for the
parental education level as a surrogate for maternal IQ; they also
adjusted for gender, serum ferritin level and serum cotinine level.
\47\ The Criteria Document notes that a ``major challenge to
observational studies examining the impact of Pb on parameters of
child development has been the assessment and control for
confounding factors'' (CD, p. 6-73). However, the Criteria Document
further recognizes that ``[m]ost of the important confounding
factors in Pb studies have been identified, and efforts have been
made to control them in studies conducted since the 1990
Supplement'' (CD, p. 6-75). On this subject, the Criteria Document
further concludes the following: ``Invocation of the poorly measured
confounder as an explanation for positive findings is not
substantiated in the database as a whole when evaluating the impact
of Pb on the health of U.S. children (Needleman, 1995). Of course,
it is often the case that following adjustment for factors such as
social class, parental neurocognitive function, and child rearing
environment using covariates such as parental education, income, and
occupation, parental IQ, and HOME scores, the Pb coefficients are
substantially reduced in size and statistical significance (Dietrich
et al., 1991). This has sometimes led investigators to be quite
cautious in interpreting their study results as being positive
(Wasserman et al., 1997). This is a reasonable way of appraising any
single study, and such extreme caution would certainly be warranted
if forced to rely on a single study to confirm the Pb effects
hypothesis. Fortunately, there exists a large database of high
quality studies on which to base inferences regarding the
relationship between Pb exposure and neurodevelopment. In addition,
Pb has been extensively studied in animal models at doses that
closely approximate the human situation. Experimental animal studies
are not compromised by the possibility of confounding by such
factors as social class and correlated environmental factors. The
enormous experimental animal literature that proves that Pb at low
levels causes neurobehavioral deficits and provides insights into
mechanisms must be considered when drawing causal inferences
(Bellinger, 2004; Davis et al., 1990; U.S. Environmental Protection
Agency, 1986a, 1990).'' (CD, p. 6-75)
\48\ The tests for cognitive function in these studies include
age-appropriate Wechsler intelligence tests (Lanphear et al., 2005),
the Stanford-Binet intelligence test (Canfield et al., 2003a), and
the Bayley Scales of Infant Development (Tellez-Rojo et al., 2006).
In some cases, individual subtests of the Wechsler intelligence
tests (Lanphear et al., 2000; Walkowiak et al., 1998), and
individual subtests of the Wide Range Achievement Test (Lanphear et
al., 2000) were used. The Wechsler and Stanford-Binet tests are
widely used to assess neurocognitive function in children and
adults, however, these tests are not appropriate for children under
age three. For such children, studies generally use the age-
appropriate Bayley Scales of Infant Development as a measure of
cognitive development. See footnote 63 for further information.
---------------------------------------------------------------------------
The study by Lanphear et al. (2000) is a large cross-sectional
study using NHANES III dataset, with 4853 subjects in the full study
and more than 4000 in the subgroup analyses, that reports statistically
significant \49\ associations of concurrent blood Pb levels \50\ with
neurocognitive decrements in the full study population and in subgroup
analyses down to and including the subgroup with individual blood Pb
levels below 5 [mu]g/dL (CD, pp. 6-31 to 6-32; Lanphear et al., 2000).
Specifically the study by Lanphear et al. (2000) reported a
statistically significant association between math (p<0.001), reading
(p<0.001), block design (p=0.009), and digit span (p=0.04) scores and
blood Pb levels in the analysis that included all study subjects.
Additionally, the study reports statistically significant associations
for block design and digit span scores down to and including the
subgroup with individual blood Pb levels below 7.5 [mu]g/dL and 10
[mu]g/dL, respectively.\51\ Further, statistically significant
associations were observed for reading and math scores down to and
including the subgroup with individual blood Pb levels below 5 [mu]g/
dL, which included 4043 of the 4853 children.\52\ A similar pattern in
the magnitude of the effect estimates was observed across all the
subgroup analyses and for all four tests, including the subgroup with
individual blood Pb levels less than 2.5 [mu]g/dL, although not all the
effect estimates were statistically significant (Lanphear et al.,
2000).\53\ In particular, the lack of statistical significance in the
subset of individuals with blood Pb levels less than 2.5 [mu]g/dL may
be attributable to the smaller sample size (2467 children) and reduced
variability of blood Pb levels.\54\ Blood Pb levels in the full study
population ranged from below detection to above 10 [mu]g/dL, with a
population geometric mean of 1.9 [mu]g/dL, and the subgroups were
composed of children with blood Pb levels less than 10 [mu]g/dL
(geometric mean of 1.8 [mu]g/dL), less than 7.5 [mu]g/dL (geometric
mean of 1.8 [mu]g/dL), less than 5 [mu]g/dL (geometric mean of 1.7
[mu]g/dL), and less than 2.5 [mu]g/dL (geometric mean of 1.2 [mu]g/dL),
respectively (Lanphear et al., 2000; Auinger, 2008).\55\
---------------------------------------------------------------------------
\49\ The statistical significance refers to the effect estimate
of the linear relationship across the range of data, as presented in
Table 4 of Lanphear et al. (2000).
\50\ A limitation noted for this study is with regard to the use
of concurrent blood Pb levels in children of this age. The authors
state that ``it is not clear whether the cognitive and academic
deficits observed in the present analysis are due to lead exposure
that occurred during early childhood or due to concurrent
exposure'', however, they further note that ``concurrent blood lead
concentration was the best predictor of adverse neurobehavioral
effects of lead exposure in all but one of the published prospective
studies''. The average blood Pb level for 1-5 year olds was
approximately 15 [mu]g/dL in the 1976-1980 NHANES. When in that age
range, some of the children included in the NHANES III dataset may
have had blood Pb levels comparable to those of the earlier NHANES.
The general issue regarding blood Pb metrics is further discussed in
subsequent text.
\51\ The associations with block design score were not
statistically significant for subgroups limited to blood Pb of <5
and <2.5 [mu]g/dL. The associations with digit span score were not
statistically significant for the blood Pb subgroups of <7.5 and
lower.
\52\ The associations with math and reading scores were not
statistically significant for the subgroup limited to blood Pb <2.5
[mu]/dL.
\53\ For example, for reading scores, effect estimates were -
0.99, -1.44, -1.53, -1.66, and -1.71 points per [mu]g/dL for all
children, the subgroup with blood Pb <10 [mu]g/dL, the subgroup with
blood Pb <7.5, the subgroup with blood Pb <5 and the subgroup with
blood Pb<2.5, respectively (Lanphear et al., 2000, Table 4).
\54\ The authors state ``Indeed, while the average effects of
lead exposure on reading scores were not significant for blood lead
concentrations less that 2.5 [mu]g/dL, the size of the effect and
the borderline significance level ([beta] = -1.71, p=0.07) suggests
that the smaller sample size and the imprecision of the relationship
of blood Pb concentration with performance on the reading subtest--
as indicated by the large standard error--may be the reason we did
not find a statistically significant association for children in
that range.''
\55\ We note that the datasets for each subgroup include
children for the lower blood Pb subgroups (in Table 4 of Lanphear et
al., 2000). For example, the dataset of children with blood Pb
levels <2.5 is a component of the dataset of children with blood Pb
levels <5 (Lanphear et al., 2000).
---------------------------------------------------------------------------
The epidemiological studies that have investigated blood Pb effects
on IQ (as discussed in the CD, Section 6.2.3) have considered a variety
of specific blood Pb metrics, including: (1) Blood concentration
``concurrent'' with the response assessment (e.g., at the time of IQ
testing), (2) average blood concentration over the ``lifetime'' of the
child at the time of response assessment (e.g., average of measurements
taken over child's first 6 or 7 years), (3) peak blood concentration
during a particular age range, and (4) early childhood blood
concentration (e.g., the mean of measurements between 6 and 24 months
age). With regard to the latter two, the Criteria Document (e.g., CD,
chapters 3 and 6) has noted that age has been observed to strongly
predict the period of peak exposure (around 18-27 months when there is
maximum hand-to-mouth activity). The CD further notes, this maximum
exposure period coincides with a period of time in which major events
are occurring in central nervous system (CNS) development (CD, p. 6-
60). Accordingly, the belief that the first few years of life are a
critical window of vulnerability is evident particularly in the earlier
literature (CD, p. 6-60). However, more recent analyses have found even
stronger associations between blood Pb at school age and IQ at school
age (i.e., concurrent blood Pb), indicating the important role that is
continued to be played by Pb exposures later in life. In fact,
concurrent and lifetime averaged measurements were stronger predictors
of adverse neurobehavioral effects (better than the peak or 24 month
metrics) in all but one of the prospective cohort studies (CD, pp. 6-61
to 6-62). While all four specific blood Pb metrics were correlated with
IQ in the international pooled analysis by Lanphear and others (2005),
the concurrent blood Pb level exhibited the strongest relationship with
intellectual deficits (CD, p. 6-29).
The Criteria Document presentation on toxicological evidence also
recognizes neurological effects elicited by exposures subsequent to
earliest childhood (CD, sections 5.3.5 and 5.3.7). For example,
research with monkeys has indicated that while exposure only during
infancy may elicit a response, exposures (with similar blood Pb levels)
that only occurred post-infancy also elicit responses. Further, in the
monkey research, exposures limited to post-infancy resulted in a
greater response than exposures limited to infancy (Rice and Gilbert,
1990; Rice, 1992).
A study by Chen and others (2005) involving 622 children has
attempted to directly address the question regarding periods of
enhanced susceptibility to Pb effects (CD, pp. 6-62 to 6-64).\56\ The
authors found that the concurrent blood
[[Page 29201]]
Pb association with IQ was always stronger than that for 24-month blood
Pb. As children aged, the relationship with concurrent blood Pb grew
stronger while that with 24-month blood Pb grew weaker. Further, in
models including both prior blood Pb (at 24-months age) and concurrent
blood Pb (at 7-years age), concurrent blood Pb was always more
predictive of IQ. In fact, concurrent blood Pb explained most of Pb-
related variation in IQ such that prior blood Pb (at 24-months age) was
rendered nonsignificant and nearly null.\57\ The effect estimate for
concurrent blood Pb was robust and remained significant, little changed
from its value without adjustment for 24-month blood Pb level. The
Criteria Document concluded the following regarding the results of this
study (CD, pp. 6-63 to 6-64).
---------------------------------------------------------------------------
\56\ In the children in this study, the mean blood Pb
concentration was 26.2 [mu]g/dL at age 2, 12.0 [mu]g/dL at age 5 and
8.0 [mu]g/dL at age 7 (Chen et al. 2005).
\57\ We note that blood Pb levels at any point in time are
influenced by current as well as past exposures, e.g., through
exchanges between blood and bone (as summarized in section II.B.1
above and discussed in more detail in the Criteria Document).
These results support the idea that Pb exposure continues to be
toxic to children as they reach school age, and do not lend support
to the interpretation that all the damage is done by the time the
child reaches 2 to 3 years of age. These findings also imply that
cross-sectional associations seen in children, such as the study
recently conducted by Lanphear et al. (2000) using data from NHANES
III, should not be dismissed. Chen et al. (2005) concluded that if
concurrent blood Pb remains important until school age for optimum
cognitive development, and if 6- and 7-year-olds are as or more
sensitive to Pb effects than 2-year-olds, then the difficulties in
preventing Pb exposure are magnified but the potential benefits of
---------------------------------------------------------------------------
prevention are greater.
In addition to findings of association with neurocognitive
decrement (including IQ) at study group mean blood Pb levels well below
10 [mu]g/dL, the evidence indicates that the slope for Pb effects on IQ
is steeper at lower blood Pb levels (CD, section 6.2.13). As stated in
the CD, ``the most compelling evidence for effects at blood Pb levels
<10 [mu]g/dL, as well as a nonlinear relationship between blood Pb
levels and IQ, comes from the international pooled analysis of seven
prospective cohort studies (n=1,333) by Lanphear et al. (2005)'' (CD,
pp. 6-67 and 8-37 and section 6.2.3.1.11).\58\ Using the full pooled
dataset with concurrent blood Pb level as the exposure metric and IQ as
the response from the pooled dataset of seven international studies,
Lanphear and others (2005) employed mathematical models of various
forms, including linear, cubic spline, log-linear, and piece-wise
linear, in their investigation of the blood Pb concentration-response
relationship (CD, p. 6-29; Lanphear et al., 2005). They observed that
the shape of the concentration-response relationship is nonlinear and
the log-linear model provides a better fit over the full range of blood
Pb measurements \59\ than a linear one (CD, p. 6-29 and pp. 6-67 to 6-
70; Lanphear et al., 2005). In addition, they found that no individual
study among the seven was responsible for the estimated nonlinear
relationship between Pb and deficits in IQ (CD p. 6-30). Others have
also analyzed the same dataset and similarly concluded that, across the
range of the dataset's blood Pb levels, a log-linear relationship was a
significantly better fit than the linear relationship (p=0.009) with
little evidence of residual confounding from included model variables
(CD, Section 6.2.13; Rothenberg and Rothenberg, 2005).
---------------------------------------------------------------------------
\58\ We note that a public comment submitted on March 19, 2008
on behalf of the Association of Battery Recyclers described concerns
the commenter had with the conclusion by Lanphear et al. (2005) of a
nonlinear relationship of blood Pb with IQ, citing a publication by
Surkan et al. (2007), a study published since the completion of the
Criteria Document, and the Tellez-Rojo et al. (2006) finding,
discussed in the Criteria Document, of two different slopes for
their study subgroups of young children with blood Pb levels below 5
[mu]g/d (n=193, for which the slope of -1.7 was statistically
significant, p=0.01) and those with blood Pb levels between 5 and 10
[mu]g/dL (n=101, for which the slope of -0.94 was not statistically
significant, p=0.12). The commenter also cites another publication
published since the completion of the Criteria Document, Jusko et
al. (2007) related to this issue. EPA notes that it is not basing
its proposed decisions on studies that are not included in the
Criteria Document.
\59\ The geometric mean of the concurrent blood Pb levels
modeled was 9.7 [mu]g/dL; the 5th and 95th percentile values were
2.5 and 33.2 [mu]g/dL, respectively (Lanphear et al., 2005).
---------------------------------------------------------------------------
The impact of the nonlinear slope is illustrated by the log-linear
model-based estimates of IQ decrements for similar changes in blood Pb
level at different absolute values of blood Pb level (Lanphear et al.,
2005). These estimates of IQ decrement are 3.9 (with 95% confidence
interval, CI, of 2.4-5.3), 1.9 (95% CI, 1.2-2.6) and 1.1 IQ points per
[mu]g/dL blood Pb (95% CI, 0.7-1.5), for increases in concurrent blood
Pb from 2.4 to 10 [mu]g/dL, 10 to 20 [mu]g/dL, and 20 to 30 [mu]g/dL,
respectively (Lanphear et al., 2005). For an increase in concurrent
blood Pb levels from <1 to 10 [mu]g/dL, the log-linear model estimates
a decline of 6.2 points in full scale IQ which is comparable to the 7.4
point decrement in IQ for an increase in lifetime mean blood Pb levels
up to 10 [mu]g/dL observed in the Rochester study (CD, pp. 6-30 to 6-
31).
A nonlinear blood Pb concentration-response relationship is also
suggested by several other analyses that have observed that each [mu]g/
dL increase in blood Pb may have a greater effect on IQ at lower blood
Pb levels (e.g., below 10 [mu]g/dL) than at higher levels (CD, pp. 8-63
to 8-64; Figure 8-7). As noted in the Criteria Document, while this may
at first seem at odds with certain fundamental toxicological concepts,
a number of examples of non- or supralinear dose-response relationships
exist in toxicology (CD, pp. 6-76 and 8-38 to 8-39). With regard to the
effects of Pb on neurodevelopmental outcome such as IQ, the CD suggests
that initial neurodevelopmental effects at lower Pb levels may be
disrupting very different biological mechanisms (e.g., early
developmental processes in the central nervous system) than more severe
effects of high exposures that result in symptomatic Pb poisoning and
frank mental retardation (CD, p. 6-76).
The Criteria Document describes this issue with regard to Pb as
follows (CD, p. 8-39).
In the case of Pb, this nonlinear dose-effect relationship
occurs in the pattern of glutamate release (Section 5.3.2), in the
capacity for long term potentiation (LTP; Section 5.3.3), and in
conditioned operant responses (Section 5.3.5). The 1986 Lead AQCD
also reported U-shaped dose-effect relationships for maze
performance, discrimination learning, auditory evoked potential, and
locomotor activity. Davis and Svendsgaard (1990) reviewed U-shaped
dose-response curves and their implications for Pb risk assessment.
An important implication is the uncertainty created in
identification of thresholds and ``no-observed-effect-levels''
(NOELS). As a nonlinear relationship is observed between IQ and low
blood Pb levels in humans, as well as in new toxicologic studies
wherein neurotransmitter release and LTP show this same
relationship, it is plausible that these nonlinear cognitive
outcomes may be due, in part, to nonlinear mechanisms underlying
these observed Pb neurotoxic effects.
More specifically, various findings within the toxicological
evidence presented in the Criteria Document provides biologic
plausibility for a steeper IQ loss at low blood levels, with a
potential explanation being that the predominant mechanism at very low
blood-Pb levels is rapidly saturated and that a different, less-
rapidly-saturated process, becomes predominant at blood-Pb levels
greater than 10 [mu]g/dL.\60\
---------------------------------------------------------------------------
\60\ The toxicological evidence presented in the Criteria
Document of biphasic dose-effect relationships includes: Suppression
of stimulated hippocampal glutamate release at low exposure levels
and induction of glutamate exocytosis at higher exposure levels (CD,
Section 5.3.2); downregulation of NMDA receptors at low blood Pb
levels and upregulation at higher levels (CD, section 5.3.2); Pb
causes elevated induction threshold and diminished magnitude of
long-term potentiation at low exposures, but not at higher exposures
(CD, section 5.3.3); and low-level Pb exposures increase fixed-
interval response rates and high-level Pb exposures decrease fixed
interval response rates in learning deficit testing in rats (CD,
section 5.3.5). Additional in vitro evidence includes Pb stimulation
of PKC activity at picomolar concentrations and inhibition of PKC
activity at nano- and micro-molar concentrations (CD, section
5.3.2).
---------------------------------------------------------------------------
[[Page 29202]]
In addition to the observed associations between neurocognitive
decrement (including IQ) and blood Pb at study group mean levels well
below 10 [mu]g/dL (described above), the current evidence includes
multiple studies that have examined the quantitative relationship
between IQ and blood Pb level in analyses of children with individual
blood Pb concentrations below 10 [mu]g/dL. In comparing across the
individual epidemiological studies and the international pooled
analysis, the Criteria Document observed that at higher blood Pb levels
(e.g., above 10 [mu]g/dL), the slopes (for change in IQ with blood Pb)
derived for log-linear and linear models are almost identical, and for
studies with lower blood Pb levels, the slopes appear to be steeper
than those observed in studies involving higher blood Pb levels (CD, p.
8-78, Figure 8-7). In making these observations, the Criteria Document
focused on the curves from the models from the 10th percentile to the
90th percentile saying that the ``curves are restricted to that range
because log-linear curves become very steep at the lower end of the
blood Pb levels, and this may be an artifact of the model chosen.''
The quantitative relationship between IQ and blood Pb level has
been examined in the Criteria Document using studies where all or the
majority of study subjects had blood Pb levels below 10 [mu]g/dL and
also where an analysis was performed on a subset of children whose
blood Pb levels have never exceeded 10 [mu]g/dL (CD, Table 6-1). The
datasets for three of these studies included concurrent blood Pb levels
above 10 [mu]g/dL; the C-R relationship reported for one of the three
was linear while it was log-linear for the other two. For the one of
these three studies with the linear C-R relationship, the highest blood
Pb level was just below 12 [mu]g/dL (Kordas et al., 2006). Of the two
studies with log-linear functions, one reported 69% of the children
with blood Pb levels below 10 [mu]g/dL and a population mean blood Pb
level of 7.44 [mu]g/dL (Al-Saleh et al., 2001), and the second reported
a population median blood Pb level of 9.7 [mu]g/dL and a 95th
percentile of 33.2 [mu]g/dL (Lanphear et al., 2005). In order to
compare slopes across all of these studies (linear and log-linear), EPA
estimated, for each, the average slope of change in IQ with change in
blood Pb between the 10th percentile \61\ blood Pb level and 10 [mu]g/
dL (CD, Table 6-1). The resultant group of reported and estimated
average linear slopes for IQ change with blood Pb levels up to 10
[mu]g/dL range from -0.4 to -1.8 IQ points per [mu]g/dL blood Pb (CD,
Tables 6-1 and 8-7), with a median of -0.9 IQ points per [mu]g/dL blood
Pb (CD, pp. 8-80).\62\
---------------------------------------------------------------------------
\61\ In the Criteria Document analysis, the 10th percentile was
chosen as a common point of comparison for the loglinear (and
linear) models at a point prior to the lowest end of the blood Pb
levels.
\62\ Among this group of slopes (CD, Table 6-1) is that from the
analysis of the IQ-blood Pb (concurrent) relationship for children
whose peak blood Pb levels are below 10 [mu]g/dL in the
international pooled dataset studied by Lanphear and others (2005);
these authors reported this slope along with the companion slope for
blood Pb levels for the remaining children with peak blood Pb level
equal to or above 10 [mu]g/dL (Lanphear et al., 2005). In the
economic analysis for EPA's recent Lead Renovation, Repair and
Painting (RRP) Program rule (described above in section I.C),
changes in IQ loss as a function of changes in lifetime average
blood Pb level were estimated using the corresponding piecewise
model for lifetime average blood Pb derived from the pooled dataset
(USEPA, 2008; USEPA, 2007e). Selection of this model for the RRP
economic analysis reflects consideration of the distribution of
blood Pb levels in that analysis, those for children living in
houses with Pb-based paint. With consideration of these blood Pb
levels, the economic analysis document states that ``[s]electing a
model with a node, or changing one segment to the other, at a
lifetime average blood Pb concentration of 10 [mu]g/dL rather than
at 7.5 [mu]g/dL, is a small protection against applying an
incorrectly rapid change (steep slope with increasingly smaller
effect as concentrations lower) to the calculation''. We note that
the slope for the less-than-10-[mu]g/dL portion of the model used in
the RRP analysis (-0.88) is similar to the median for the slopes
included in the Criteria Document analysis of quantitative
relationships for distributions of blood Pb levels extending from
just below 10 [mu]g/dL and lower.
---------------------------------------------------------------------------
Among this group of quantitative IQ-blood Pb relationships examined
in the Criteria Document (CD, Tables 6-1 and 8-7), the steepest slopes
for change in IQ with change in blood Pb level are those derived for
the subsets of children in the Rochester and Boston cohorts for which
peak blood Pb levels were <10 [mu]g/dL; these slopes, in terms of IQ
points per [mu]g/dL blood Pb, are -1.8 (for concurrent blood Pb
influence on IQ) and -1.6 (for 24-month blood Pb influence on IQ),
respectively. The mean blood Pb levels for children in these subsets of
the Rochester and Boston cohorts are 3.32 and 3.8 [mu]g/dL,
respectively, which are the lowest population mean levels among the
datasets included in the table (Canfield, 2008; Bellinger, 2008). Other
studies with analyses involving similarly low blood Pb levels (e.g.,
mean levels below 4 [mu]g/dL) also had slopes steeper than -1.5 points
per [mu]g/dL blood Pb. These include the slope of -1.71 points per
[mu]g/dL blood Pb \63\ for the subset of 24-month-old children in the
Mexico City cohort with blood Pb levels less than 5 [mu]g/dL (n=193),
for which the mean concurrent blood Pb level was 2.9 [mu]g/dL (Tellez-
Rojo et al. 2006, 2008) \64\ and also the slope of -2.94 points per
[mu]g/dL blood Pb for the subset of 6-10-year-old children whose peak
blood Pb levels never exceeded 7.5 [mu]g/dL (n=112), and for which the
mean concurrent blood Pb level was 3.24 [mu]g/dL (Lanphear et al. 2005;
Hornung 2008). Thus, from these subset analyses, the slopes range from
-1.71 to -2.94 IQ points per [mu]g/dL of concurrent blood Pb. We also
note that the nonlinear C-R function in which greatest confidence is
placed in estimating IQ loss in the quantitative risk assessment
(described below in section II.C) has a slope that falls
[[Page 29203]]
intermediate between these two for blood Pb levels up to approximately
3.7 [mu]g/dL (USEPA, 2007b).
---------------------------------------------------------------------------
\63\ This slope reflects effects on cognitive development in
this cohort of 24-month-old children based on the age-appropriate
test described earlier, and is similar in magnitude to slopes for
the cohorts of older children described here. The strengths and
limitations of this age-appropriate text, the Mental Development
Index (MDI) of the Bayley Scales of Infant Development (BSID), were
discussed in a letter to the editor by Black and Baqui (2005). The
authors state that ``the MDI is a well-standardized,
psychometrically strong measure of infant mental development.'' The
MDI represents a complex integration of empirically-derived
cognitive skills, for example, sensory/perceptual acuities,
discriminations, and response; acquisition of object constancy;
memory learning and problem solving; vocalization and beginning of
verbal communication; and basis of abstract thinking. Black and
Baqui state that although the MDI is one of the most well-
standardized, widely used assessment of infant mental development,
evidence indicates low predictive validity of the MDI for infants
younger than 24 months to subsequent measures of intelligence. They
explain that the lack of continuity may be partially explained by
``the multidimensional and rapidly changing aspects of infant mental
development and by variations in performance during infancy,
variations in tasks used to measure intellectual functioning
throughout childhood, and variations in environmental challenges and
opportunities that may influence development.'' Martin and Volkmar
(2007) also noted that correlations between BSID performance and
subsequent IQ assessments were variable, but they also reported high
test-retest reliability and validity, as indicated by the
correlation coefficients of 0.83 to 0.91, as well as high interrater
reliability, correlation coefficient of 0.96, for the MDI.
Therefore, the BSID has been found to be a reliable indicator of
current development and cognitive functioning of the infant. Martin
and Volkmar (2007) further note that ``for the most part,
performance on the BSID does not consistently predict later
cognitive measures, particularly when socioeconomic status and level
of functioning are controlled''.
\64\ In this study, the slope for blood Pb levels between 5 and
10 [mu]g/dL (population mean blood Pb of 6.9 [mu]g/dL; n=101) was -
0.94 points per [mu]g/dL blood Pb but was not statistically
significant, with a P value of 0.12. The difference in the slope
between the <5 [mu]g/dL and the 5-10 [mu]g/dL groups was not
statistically significant (Tellez-Rojo et al., 2006; Tellez-Rojo,
2008).
---------------------------------------------------------------------------
The C-R functions discussed above are presented in two sets in
Table 1 below.
Table 1. Summary of Quantitative Relationships of IQ and Blood Pb for Two Sets of Studies Discussed Above
--------------------------------------------------------------------------------------------------------------------------------------------------------
Form of model Average
Range BLL ([mu]g/ Geometric mean BLL from which linear slope
Study/Analysis Study cohort Analysis dataset N dL) 5th-95th ([mu]g/dL) average slope \A\ (points
percentile] derived per [mu]g/dL)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Set of studies from which steeper slopes are drawn
--------------------------------------------------------------------------------------------------------------------------------------------------------
Tellez-Rojo <5 subgroup based Mexico City, age Children--BLL<5 193 0.8-4.9........... 2.9............... Linear.......... -1.71
on Lanphear et al. 2005,\B\ 24 mo. [mu]g/dL.
Log-linear with low-exposure
linearization (LLL) \B\.
Dataset from which the log-linear function is derived is the pooled International LLL\C\.......... -2.29 at 2
dataset of 1333 children, age 6-10 yr, having median blood Pb of 9.7 [mu]g/dL and 5th- [mu]g/dL\C\
95th percentile of 2.5-33.2 [mu]g/dL.Slope presented here is the slope at a blood Pb
level of 2 [mu]g/dL.\C\
Lanphear et al. 2005,\B\ <7.5 Pooled Children--peak 103 [1.3-6.0]......... 3.24.............. Linear.......... -2.94
peak subgroup. International, BLL <7.5 [mu]g/
age 6-10 yr. dL.
--------------------------------------------------------------------------------------------------------------------------------------------------------
Set of studies with shallower slopes (Criteria Document, Table 6-1) \D\
--------------------------------------------------------------------------------------------------------------------------------------------------------
Canfield et al. 2003 \B\, <10 Rochester, age 5 Children--peak 71 Unspecified....... 3.32.............. Linear.......... -1.79
peak subgroup. yr. BLL <10 [mu]g/dL.
Bellinger and Needleman Boston\A\ \E\.... Children--peak 48 1-9.3\E\.......... 3.8\E\............ Linear.......... -1.56
2003\B\. BLL <10 [mu]g/dL.
Tellez-Rojo et al. 2006....... Mexico City, age Full dataset..... 294 0.8-<10........... 4.28.............. Linear.......... -1.04
24 mo.
Tellez-Rojo et al. 2006 full-- Mexico City, age Full dataset..... 294 0.8-<10........... 4.28.............. Log-linear...... -0.94
loglinear. 24 mo.
Lanphear et al. 2005,\B\ <10 Pooled Children--peak 244 [1.4-8.0]......... 4.30.............. Linear.......... -0.80
peak\F\ subgroup. International, BLL <10 [mu]g/dL.
age 6-10 yr.
Al-Saleh et al. 2001 full-- Saudi Arabia, age Full dataset..... 533 2.3-27.36\G\...... 7.44.............. Log-linear...... -0.76
loglinear. 6-12 yr.
Kordas et al. 2006, <12 Torreon, Mexico, Children--BLL<12 377 2.3-<12........... 7.9............... Linear.......... -0.40
subgroup. age 7 yr. [mu]g/dL.
Lanphear et al. 2005\B\ full-- Pooled Full dataset..... 1333 [2.5-33.2]........ 9.7 (median)...... Log-linear...... -0.41
loglinear. International,
age 6-10 yr.
Median value.... -0.9 \D\
--------------------------------------------------------------------------------------------------------------------------------------------------------
\A\ Average slope for change in IQ from 10th percentile to 10 [mu]g/dL Slope estimates here are for relationship between IQ and concurrent blood Pb
levels (BLL), except for Bellinger & Needleman which used 24 month BLLs with 10 year old IQ.
\B\ The Lanphear et al. 2005 pooled International study includes blood Pb data from the Rochester and Boston cohorts, although for different ages (6 and
5 years, respectively) than the ages analyzed in Canfield et al. 2003 and Bellinger and Needleman 2003.
\C\ The LLL function (described in section II.C.2.b) was developed from Lanphear et al. 2005 loglinear model with a linearization of the slope at BLL
below 1 [mu]g/dL. The slope shown is that at 2 [mu]g/dL. In estimating IQ loss with this function in the risk assessment (section II.C) and in the
evidence-based considerations in section II.E.3, the nonlinear form of the model was used, with varying slope for all BLL above 1 [mu]g/dL.
\D\ These studies and quantitative relationships are discussed in the Criteria Document (CD, sections 6.2, 6.2.1.3 and 8.6.2).
\E\ The BLL for Bellinger and Needleman (2003) are for age 24 months.
\F\ As referenced above and in section II.C.2.b, the form of this function derived for lifetime average blood Pb was used in the economic analysis for
the RRP rule. The slope for that function was -0.88 IQ points per [mu]g/dL lifetime averaged blood Pb.
\G\ 69% of children in Al-Saleh et al. (2001) study had BLL<10 [mu]g/dL.
[[Page 29204]]
3. Lead-Related Impacts on Public Health
In addition to the advances in our knowledge and understanding of
Pb health effects at lower exposures (e.g., using blood Pb as the
index), there has been some change with regard to the U.S. population
Pb burden since the time of the last Pb NAAQS review. For example, the
geometric mean blood Pb level for U.S. children aged 1-5, as estimated
by the U.S. Centers for Disease Control, declined from 2.7 [mu]g/dL
(95% CI: 2.5-3.0) in the 1991-1994 survey period to 1.7 [mu]g/dL (95%
CI: 1.55-1.87) in the 2001-2002 survey period (CD, Section 4.3.1.3) and
1.8 [mu]g/dL in the 2003-2004 survey period (Axelrad, 2008).\65\ Blood
Pb levels have also declined in the U.S. adult population over this
time period (CD, Section 4.3.1.3).\66\ As noted in the Criteria
Document, ``blood-Pb levels have been declining at differential rates
for various general subpopulations, as a function of income, race, and
certain other demographic indicators such as age of housing'' (CD, pp.
8-21). For example, the geometric mean blood Pb level for children
(aged one to five) living in poverty in the 2003-2004 survey period is
2.4 [mu]g/dL. For black, non-Hispanic children, the geometric mean is
2.7 [mu]g/dL, and for the subset of this group that is living in
poverty, the geometric mean is 3.1 [mu]g/dL. Further, the 95th
percentile blood Pb level in the 2003-2004 NHANES for children aged 1-5
of all races and ethnic groups is 5.1 [mu]g/dL, while the corresponding
level for the subset of children living below the poverty level is 6.6
[mu]g/dL. The 95th percentile level for black, non-Hispanic children is
8.9 [mu]g/dL, and for the subset of that group living below the poverty
level, it is 10.5 [mu]g/dL (Axelrad, 2008).\67\
---------------------------------------------------------------------------
\65\ These levels are in contrast to the geometric mean blood Pb
level of 14.9 [mu]g/dL reported for U.S. children (aged 6 months to
5 years) in 1976-1980 (CD, Section 4.3.1.3).
\66\ For example, NHANES data for older adults (60 years of age
and older) indicate a decline in overall population geometric mean
blood Pb level from 3.4 [mu]g/dL in 1991-1994 to 2.2 [mu]g/dL in
1999-2002; the trend for adults between 20 and 60 years of age is
similar to that for children 1 to 5 years of age (http://
www.cdc.gov/mmwr/preview/mmwrhtml/mm5420a5.htm).
\67\ Although the 90th percentile statistic for these subgroups
is not currently available for the 2003-04 survey period, the 2001-
2004 90th percentile blood Pb level for children aged 1-5 of all
races and ethnic groups is 4.0 [mu]g/dL, while the corresponding
level for the subset of children living below the poverty level is
5.4 [mu]g/dL, and that level for black, non-Hispanic children living
below the poverty level is 7.7 [mu]g/dL (http://www.epa.gov/
envirohealth/children/body_burdens/b1-table.htm--then click on
``Download a universal spreadsheet file of the Body Burdens data
tables'').
---------------------------------------------------------------------------
a. At-Risk Subpopulations
Potentially at-risk subpopulations include those with increased
susceptibility (i.e., physiological factors contributing to a greater
response for the same exposure) and those with increased exposure
(including that resulting from behavior leading to increased contact
with contaminated media) (USEPA 1986a, pp. 1-154). A behavioral factor
of great impact on Pb exposure is the incidence of hand-to-mouth
activity that is prevalent in very young children (CD, Section 4.4.3).
Physiological factors include both conditions contributing to a
subgroup's increased risk of effects at a given blood Pb level, and
those that contribute to blood Pb levels higher than those otherwise
associated with a given Pb exposure (CD, Section 8.5.3). These factors
include nutritional status (e.g., iron deficiency, calcium intake), as
well as genetic and other factors (CD, chapter 4 and sections 3.4,
5.3.7 and 8.5.3).
We also considered evidence pertaining to vulnerability to
pollution-related effects which additionally encompasses situations of
elevated exposure, such as residing in older housing with Pb-containing
paint or near sources of ambient Pb, as well as socioeconomic factors,
such as reduced access to health care or low socioeconomic status (SES)
(USEPA, 2003, 2005c) that can contribute to increased risk of adverse
health effects from Pb. With regard to elevated exposures in particular
socioeconomic and minority subpopulations, we observe notably higher
blood Pb levels in children in poverty and in black, non-Hispanic
children compared to those for more economically well-off children and
white children, in general (as recognized in section II.B.1.b above).
Three particular physiological factors contributing to increased
risk of Pb effects at a given blood Pb level are recognized in the
Criteria Document (e.g., CD, Section 8.5.3): age, health status, and
genetic composition. With regard to age, the susceptibility of young
children to the neurodevelopmental effects of Pb is well recognized
(e.g., CD, Sections 5.3, 6.2, 8.4, 8.5, 8.6.2), although the specific
ages of vulnerability have not been established (CD, pp. 6-60 to 6-64).
Early childhood may also be a time of increased susceptibility for Pb
immunotoxicity (CD, Sections 5.9.10, 6.8.3 and 8.4.6). Further early
life exposures have been associated with increased risk of
cardiovascular effects in humans later in life (CD, pp. 8-74). Early
life exposures have also been associated with increased risk, in
animals, of neurodegenerative effects later in life (CD, pp. 8-74).\68\
Health status is another physiological factor in that subpopulations
with pre-existing health conditions may be more susceptible (as
compared to the general population) for particular Pb-associated
effects, with this being most clear for renal and cardiovascular
outcomes. For example, African Americans as a group have a higher
frequency of hypertension than the general population or other ethnic
groups (NCHS, 2005), and as a result may face a greater risk of adverse
health impact from Pb-associated cardiovascular effects. A third
physiological factor relates to genetic polymorphisms. That is,
subpopulations defined by particular genetic polymorphisms (e.g.,
presence of the [delta]-aminolevulinic acid dehydratase-2 [ALAD-2]
allele) have also been recognized as sensitive to Pb toxicity, which
may be due to increased susceptibility to the same internal dose and/or
to increased internal dose associated with the same exposure (CD, pp.
8-71, Sections 6.3.5, 6.4.7.3 and 6.3.6).
---------------------------------------------------------------------------
\68\ Specifically, among young adults who lived as children in
an area heavily polluted by a smelter and whose current Pb exposure
was low, higher bone Pb levels were associated with higher systolic
and diastolic blood pressure (CD, pp. 8-74). In adult rats, greater
early exposures to Pb are associated with increased levels of
amyloid protein precursor, a marker of risk for neurodegenerative
disease (CD, pp. 8-74).
---------------------------------------------------------------------------
Childhood is well recognized as a time of increased susceptibility,
and as summarized in section II.B.2.b above and described in more
detail in the Criteria Document, a large body of epidemiological
evidence describes neurological effects on children at low blood Pb
levels. The toxicological evidence further helps inform an
understanding of specific periods of development with increased
vulnerability to specific types of neurological effect (CD, Section
5.3). Additionally, the toxicological evidence of a differing
sensitivity of the immune system to Pb across and within different
periods of life stages indicates the potential importance of exposures
of duration as short as weeks to months. For example, the animal
studies suggest that, for immune effects, the gestation period is the
most sensitive life stage followed by early neonatal stage, and that
within these life stages, critical windows of vulnerability are likely
to exist (CD, Section 5.9 and p. 5-245).
In summary, there are a variety of ways in which Pb exposed
populations might be characterized and stratified for consideration of
public health impacts. Age or lifestage was used to distinguish
[[Page 29205]]
potential groups on which to focus the quantitative risk assessment
because of its influence on exposure and susceptibility. Young children
were selected as the priority population for the risk assessment in
consideration of the health effects evidence regarding endpoints of
greatest public health concern. The Criteria Document recognizes,
however, other population subgroups as described above may also be at
risk of Pb-related health effects of public health concern.
b. Potential Public Health Impacts
As discussed in the Criteria Document, there are potential public
health implications of low-level Pb exposure, indexed by blood Pb
levels, associated with several health endpoints identified in the
Criteria Document (CD, Section 8.6).\69\ These include potential
impacts on population IQ, which is the focus of the quantitative risk
assessment conducted for this review, as well as heart disease and
chronic kidney disease, which are not included in the quantitative risk
assessment (CD, Sections 8.6, 8.6.2, 8.6.3 and 8.6.4). It is noted that
there is greater uncertainty associated with effects at the lower
levels of blood Pb, and that there are differing weights of evidence
across the effects observed.\70\ With regard to potential implications
of Pb effects on IQ, the Criteria Document recognizes the ``critical''
distinction between population and individual risk, noting that a
``point estimate indicating a modest mean change on a health index at
the individual level can have substantial implications at the
population level'' (CD, p. 8-77).\71\ A downward shift in the mean IQ
value is associated with both substantial decreases in percentages
achieving very high scores and substantial increases in the percentage
of individuals achieving very low scores (CD, p. 8-81).\72\ For an
individual functioning in the low IQ range due to the influence of
developmental risk factors other than Pb, a Pb-associated IQ decline of
several points might be sufficient to drop that individual into the
range associated with increased risk of educational, vocational, and
social handicap (CD, p. 8-77).
---------------------------------------------------------------------------
\69\ The differing evidence and associated strength of the
evidence for these different effects is described in detail in the
Criteria Document.
\70\ As is described in Section II.C.2.a, CASAC, in their
comments on the analysis plan for the risk assessment described in
this notice, placed higher priority on modeling the child IQ metric
than the adult endpoints (e.g., cardiovascular effects).
\71\ Similarly, ``although an increase of a few mmHg in blood
pressure might not be of concern for an individual's well-being, the
same increase in the population mean might be associated with
substantial increases in the percentages of individuals with values
that are sufficiently extreme that they exceed the criteria used to
diagnose hypertension'' (CD, p. 8-77).
\72\ For example, for a population mean IQ of 100 (and standard
deviation of 15), 2.3% of the population would score above 130, but
a shift of the population to a mean of 95 results in only 0.99% of
the population scoring above 130 (CD, pp. 8-81 to 8-82).
---------------------------------------------------------------------------
The magnitude of a public health impact is dependent upon the size
of population affected and type or severity of the effect. As
summarized above, there are several population groups that may be
susceptible or vulnerable to effects associated with exposure to Pb,
including young children, particularly those in families of low SES
(CD, p. E-15), as well as individuals with hypertension, diabetes, and
chronic renal insufficiency (CD, p. 8-72). Although comprehensive
estimates of the size of these groups residing in proximity to sources
of ambient Pb have not been developed, total estimates of these
population subpopulations within the U.S. are substantial (as noted in
Table 3-3 of the Staff Paper).\73\
---------------------------------------------------------------------------
\73\ For example, approximately 4.8 million children live in
poverty, while the estimates of numbers of adults with hypertension,
diabetes or chronic kidney disease are on the order of 20 to 50
million (see Table 3-3 of Staff Paper).
---------------------------------------------------------------------------
With regard to estimates of the size of potentially vulnerable
subpopulations living in areas of increased exposure related to ambient
Pb, the information is still more limited. The limited information
available on air and surface soil concentrations of Pb indicates
elevated concentrations near stationary sources as compared with areas
remote from such sources (CD, Sections 3.2.2 and 3.8). Air quality
analyses (presented in Chapter 2 of the Staff Paper) indicate
dramatically higher Pb concentrations at monitors near sources as
compared with those more remote. As described in Section 2.3.2.1 of the
Staff Paper, however, since the 1980s the number of Pb monitors has
been significantly reduced by states (with EPA guidance that monitors
well below the current NAAQS could be shut down) and a lack of monitors
near some large sources may lead to underestimates of the extent of
occurrences of relatively higher Pb concentrations. The significant
limitations of our monitoring and emissions information constrain our
efforts to characterize the size of at-risk populations in areas
influenced by sources of ambient Pb. For example, the limited size and
spatial coverage of the current Pb monitoring network constrains our
ability to characterize current levels of airborne Pb in the U.S.
Further, as noted above in section II.A.1, the Staff Paper review of
the available information on emissions and locations of sources (as
described in section 2.3.2.1 of the Staff Paper) indicates that the
network is inconsistent in its coverage of the largest sources
identified in the 2002 National Emissions Inventory (NEI). The most
recent analysis of monitors near sources greater than 1 ton per year
(tpy) indicates that less than 15% of stationary sources with emissions
greater than or equal to 1 tpy have a monitor within one mile.
Additionally, there are various uncertainties and limitations
associated with source information in the NEI (as described in section
2.2.5 of the Staff Paper; USEPA, 2007c).
In recognition of the significant limitations associated with the
currently available information on Pb emissions and airborne
concentrations in the U.S. and the associated exposure of potentially
at-risk populations, Chapter 2 of the Staff Paper summarizes the
information in several different ways. For example, analyses of the
current monitoring network indicated the numbers of monitoring sites
that would exceed alternate standard levels, taking into consideration
different statistical forms. These analyses are also summarized with
regard to population size in counties home to those monitoring sites
(as presented in Appendix 5.A of the Staff Paper). Information for the
monitors and from the NEI indicates a range of source sizes in
proximity to monitors at which various levels of Pb are reported.
Together this information suggests that there is variety in the
magnitude of Pb emissions from sources that could influence air Pb
concentrations. Identifying specific emissions levels of sources
expected to result in air Pb concentrations of interest, however, would
be informed by a comprehensive analysis using detailed source
characterization information, which was not feasible within the time
and data constraints of this review. Instead, we have developed a
summary of the emissions and demographic information for Pb sources
that includes estimates of the numbers of people residing in counties
in which the aggregate Pb emissions from NEI sources is greater than or
equal to 0.1 tpy or in counties in which the aggregate Pb emissions is
greater than or equal to 0.1 tpy per 1000 square miles (as presented in
Tables 3-4 and 3-5, respectively, in the Staff Paper).
Additionally, the potential for resuspension of recently and
historically deposited Pb near roadways to contribute to increased
risks of Pb exposure to populations residing nearby is suggested in the
Criteria Document (e.g., CD, pp. 2-62 and 3-32).
[[Page 29206]]
4. Key Observations
The following key observations are based on the available health
effects evidence and the evaluation and interpretation of that evidence
in the Criteria Document.
Lead exposures occur both by inhalation and by ingestion
(CD, Chapter 3). As stated in the Criteria Document, ``given the large
amount of time people spend indoors, exposure to Pb in dusts and indoor
air can be significant'' (CD, p. 3-27).
Children, in general and especially those of low SES, are
at increased risk for Pb exposure and Pb-induced adverse health
effects. This is due to several factors, including enhanced exposure to
Pb via ingestion of soil Pb and/or dust Pb due to normal childhood
hand-to-mouth activity (CD, p. E-15, Chapter 3 and Section 6.2.1).
Once inhaled or ingested, Pb is distributed by the blood,
with long-term storage accumulation in the bone. Bone Pb levels provide
a strong measure of cumulative exposure which has been associated with
many of the effects summarized below, although difficulty of sample
collection has precluded widespread use in epidemiological studies to
date (CD, Chapter 4).
Blood levels of Pb are well accepted as an index of
exposure (or exposure metric) for which associations with the key
effects (see below) have been observed. In general, associations with
blood Pb are most robust for those effects for which past exposure
history poses less of a complicating factor, i.e., for effects during
childhood (CD, Section 4.3).
Both epidemiological and toxicologic studies have shown
that environmentally relevant levels of Pb affect many different organ
systems (CD, p. E-8). With regard to the most important such effects
observed in children and adults, the Criteria Document states (CD, p.
8-60) that ``neurotoxic effects in children and cardiovascular effects
in adults are among those best substantiated as occurring at blood-Pb
concentrations as low as 5 to 10 [mu]g/dL (or possibly lower); and
these categories of effects are currently clearly of greatest public
health concern. Other newly demonstrated immune and renal system
effects among general population groups are also emerging as low-level
Pb-exposure effects of potential public health concern.''
Many associations of health effects with Pb exposure have
been found at levels of blood Pb that are currently relevant for the
U.S. population, with individual children having blood Pb levels of 5-
10 [mu]g/dL and lower, being at risk for neurological effects (as
described in the subsequent bullet). Supportive evidence from
toxicological studies provides biological plausibility for the observed
effects. (CD, Chapters 5, 6 and 8)
Pb exposure is associated with a variety of neurological
effects in children, notably intellectual attainment and school
performance. Both qualitative and quantitative evidence, with further
support from animal research, indicates a robust and consistent effect
of Pb exposure on neurocognitive ability at mean concurrent blood Pb
levels in the range of 5 to 10 [mu]g/dL. Specific epidemiological
analyses have further indicated association with neurocognitive effects
in analyses restricted to children with individual blood Pb levels
below 5-10 [mu]g/dL, and for which group mean levels are lower.
Further, ``[s]ome newly available analyses appear to show Pb effects on
the intellectual attainment of preschool and school age children at
population mean concurrent blood-Pb levels ranging down to as low as 2
to 8 [mu]g/dL'' (CD, p. E-9; Sections 5.3, 6.2, 8.4.2 and 6.10).
Deficits in cognitive skills may have long-term
consequences over a lifetime. Poor academic skills and achievement can
have enduring and important effects on objective parameters of success
in life as well as increased risk of antisocial and delinquent
behavior. (CD, Sections 6.1 and 8.4.2)
The current epidemiological evidence indicates a steeper
slope of the blood Pb concentration-response relationship at lower
blood Pb levels, particularly those below 10 [mu]g/dL (CD, Sections
6.2.13 and 8.6).
At mean blood Pb levels, in children, on the order of 10
[mu]g/dL, and somewhat lower, associations have been found with effects
to the immune system, including altered macrophage activation,
increased IgE levels and associated increased risk for autoimmunity and
asthma (CD, Sections 5.9, 6.8, and 8.4.6).
In adults, with regard to cardiovascular outcomes, the
Criteria Document included the following summary (CD, p. E-10).
Epidemiological studies have consistently demonstrated
associations between Pb exposure and enhanced risk of deleterious
cardiovascular outcomes, including increased blood pressure and
incidence of hypertension.\74\ A meta-analysis of numerous studies
estimates that a doubling of blood-Pb level (e.g., from 5 to 10
[mu]g/dL) is associated with ~1.0 mm Hg increase in systolic blood
pressure and ~0.6 mm Hg increase in diastolic pressure. Studies have
also found that cumulative past Pb exposure ( e.g., bone Pb) may be
as important, if not more, than present Pb exposure in assessing
cardiovascular effects. The evidence for an association of Pb with
cardiovascular morbidity and mortality is limited but supportive.
---------------------------------------------------------------------------
\74\ The Criteria Document states that ``While several studies
have demonstrated a positive correlation between blood pressure and
blood Pb concentration, others have failed to show such association
when controlling for confounding factors such as tobacco smoking,
exercise, body weight, alcohol consumption, and socioeconomic
status. Thus, the studies that have employed blood Pb level as an
index of exposure have shown a relatively weak association with
blood pressure. In contrast, the majority of the more recent studies
employing bone Pb level have found a strong association between
long-term Pb exposure and arterial pressure (Chapter 6). Since the
residence time of Pb in the blood is relatively short but very long
in the bone, the latter observations have provided rather compelling
evidence for a positive relationship between Pb exposure and a
subsequent rise in arterial pressure'' (CD, pp. 5-102 to 5-103).
Further, in consideration of the meta-analysis also described here,
the Criteria Document stated that ``The meta-analysis provides
strong evidence for an association between increased blood Pb and
increased blood pressure over a wide range of populations'' (CD, p.
6-130) and ``the meta-analyses results suggest that studies not
detecting an effect may be due to small sample sizes or other
factors affecting precision of estimation of the exposure effect
relationship'' (CD, p. 6-133).
Studies of nationally representative U.S. samples observed associations
between blood Pb levels and increased systolic blood pressure at
population mean blood Pb levels less than 5 [mu]g/dL, particularly
among African Americans (CD, Section 6.5.2). With regard to gender
---------------------------------------------------------------------------
differences, the Criteria Document states the following (CD, p. 6-154).
Although females often show lower Pb coefficients than males,
and Blacks higher Pb coefficients than Whites, where these
differences have been formally tested, they are usually not
statistically significant. The tendencies may well arise in the
differential Pb exposure in these strata, lower in women than in
men, higher in Blacks than in Whites. The same sex and race
differential is found with blood pressure.
Animal evidence provides confirmation of Pb effects on cardiovascular
functions (CD, Sections 5.5, 6.5, 8.4.3 and 8.6.3).
Renal effects, evidenced by reduced renal filtration, have
also been associated with Pb exposures indexed by bone Pb levels and
also with mean blood Pb levels in the range of 5 to 10 [mu]g/dL in the
general adult population, with the potential adverse impact of such
effects being enhanced for susceptible subpopulations including those
with diabetes, hypertension, and chronic renal insufficiency (CD,
Sections 6.4, 8.4.5, and 8.6.4). The full significance of this effect
is unclear,
[[Page 29207]]
given that other evidence of more marked signs of renal dysfunction
have not been detected at blood Pb levels below 30-40 [mu]g/dL in large
studies of occupationally exposed Pb workers (CD, pp. 6-270 and 8-
50).\75\
---------------------------------------------------------------------------
\75\ In the general population, both cumulative and circulating
Pb has been found to be associated with longitudinal decline in
renal functions. In the large NHANES III study, alterations in
urinary creatinine excretion rate (one indicator of possible renal
dysfunction) were observed in hypertensives at a mean blood Pb of
only 4.2 [mu]g/dL. These results provide suggestive evidence that
the kidney may well be a target organ for effects from Pb in adults
at current U.S. environmental exposure levels. The magnitude of the
effect of Pb on renal function ranged from 0.2 to -1.8 mL/min change
in creatinine clearance per 1.0 [mu]g/dL increase in blood Pb in
general population studies. However, the full significance of this
effect is unclear, given that other evidence of more marked signs of
renal dysfunction have not been detected at blood Pb levels below
30-40 [mu]g/dL among thousands of occupationally exposed Pb workers
that have been studied (CD, p. 6-270).
---------------------------------------------------------------------------
Other Pb associated effects in adults occurring at or just
above 10 [mu]g/dL include hematological (e.g., impact on heme synthesis
pathway) and neurological effects, with animal evidence providing
support of Pb effects on these systems and evidence regarding mechanism
of action (CD, Sections 5.2, 5.3, 6.3 and 6.9.2).
C. Human Exposure and Health Risk Assessments
This section presents a brief summary of the human exposure and
health risk assessments conducted by EPA for this review. The complete
full-scale assessment, which includes specific analyses conducted to
address CASAC comments and advice on an earlier draft assessment, is
presented in the final Risk Assessment Report (USEPA, 2007b).
The focus of this Pb NAAQS risk assessment is on characterizing
risk resulting from exposure to policy-relevant Pb (i.e., exposure to
Pb that has passed through ambient air on its path from source to human
exposure--as described in section II.A.2). The design and
implementation of this assessment needed to address significant
limitations and complexity that go far beyond the situation for similar
assessments typically performed for other criteria pollutants. Not only
was the risk assessment constrained by the timeframe allowed for this
review in the context of breadth of information to address, it was also
constrained by significant limitations in data and modeling tools for
the assessment, as discussed further in section II.C.2.h below.
Furthermore, the multimedia and persistent nature of Pb, and the role
of multiple exposure pathways (discussed in section II.A), add
significant complexity to the assessment as compared to other
assessments that focus only on the inhalation pathway. The impact of
this on our estimates for air-related exposure pathways is discussed in
section II.C.2.e.
The remainder of this overview of the human health risk assessment
is organized as follows. An overview of the human health risk
assessment completed in the last review of the Pb NAAQS in 1990 (USEPA,
1990a) is presented first. Next, design aspects of the current risk
assessment are presented, including: (a) CASAC advice regarding the
design of the risk assessment, (b) description of health endpoints and
associated risk metrics modeled, including the concentration-response
functions used, (c) overview of the case study approach employed, (d)
description of air quality scenarios modeled, (e) explanation of air-
related versus background classification of risk results in the context
of this analysis, (f) overview of analytical (modeling) steps completed
for the risk assessment and (g) description of the multiple sets of
risk results generated for the analysis. Then, key sources of
uncertainty associated with the analysis are presented. And finally, a
summary of exposure and risk estimates and key observations is
presented.
1. Overview of Risk Assessment From Last Review
The risk assessment conducted in support of the last review used a
case study approach to compare air quality scenarios in terms of their
impact on the percentage of modeled populations that exceeded specific
blood Pb levels chosen with consideration of the health effects
evidence at that time (USEPA, 1990b; USEPA, 1989). The case studies in
that analysis, however, focused exclusively on Pb smelters including
two secondary and one primary smelter and did not consider exposures in
a more general urban context. The analysis focused on children (birth
through 7 years of age) and middle-aged men. The assessment evaluated
impacts of alternate NAAQS on numbers of children and men with blood Pb
levels above levels of concern based on health effects evidence at that
time. The primary difference between the risk assessment approach used
in the current analysis and the assessment completed in 1990 involves
the risk metric employed. Rather than estimating the percentage of
study populations with exposures above blood Pb levels of interest as
was done in the last review (i.e., 10, 12 and 15 [mu]g/dL), the current
analysis estimates changes in health risk, specifically IQ loss,
associated with Pb exposure for child populations at each of the case
study locations with that estimated IQ loss further differentiated
between air-related and background Pb exposure categories.
2. Design Aspects of Exposure and Risk Assessments
This section provides an overview of key elements of the assessment
design, inputs, and methods, and includes identification of key
uncertainties and limitations.
a. CASAC Advice
The CASAC conducted a consultation on the draft analysis plan for
the risk assessment (USEPA, 2006c) in June, 2006 (Henderson, 2006).
Some key comments provided by CASAC members on the plan included: (1)
Placing a higher priority on modeling the child IQ metric than the
adult endpoints (e.g., cardiovascular effects), (2) recognizing the
importance of indoor dust loading by Pb contained in outdoor air as a
factor in Pb-related exposure and risk for sources considered in this
analysis, and (3) concurring with use of the IEUBK biokinetic blood Pb
model. Taking these comments into account, a pilot phase assessment was
conducted to test the risk assessment methodology being developed for
the subsequent full-scale assessment. The pilot phase assessment is
described in the first draft Staff Paper and accompanying technical
report (ICF 2006), which was discussed by the CASAC Pb panel on
February 6-7 (Henderson, 2007a).
Results from the pilot assessment, together with comments received
from CASAC and the public, informed the design of the full-scale
analysis. The full-scale analysis included a substitution of a more
generalized urban case study for the location-specific near-roadway
case study evaluated in the pilot. In addition, a number of changes
were made in the exposure and risk assessment approaches, including the
development of a new indoor dust Pb model focused specifically on urban
residential locations and specification of additional IQ loss
concentration-response (C-R) functions to provide greater coverage for
potential impacts at lower exposure levels.
The draft full-scale assessment was presented in the July 2007
draft risk assessment report (USEPA, 2007a) that was released for
public comment and provided to CASAC for review. In their review of the
July draft risk assessment report, the CASAC Pb Panel made several
recommendations for additional exposure and health risk analyses
(Henderson, 2007b). These included a recommendation that the general
urban
[[Page 29208]]
case study be augmented by the inclusion of risk analyses in specific
urban areas of the U.S. In this regard, they specifically stated the
following (Henderson, 2007b, p. 3)
* * * the CASAC strongly believes that it is important that EPA
staff make estimates of exposure that will have national
implications for, and relevance to, urban areas; and that,
significantly, the case studies of both primary lead (Pb) smelter
sites as well as secondary smelter sites, while relevant to a few
atypical locations, do not meet the needs of supporting a Lead
NAAQS. The Agency should also undertake case studies of several
urban areas with varying lead exposure concentrations, based on the
prototypic urban risk assessment that OAQPS produced in the 2nd
Draft Lead Human Exposure and Health Risk Assessments. In order to
estimate the magnitude of risk, the Agency should estimate exposures
and convert these exposures to estimates of blood levels and IQ loss
for children living in specific urban areas.
Hence, EPA included additional case studies in the risk assessment
focused on characterizing risk for residential populations in three
specific urban locations. Further, CASAC recommended using a
concentration-response function with a change in slope near 7.5 [mu]g/
dL. Accordingly, EPA included such an additional concentration-response
function in the risk assessment. Results from the initial full-scale
analyses, along with comments from CASAC, such as those described here,
and the public resulted in a final version of the full-scale
assessments which is briefly summarized here and presented in greater
detail in the Risk Assessment Report and associated appendices (USEPA,
2007b).
In their review of the final risk assessment, CASAC expressed
strong support, stating as follows (Henderson, 2008a, p. 4):
The Final Risk Assessment report captures the breadth of issues
related to assessing the potential public health risk associated
with lead exposures; it competently documents the universe of
knowledge and interpretations of the literature on lead toxicity,
exposures, blood lead modeling and approaches for conducting risk
assessments for lead.
b. Health Endpoint, Risk Metric and Concentration-Response Functions
The health endpoint on which the quantitative health risk
assessment focuses is developmental neurotoxicity in children, with IQ
decrement (or loss) as the risk metric. Among the wide variety of
health endpoints associated with Pb exposures, there is general
consensus that the developing nervous system in young children is the
most sensitive and that neurobehavioral effects (specifically
neurocognitive deficits), including IQ decrements, appear to occur at
lower blood levels than previously believed (i.e., at levels <10 [mu]g/
dL). The selection of children's IQ for the quantitative risk
assessment reflects consideration of the evidence presented in the
Criteria Document as well as advice received from CASAC (Henderson,
2006, 2007a).
Given the evidence described in detail in the Criteria Document
(Chapters 6 and 8), and in consideration of CASAC recommendations
(Henderson, 2006, 2007a, 2007b), the risk assessment for this review
relies on the functions presented by Lanphear and others (2005) that
relate absolute IQ as a function of concurrent blood Pb or of the log
of concurrent blood Pb, and lifetime average blood Pb, respectively. As
discussed in the Criteria Document (CD, p. 8-63 to 8-64), the slope of
the concentration-response relationship described by these functions is
greater at the lower blood Pb levels (e.g., less than 10 [mu]g/dL). As
discussed in the Criteria Document and summarized in section II.B.2,
threshold blood Pb levels for these effects cannot be discerned from
the currently available epidemiological studies, and the evidence in
the animal Pb neurotoxicity literature does not define a threshold for
any of the toxic mechanisms of Pb (CD, Sections 5.3.7 and 6.2).
In applying relationships observed with the international pooled
analysis by Lanphear and others (2005) to the risk assessment, which
includes blood Pb levels below the range represented by the pooled
analysis, several alternative blood Pb concentration-response models
were considered in recognition of a reduced confidence in our ability
to characterize the quantitative blood Pb concentration-response
relationship at the lowest blood Pb levels represented in the recent
epidemiological studies. The functions considered and employed in the
initial risk analyses for this review include the following.
Log-linear function with low-exposure linearization, for
both concurrent and lifetime average blood metrics, applies the
nonlinear relationship down to the blood Pb concentration representing
the lower bound of blood Pb levels for that blood metric in the pooled
analysis and applies the slope of the tangent at that point to blood Pb
concentrations estimated in the risk assessment to fall below that
level.
Log-linear function with cutpoint, for both concurrent and
lifetime average blood metrics, also applies the nonlinear relationship
at blood Pb concentrations above the lower bound of blood Pb
concentrations in the pooled analysis dataset for that blood metric,
but then applies zero risk to all lower blood Pb concentrations
estimated in the risk assessment (this cutpoint is 1 [mu]g/dL for the
concurrent blood Pb).
In the additional risk analyses performed subsequent to the August
2007 CASAC public meeting, the two functions listed above and the
following two functions were employed (details on the forms of these
functions as applied in this risk assessment are described in Section
5.3.1 of the Risk Assessment Report).
Population stratified dual linear function for concurrent
blood Pb, derived from the pooled dataset stratified at peak blood Pb
of 10 [mu]g/dL \76\ and
---------------------------------------------------------------------------
\76\ As mentioned above (section II.B.2.b), this function
(derived for lifetime average blood Pb), was used in the economic
analysis for the RRP rule. This model was selected for the RRP
economic analysis with consideration of advice from CASAC and of the
distribution of blood Pb levels being considered in that analysis,
which focused on children living in houses with lead-based paint
(USEPA, 2008). With consideration of these blood Pb levels, the
economic analysis document states that ``[s]electing a model with a
node, or changing one segment to the other, at a lifetime average
blood Pb concentration of 10 [mu]g/dL rather than at 7.5 [mu]g/dL,
is a small protection against applying an incorrectly rapid change
(steep slope with increasingly smaller effect as concentrations
lower) to the calculation'' (USEPA, 2008).
---------------------------------------------------------------------------
Population stratified dual linear function for concurrent
blood Pb, derived from the pooled dataset stratified at 7.5 [mu]g/dL
peak blood Pb.
In interpreting risk estimates derived using the various functions,
consideration should be given to the uncertainties with regard to the
precision of the coefficients used for each analysis. The coefficients
for the log-linear model from Lanphear et al. (2005) had undergone a
careful development process, including sensitivity analyses, using all
available data from 1,333 children. The shape of the exposure-response
relationship was first assessed through tests of linearity, then by
evaluating the restricted cubic spline model. After determining that
the log-linear model provided a good fit to the data, covariates to
adjust for potential confounding were included in the log-linear model
with careful consideration of the stability of the parameter estimates.
After the multiple regression models were developed, regression
diagnostics were employed to ascertain whether the Pb coefficients were
affected by collinearity or influential observations. To further
investigate the stability of the model, a random-effects model (with
sites
[[Page 29209]]
random) was applied to evaluate the results and also the effect of
omitting one of the seven cohorts on the Pb coefficient. In the various
sensitivity analyses performed, the coefficient from the log-linear
model was found to be robust and stable. The log-linear model, however,
is not biologically plausible at the very lowest blood Pb
concentrations as they approach zero; therefore, in the first two
functions the log-linear model is applied down to a cutpoint (of 1
[mu]g/dL for the concurrent blood Pb metric), selected based on the low
end of the blood Pb levels in the pooled dataset, followed by a
linearization or an assumption of zero risk at levels below that point.
In contrast, the coefficients from the two analyses using the
population stratified dual linear function with stratification at 7.5
[mu]g/dL and 10 [mu]g/dL,\77\ peak blood Pb, have not undergone as
careful development. These analyses were primarily done to compare the
lead-associated decrement at lower blood Pb concentrations and higher
blood Pb concentrations. For these analyses, the study population was
stratified at the specified peak blood Pb level and separate linear
models were fitted to the concurrent blood Pb data for the children in
the two study population subgroups.\78\ While these analyses are quite
suitable for the purpose of investigating whether the slope at lower
concentration levels is greater compared to higher concentration
levels, use of such coefficients as the primary C-R function in a risk
analysis such as this may be inappropriate. Further, only 103 children
had maximal blood Pb levels less than 7.5 [mu]g/dL and 244 children had
maximal blood Pb levels less than 10 [mu]g/dL. While these children may
better represent current blood Pb levels, not fitting a single model
using all available data may lead to bias. Slob et al. (2005) noted
that the usual argument for not considering data from the high dose
range is that different biological mechanisms may play a role at higher
doses compared to lower doses. However, this does not mean a single
curve across the entire exposure range cannot describe the
relationship. The fitted curve merely assumes that the underlying dose-
response follows a smooth curve over the whole dose range. If
biological mechanisms change when going from lower to higher doses,
this change will result in a gradually changing slope of the dose-
response. The major strength of the Lanphear et al. (2005) study was
the large sample size and the pooled analysis of data from seven
different cohorts. In the case of the study population subgroup with
peak blood Pb below 7.5 [mu]g/dL, less than 10% of the available data
is used in the analysis (103 of the 1333 subjects in the pooled
dataset), with more than half of the data coming from one cohort
(Rochester) and the six other cohorts contributing zero to 13 children
to the analysis. Such an analysis consequently does not make full use
of the strength of the pooled study by Lanphear and others (2005).
---------------------------------------------------------------------------
\77\ See previous footnote.
\78\ Neither fit of the model nor other sensitivity analyses
were conducted (or reported) for these coefficients.
---------------------------------------------------------------------------
In consideration of the preceding discussion and the range of blood
Pb levels assessed in this analysis,\79\ greater confidence is placed
in the log-linear model form compared to the dual-linear stratified
models for purposes of the risk assessment described in this notice.
Further, in considering risk estimates derived from the four core
functions (log-linear function with low-exposure linearization, log-
linear function with cutpoint, dual linear function, stratified at 7.5
[mu]g/dL peak blood Pb, and dual linear function, stratified at 10
[mu]g/dL peak blood Pb), greatest confidence is assigned to risk
estimates derived using the log-linear function with low-exposure
linearization since this function (a) is a nonlinear function that
describes greater response per unit blood Pb at lower blood Pb levels
consistent with multiple studies identified in the discussion above,
(b) is based on fitting a function to the entire pooled dataset (and
hence uses all of the data in describing response across the range of
exposures), (c) is supported by sensitivity analyses showing the model
coefficients to be robust, and (d) provides an approach for predicting
IQ loss at the lowest exposures simulated in the assessment (consistent
with the lack of evidence for a threshold). Note, however, that risk
estimates generated using the other three concentration-response
functions are also presented to provide perspective on the impact of
uncertainty in this key modeling step. We additionally note that the
CASAC Pb Panel recommended that C-R function derived from the pooled
dataset stratified at 7.5 [mu]g/dL, peak blood Pb, be given weight in
this analysis (Henderson, 2008).
---------------------------------------------------------------------------
\79\ The median concurrent values in all case studies and air
quality scenarios are below 5 [mu]g/dL and those for air quality
scenarios within the range of standard levels proposed in this
notice are below 3 [mu]g/dL (as shown in Table 1).
---------------------------------------------------------------------------
c. Case Study Approach
For the risk assessment described in this notice, a case study
approach was employed as described in Sections 2.2 (and subsections)
and 5.1.3 of the Risk Assessment Report (USEPA, 2007b). In summarizing
the assessment in this proposal, we have focused on five \80\ case
studies that generally represent two types of population exposures: (1)
More highly air-pathway exposed children (as described below) residing
in small neighborhoods or localized residential areas with air
concentrations somewhat near the standard level being evaluated, and
(2) urban populations with a broader range of air-related exposures.
These five case studies are:
---------------------------------------------------------------------------
\80\ A sixth case study (the secondary Pb smelter case study) is
also described in the Risk Assessment Report. However, as discussed
in Section 4.3.1 of that document (USEPA, 2007b), significant
limitations in the approaches employed for this case study have
contributed to large uncertainties in the corresponding estimates.
---------------------------------------------------------------------------
A general urban case study: This case study is not based
on a specific geographic location and reflects several simplifying
assumptions used in representing exposure including uniform ambient air
Pb levels associated with the standard of interest across the
hypothetical study area and a uniform study population. This case study
characterizes risk for a localized part of an urban area at different
standard levels, but based on national average estimates of the
relationships between the different standard form assessed and ambient
air exposure concentrations. Thus, while this provides characterization
of risk to children that are relatively more highly air pathway exposed
(as compared to the location-specific case studies), this case study is
not considered to represent a high-end scenario with regard to the
characterization of ambient air Pb levels and associated risk.\81\
---------------------------------------------------------------------------
\81\ In representing the different forms of each standard level
assessed (maximum monthly or maximum quarterly) as annual air
concentrations for input to the blood Pb model for this case study,
however, we relied on averages of these relationships for large
urban areas nationally. As the averages are higher than the medians,
localized areas near more than half the urban monitoring locations
would have higher exposures and associated risks than those reported
for this case study. Further, we note that exposure concentrations
would be twice those used here if the 25th percentile values for
these relationships had been used in place of the averages. For this
reason, this case study should not be interpreted as representing a
high-end scenario with regard to the characterization of ambient air
Pb levels and associated risk.
---------------------------------------------------------------------------
A primary Pb smelter case study: \82\ This case study
estimates risk for children living in an area currently not in
attainment with the current NAAQS that is impacted by Pb emissions from
a primary Pb smelter. Results described
[[Page 29210]]
here are those for the area within 1.5 km of the facility (the
``subarea'') where airborne Pb concentrations are closest to the
current standard. As such, this case study characterizes risk for a
specific more highly exposed population and also provides insights on
risk to child populations living in areas near large sources of Pb
emissions.\83\
---------------------------------------------------------------------------
\82\ See Section II.C.2.a for a summary of CASAC's comment with
regard to the primary and secondary Pb smelter case studies.
\83\ Result for the full study area, which extends 10 km out
from the facility, are presented in the Risk Assessment Report
(USEPA, 2007a), but are not presented here. Exposures in the full
study area were dominated by modeled children farther from the
facility where, as discussed in the ANPR (section III.B.2.h), there
is likely underestimation of ambient air-related Pb exposure due to
increasing influence of other sources relative to that of the
facility, which were not included in the dispersion modeling
performed to estimate air concentrations for this case study.
---------------------------------------------------------------------------
Three location-specific urban case studies: These urban
case studies focus on specific urban areas (Cleveland, Chicago and Los
Angeles) to provide representations of the distribution of ambient air-
related risk in specific densely populated urban locations. These case
studies represent areas with specific population distributions and that
experience a broader range of air-related exposures due both to
potential spatial gradients in ambient air Pb levels and population
density. A large majority of the population in these case studies
resides in areas with much lower air concentrations than those in the
very small subareas of these case studies with the highest
concentrations. Ambient air Pb concentrations are characterized using
source-oriented and other Pb-TSP monitors in these cities, while
location-specific U.S. Census demographic data are used to characterize
the spatial distribution of residential child populations in these
study areas.
These different case studies generally represent two types of
population exposures. The general urban and primary Pb smelter subarea
provide estimates of risk for more highly air-pathway exposed children
residing in small neighborhoods or localized residential areas with air
concentrations somewhat near the standard level being evaluated. By
contrast, the three location-specific urban case studies included in
the analysis provide risk estimates for an urban population with a
broader range of air-related exposures. In fact, for the location-
specific urban case studies, the majority of the modeled populations
experience ambient air Pb levels significantly lower than the standard
level being evaluated, with only a small population experiencing
ambient air Pb levels at or near the standard.\84\
---------------------------------------------------------------------------
\84\ Based on the nature of the population exposures represented
by the two categories of case study, the first category (the general
urban and primary Pb smelter case studies) relates more closely to
the second evidence-based framework (see Sections II.D.2.a and
II.E.3.a) with regard to estimates of air-related IQ loss. As
mentioned above these case studies, as compared to the other
category of case studies, include populations that are relatively
more highly air pathway exposed to air Pb concentrations somewhat
near the standard level evaluated.
---------------------------------------------------------------------------
In considering risk results generated for the location-specific
urban case studies, we note that, given the wide range of monitored Pb
levels in urban areas, combined with the relatively limited monitoring
network characterizing ambient levels in the urban setting, it is not
possible to determine where these case studies fall within the
distribution of ambient air-related risk in U.S. cities.
d. Air Quality Scenarios
Air quality scenarios assessed include (a) a current conditions
scenario for the location-specific urban case studies and the general
urban case study, (b) a current NAAQS scenario for the location-
specific urban case studies, the general urban case study and the
primary Pb smelter case study, and (c) a range of alternative NAAQS
scenarios for all case studies. The alternative NAAQS scenarios include
levels of 0.5, 0.2, 0.05, and 0.02 [mu]g/m\3\, with a monthly averaging
time, as well as a level of 0.2 [mu]g/m\3\ scenario using a quarterly
averaging time.\85\
---------------------------------------------------------------------------
\85\ For further discussion of the air quality scenarios and
averaging times included in the risk assessment, see section 2.3.1
of the Risk Assessment Report (USEPA, 2007b).
---------------------------------------------------------------------------
The current NAAQS scenario for the urban case studies assumes
ambient air Pb concentrations higher than those currently occurring in
nearly all urban areas nationally.\86\ While it is extremely unlikely
that Pb concentrations in urban areas would rise to meet the current
NAAQS and there are limitations and uncertainties associated with the
roll-up procedure used for the location-specific urban case studies (as
described in Section III.B.2.h below), this scenario was included for
those case studies to provide perspective on potential risks associated
with raising levels to the point that the highest level across the
study area just meets the current NAAQS. When evaluating these results
it is important to keep these limitations and uncertainties in mind.
---------------------------------------------------------------------------
\86\ This scenario was simulated for the location-specific urban
case studies using a proportional roll-up procedure. For the general
urban case study, the maximum quarterly average ambient air
concentration was set equal to the current NAAQS.
---------------------------------------------------------------------------
Current conditions for the three location-specific urban case
studies in terms of maximum quarterly average air Pb concentrations are
0.09, 0.14 and 0.36 [mu]g/m\3\ for the study areas in Los Angeles,
Chicago and Cleveland, respectively. In terms of maximum monthly
average the values are 0.17 [mu]g/m\3\, 0.31 [mu]g/m\3\ and 0.56 [mu]g/
m\3\ for the study areas in Los Angeles, Chicago and Cleveland,
respectively.
Details of the assessment scenarios, including a description of the
derivation of Pb concentrations for air and other media are presented
in Sections 2.3 (and subsections) and Section 5.1.1 of the Risk
Assessment Report (USEPA, 2007b).
e. Categorization of Policy-Relevant Exposure Pathways
As discussed in Section IIA, this review focuses on air-related
exposure pathways (i.e., those pathways where Pb passes through ambient
air on its path from source to human exposure). These include both
inhalation of ambient air Pb (including both Pb emitted directly into
ambient air as well as resuspended Pb); and ingestion of Pb that, once
airborne, has made its way into indoor dust, outdoor dust or soil,
dietary items (e.g., crops and livestock), and drinking water. Because
of the nonlinear response of blood Pb to exposure (simulated in the
IEUBK blood Pb model) and also the nonlinearity reflected in the C-R
functions for estimation of IQ loss, this assessment first estimates
total blood Pb and risk (air- and nonair-related), and then separates
out those estimates of blood Pb and associated risk associated with the
pathways of interest in this review.
To separate out risk for the pathways of interest in this review,
we split the estimates of total (all-pathway) blood Pb and IQ loss into
background and two air-related categories (referred to as ``recent
air'' and ``past air''). However, significant limitations in our
modeling tools and data resulted in an inability to parse specific risk
estimates into specific pathways, such that we have approximated
estimates for the air-related and background categories.
Those Pb exposure pathways identified in section II.A.2 as being
tied most directly to ambient air, which consequently have the
potential to respond relatively more quickly to changes in air Pb
(inhalation and ingestion of indoor dust loaded directly from ambient
air Pb) were placed into the ``recent air'' category. The other air-
related Pb exposure pathways, associated with atmospheric deposition,
were placed into the ``past air'' category. These include ingestion of
Pb in
[[Page 29211]]
outdoor dust/soil and ingestion of the portion of Pb in indoor dust
that after deposition from ambient air outdoors is carried indoors with
humans (as described in section II.A.2 above).\87\
---------------------------------------------------------------------------
\87\ As discussed below, due to technical limitations related to
indoor dust Pb modeling, dust from Pb paint may be included to some
extent in the ``past air'' category of exposure pathways.
---------------------------------------------------------------------------
Thus, total blood Pb and IQ loss estimates were apportioned into
the following pathways or pathway combinations:
Inhalation of ambient air Pb (i.e., ``recent air'' Pb):
This is derived using the blood Pb estimate resulting from Pb exposure
limited to the inhalation pathway (and includes inhalation of Pb in
ambient air from all sources contributing to the ambient air
concentration estimate, including potentially resuspension).
Ingestion of ``recent air'' indoor dust Pb: This is
derived using the blood Pb estimate resulting from Pb exposure limited
to ingestion of the Pb in indoor dust that is predicted in this
assessment from infiltration of ambient air indoors and subsequent
deposition.\88\
---------------------------------------------------------------------------
\88\ Recent air indoor dust Pb was estimated using the
mechanistic component of the hybrid blood Pb model (see Section
3.1.4 of the Risk Assessment Report). For the primary Pb smelter
case study, estimates for this pathway are not separated from
estimates for the pathway described in the subsequent bullet due to
uncertainty regarding this categorization with the model used for
this case study (Section 3.1.4.2 of the Risk Assessment Report).
---------------------------------------------------------------------------
Ingestion of ``other'' indoor dust Pb (considered part of
``past air'' exposure): This is derived using the blood Pb estimate
resulting from Pb exposure limited to ingestion of the Pb in indoor
dust that is not predicted from infiltration of ambient air indoors and
subsequent deposition.\89\ This is interpreted to represent indoor
paint, outdoor soil/dust, and additional sources of Pb to indoor dust
including historical air (as discussed in the Risk Assessment Report,
Section 2.4.3). As the intercept in regression dust models will be
inclusive of error associated with the model coefficients, this
category also includes some representation of dust Pb associated with
current ambient air concentrations (described in previous bullet). For
the primary Pb smelter case study, estimates for this pathway are not
separated from estimates for the pathway described above due to
uncertainty regarding this categorization with the model used for this
case study (Risk Assessment Report, Section 3.1.4.2). This pathway is
included in the ``past air'' category.
---------------------------------------------------------------------------
\89\ ``Other'' indoor dust Pb is estimated using the intercept
in the dust models plus that predicted by the outdoor soil
concentration coefficient (for models that include soil Pb as a
predictor of indoor dust Pb) (Section 3.1.4 of the Risk Assessment
Report).
---------------------------------------------------------------------------
Ingestion of outdoor soil/dust Pb: This is derived using
the blood Pb estimate resulting from Pb exposure limited to ingestion
of outdoor soil/dust Pb. This pathway is included in the ``past air''
category (and could include contamination from historic Pb emissions
from automobiles and Pb paint).
Ingestion of drinking water Pb: This is derived using the
blood Pb estimate resulting from Pb exposure limited to ingestion of
drinking water Pb. This pathway is included in the policy-relevant
background category.
Ingestion of dietary Pb: This is derived using the blood
Pb estimate resulting from Pb exposure limited to ingestion of dietary
Pb. This pathway is included in the policy-relevant background
category.
As noted above, significant limitations in our modeling tools and
data resulted in an inability to parse risk estimates for specific
pathways, such that we approximated estimates for the air-related and
background categories. Of note in this regard is the apportionment of
background (nonair) pathways. For example, while conceptually indoor Pb
paint contributions to indoor dust Pb would be considered background
and included in the ``background'' category for this assessment, due to
technical limitations related to indoor dust Pb modeling, ultimately,
dust from Pb paint was included as part of ``other'' indoor dust Pb
(i.e., as part of past air exposure). The inclusion of indoor lead Pb
as a component of ``other'' indoor air (and consequently as a component
of the ``past air'' category) represents a source of potential high
bias in our prediction of exposure and risk associated with the ``past
air'' category because conceptually, exposure to indoor paint Pb is
considered part of background exposure. Further, Pb in ambient air does
contribute to the exposure pathways included in the ``background''
category (drinking water and diet), and is likely a substantial
contribution to diet (CD, p. 3-48). But we could not separate the air
contribution from the nonair contributions, and the total contribution
from both the drinking water and diet pathways are categorized as
``background'' in this assessment. As a result, our ``background'' risk
estimate includes some air-related risk.
Further, we note that in simulating reductions in exposure
associated with reducing ambient air Pb levels through alternative
NAAQS (and increases in exposure if the current NAAQS was reached in
certain case studies) only the exposure pathways categorized as
``recent air'' (inhalation and ingestion of that portion of indoor dust
associated with outdoor ambient air) were varied with changes in air
concentration. The assessment did not simulate decreases in ``past
air'' exposure pathways (e.g., reductions in outdoor soil Pb levels
following reduction in ambient air Pb levels and a subsequent decrease
in exposure through incidental soil ingestion and the contribution of
outdoor soil to indoor dust). These exposures were held constant across
all air quality scenarios. In comparing total risk estimates between
alternate NAAQS scenarios, this aspect of the analysis will tend to
underestimate the reductions in risk associated with alternative NAAQS.
However, this does not mean that overall risk has been underestimated.
The net effect of all sources of uncertainty or bias in the analysis,
which may also tend to under- or overestimate risk, could not be
quantified. Interpretation of risk estimates is discussed more fully in
section II.C.3.b.
In summary, because of limitations in the assessment design, data
and modeling tools, our risk estimates for the ``past air'' category
include both risks that are truly air-related and potentially, some
background risk. Because we could not sharply separate Pb linked to
ambient air from Pb that is background, some of the three categories of
risk are underestimated and others overestimated. On balance, we
believe this limitation leads to a slight overestimate of the risks in
the ``past air'' category. At the same time, as discussed above, the
``recent air'' category does not fully represent the risk associated
with all air-related pathways. Thus, we consider the risk attributable
to air-related exposure pathways to be bounded on the low end by the
risk estimated for the ``recent air'' category and on the upper end by
the risk estimated for the ``recent air'' plus ``past air'' categories.
f. Analytical Steps
The risk assessment includes four analytical steps, briefly
described below and presented in detail in Sections 2.4.4, 3.1, 3.2,
4.1, and 5.1 of the Risk Assessment Report (USEPA, 2007b).
Characterization of Pb in ambient air: The
characterization of outdoor ambient air Pb levels uses different
approaches depending on the case study (as explained in more detail
below): (a) source-oriented and non-source oriented monitors are
assumed to represent different exposure zones in the city-specific case
studies, (b) a single exposure level is assumed for the entire
[[Page 29212]]
population in the general urban case study, and (c) ambient levels are
estimated using air dispersion modeling based on Pb emissions from a
particular facility in the primary Pb smelter case study.
Characterization of outdoor soil/dust and indoor dust Pb
concentrations: Outdoor soil Pb levels are estimated using empirical
data and fate and transport modeling. Indoor dust Pb levels are
predicted using a combination of (a) regression-based models that
relate indoor dust to ambient air Pb and outdoor soil Pb, and (b)
mechanistic models.\90\
---------------------------------------------------------------------------
\90\ Indoor dust Pb modeling for the urban case studies is based
on a hybrid mechanistic-empirical model which considers the direct
impact of Pb in ambient air on indoor dust Pb (i.e., which models
the infiltration of ambient air indoors and subsequent deposition of
Pb to indoor surfaces). This modeling does not consider other
ambient air-related contributions to indoor dust, such as ``tracking
in'' of outdoor soil Pb. By contrast, indoor dust Pb modeling for
the primary Pb smelter case study subarea uses a site-specific
regression model which relates average dust Pb values (based on a
recent multi-year dataset) to annual average air Pb concentrations
(based on air dispersion modeling). In this way, modeling for the
primary Pb smelter subarea may reflect some contributions to indoor
dust Pb that relate to longer term impacts of ambient air (e.g.,
``tracking in'' of outdoor soil), as well as contributions from
infiltration of ambient air. Additional detail on the methods used
in characterizing Pb concentrations in outdoor soil and indoor dust
are presented in Sections 3.1.3 and 3.1.4 of the Risk Assessment,
respectively. Data, methods and assumptions here used in
characterizing Pb concentrations in these exposure media may differ
from those in other analyses that serve different purposes.
---------------------------------------------------------------------------
Characterization of blood Pb levels: Blood Pb levels for
each exposure zone are derived from central-tendency blood Pb
concentrations estimated using the Integrated Exposure and Uptake
Biokinetic (IEUBK) model, and concurrent or lifetime average blood Pb
is estimated from these outputs as described in Section 3.2.1.1 of the
Risk Assessment Report (USEPA, 2007b). For the point source and
location-specific urban case studies, a probabilistic exposure model is
used to generate population distributions of blood Pb concentrations
based on: (a) The central tendency blood Pb levels for each exposure
zone, (b) demographic data for the distribution of children (less than
7 years of age) across exposure zones in a study area, and (c) a
geometric standard deviation (GSD) intended to characterize
interindividual variability in blood Pb (e.g., reflecting differences
in behavior and biokinetics related to Pb). For the general urban case
study, as demographic data for a specific location are not considered,
the GSD is applied directly to the central tendency blood Pb level to
estimate a population distribution of blood Pb levels. Additional
detail on the methods used to model population blood Pb levels is
presented in Sections 3.2.2 and 5.2.2.3 of the Risk Assessment Report
(USEPA, 2007b).
Risk characterization (estimating IQ loss): Concurrent or
lifetime average blood Pb estimates for each simulated child in each
case study population are converted into total Pb-related IQ loss
estimates using the concentration-response functions described above in
section II.C.2.b.\91\
---------------------------------------------------------------------------
\91\ The four C-R functions applied in the risk assessment,
which are based on analyses presented in Lanphear et al. (2005)
include a log-linear function with low-exposure linearization, a
log-linear function with a cutpoint, and two dual linear functions
(based on population stratification at peak blood Pb levels of 7.5
and 10 [mu]g/dL) (see section II.C.2.b).
---------------------------------------------------------------------------
We have also used the results of exposure modeling to estimate air-
to-blood ratios for two of the case studies (the general urban and
primary Pb smelter case studies). Specifically, we compared the change
in ambient air Pb between adjacent NAAQS levels with the associated
reduction in concurrent blood Pb levels (for the median population
percentile) to derive air-to-blood ratios. As they relate air
concentrations \92\ input to the first analytical step to blood Pb
estimates output from the third analytical step, they may be viewed as
a collapsed alternate to the three steps for the exposure pathways
directly linked to air concentrations in this assessment. The values
for these ratios are affected by design aspects of the risk assessment,
most notably those identified here:
---------------------------------------------------------------------------
\92\ Because the IEUBK blood Pb model runs with an annual time
step, the air concentrations input to the ``recent air'' pathways
modeling steps were in terms of annual average air concentration.
---------------------------------------------------------------------------
Because they are derived from differences in blood Pb
estimates between air quality scenarios and the only pathways varied
with air quality scenarios are ambient air and indoor dust (as
described in section II.C.2.e above), the exposure pathways reflected
in the ratios are generally the ``recent air'' pathways (described in
section II.C.2.e above), which include inhalation of ambient air and
ingestion of indoor dust loaded by infiltration of ambient air. Ratios
for the primary Pb smelter case study subarea may additionally reflect
some contributions to indoor dust from other ambient air-related
pathways (e.g., ``tracking in'' of soil containing ambient air Pb), yet
still not all air-related pathways. Thus, the air-to-blood ratios
derived for both case studies (described in section II.C.3.a) are lower
than they would be if they reflected all air-related pathways.
The blood Pb estimates used in this calculation are for
the ``concurrent'' metric (i.e., concentrations during the 7th year of
life). Accordingly, the resultant air-to-blood ratios are lower than
they would be if based on blood Pb estimates for the 2nd year of life
(e.g., peak) or estimates averaged over the exposure period.
Key limitations and uncertainties associated with the application
of these specific analytical steps are summarized in Section III.B.2.k
below.
g. Generating Multiple Sets of Risk Results
In the initial analyses for the full-scale assessment (USEPA,
2007a), EPA implemented multiple modeling approaches for each case
study scenario in an effort to characterize the potential impact on
exposure and risk estimates of uncertainty associated with the
limitations in the tools, data and methods available for this risk
assessment and with key analytical steps in the modeling approach.
These multiple modeling approaches are described in Section 2.4.6.2 of
the final Risk Assessment Report (USEPA, 2007b). In consideration of
comments provided by CASAC (Henderson, 2007b) on these analyses
regarding which modeling approach they felt had greater scientific
support, a pared down set of modeling combinations was identified as
the core approach for the subsequent analyses. The core modeling
approach includes the following key elements:
Ambient air Pb estimates (based on monitors or modeling
and proportional rollbacks, as described below),
Background exposure from food and water (as described
above),
The hybrid indoor dust model specifically developed for
urban residential applications (which predicts Pb in indoor dust as a
function of ambient air Pb and nonair contribution),
The IEUBK blood Pb model (which predicts blood Pb in young
children exposed to Pb from multiple exposure pathways),
The concurrent blood Pb metric,
A GSD for concurrent blood Pb of 2.1 to characterize
interindividual variability in blood Pb levels for a given ambient
level for the urban case studies,\93\ and
---------------------------------------------------------------------------
\93\ In the economic analysis for the RRP rule, a GSD of 1.6 was
used in its probabilistic simulations, reflecting the fact that the
simulated exposures focus on a subset of Pb exposure pathways
(exposure to dust and airborne Pb resulting from renovation
activity) and a CASAC recommendation to use the IEUBK-recommended
GSD with the Leggett model, where no GSD is provided. In addition,
the accompanying sensitivity analysis used a GSD of 2.1 to consider
the impact on IQ change estiamtes of using a larger GSD, which would
reflect greater heterogeneity in the study population with regard to
Pb exposure and blood Pb response.
---------------------------------------------------------------------------
[[Page 29213]]
Four different functions relating concurrent blood Pb to
IQ loss (described in section II.C.2.b), including two log-linear
models (one with a cutpoint and one with low-exposure linearization)
and two dual-linear models with stratification, one stratified at 7.5
[mu]g/dL peak blood Pb and the other at 10 [mu]g/dL peak blood Pb.
For each case study, the core modeling approach employs a single
set of modeling elements to estimate exposure and the four different
concentration-response functions referenced above to derive four sets
of risk results from the single set of exposure estimates. The spread
of estimates resulting from application of all four functions captures
much of the uncertainty associated model choice in this analytical
step. Among these four functions, EPA has greater confidence in
estimates derived using the log-linear with low-exposure linearization
concentration-response function as discussed above.
In addition to employing multiple concentration-response functions,
the assessment includes various sensitivity analyses to characterize
the potential impact of uncertainty in other key analysis steps on
exposure and risk estimates. The sensitivity analyses and uncertainty
characterization completed for the risk analysis are described in
Sections 3.5, 4.3, 5.2.5 and 5.3.3 of the Risk Assessment Report
(USEPA, 2007b).
h. Key Limitations and Uncertainties
As recognized above, EPA has made simplifying assumptions in
several areas of this assessment due to the limited data, models, and
time available. These assumptions and related limitations and
uncertainties are described in the Risk Assessment Report (USEPA,
2007b). Key assumptions, limitations and uncertainties are briefly
identified below, with emphasis on those sources of uncertainty
considered most critical in interpreting risk results. In the
presentation below, limitations (and associated uncertainty) are
listed, beginning with those regarding design of the assessment or case
studies, followed by those regarding estimation of Pb concentrations in
ambient air indoor dust, outdoor soil/dust, and blood, and lastly
regarding estimation of Pb-related IQ loss.
Temporal aspects: Exposure modeling uses a 7 year exposure
period for each simulated child, during which time, media
concentrations remain fixed (at levels associated with the ambient air
Pb level being modeled) and the child remains at the same residence,
while exposure factors and physiological parameters are adjusted to
match the age of the child. These aspects are a simplification of
population exposures that contributes some uncertainty to our exposure
and risk estimates.
General urban case study: As described in section
II.C.2.c, this case study is not based on a specific location and is
instead intended to represent a smaller neighborhood experiencing
ambient air Pb levels at or near the standard of interest.
Consequently, it assumes (a) a single exposure zone within which all
media concentrations of Pb are assumed to be spatially uniform and (b)
a uniformly distributed population of unspecified size. While these
assumptions are reasonable in the context of evaluating risk for a
smaller subpopulation located close to a monitor reporting values at or
near the standard of interest, there is significant uncertainty
associated with extrapolating these risks to a specific urban location,
particularly if that urban location is relatively large, given that
larger urban areas are expected to have increasingly varied patterns of
ambient air Pb levels and population density. The risk estimates for
this general urban case study, while generally representative of an
urban residential population exposed to the specified ambient air Pb
levels, cannot be readily related to a specific large urban population.
Location-specific urban case studies: The Pb-TSP
monitoring network is currently quite limited and consequently, the
number of monitors available to represent air concentrations in these
case studies is limited, ranged from six for Cleveland to 11 for
Chicago. Accordingly, our estimates of the magnitude of and spatial
variation of air Pb concentrations are subject to uncertainty
associated with the limited monitoring data and method used in
extrapolating from those data to characterize an ambient air Pb level
surface for these modeled urban areas. Details on the approach used to
derive ambient air Pb surfaces for the urban case studies based on
monitoring data are presented in Section 5.1.3 of the Risk Assessment
Report (USEPA, 2007b). As recognized in Section, III.B.2.a, the
analyses for these case studies were developed in response to CASAC
recommendations on the July 2007 draft Risk Assessment (Henderson,
2007b). Subsequently, the CASAC has reviewed the approach used in
conducting the final draft of the full-scale risk assessment, including
the inclusion of the location-specific urban case studies and expressed
broad support for the technical approach used (Henderson, 2008).
Current NAAQS air quality scenarios: For the location-
specific urban case studies, proportional roll-up procedures were used
to adjust ambient air Pb concentrations up to just meet the current
NAAQS (a detailed discussion is provided in Sections 2.3.1 and 5.2.2.1
of the Risk Assessment Report, USEPA, 2007b). This procedure was used
to provide insights into the degree of risk which could be associated
with ambient air Pb levels at or near the current standard in urban
areas. EPA recognizes that it is extremely unlikely that Pb
concentrations would rise to just meet the current NAAQS in urban areas
nationwide and that there is substantial uncertainty with our
simulation of such conditions. For the primary Pb smelter case study,
where current conditions exceed the current NAAQS, attainment of the
current NAAQS was simulated using air quality modeling, emissions and
source parameters used in developing the 2007 proposed revision to the
State Implementation Plan for the area (described in Section 3.1.1.2 of
the Risk Assessment Report (USEPA, 2007b)).
Alternative NAAQS air quality scenarios: In all case
studies, proportional roll-down procedures were used to adjust ambient
air Pb concentrations downward to attain alternative NAAQS (described
in Sections 2.3.1 and 5.2.2.1 of the Risk Assessment Report, USEPA,
2007b). There is significant uncertainty in simulating conditions
associated with the implementation of emissions reduction actions to
meet a lower standard.
Estimates of outdoor soil/dust Pb concentrations: Outdoor
soil Pb concentration for both the urban case studies and the primary
Pb smelter case study are based on empirical data (as described in
Section 3.1.3 of the Risk Assessment). To the extent that these data
are from areas containing older structures, the impact of Pb paint
weathered from older structures on soil Pb levels will be reflected in
these empirical estimates. In the case of the urban case studies, a
mean value from a sample of houses built between 1940 and 1998 was used
to represent soil Pb levels (as described in Section 3.1.3.1 of the
Risk Assessment). In the case of the primary Pb smelter case study
subarea, site-specific data are used. As there has been remediation of
soil in this subarea, the measurements do not reflect historical air
quality. Additionally,
[[Page 29214]]
studies since remediation have reported increasing soil Pb levels
indicating that soil concentrations are still responding to current air
quality, and consequently underestimate eventual steady state
conditions for the current air quality. In all case studies, the same
outdoor soil/dust Pb concentrations (based on these datasets) are used
for all air quality scenarios (i.e., the potential longer-term impact
of reductions in ambient air Pb on outdoor soil/dust Pb levels and
associated impacts on indoor dust Pb have not be simulated). In areas
where air concentrations have been greater in the past, however,
implementation of a reduced NAAQS might be expected to yield reduced
soil Pb levels over the long term. As described in Section 2.3.3 of the
Risk Assessment Report (USEPA, 2007b), however, there is potentially
significant uncertainty associated with this conclusion, particularly
with regard to implications for areas in which a Pb source may locate
where one of comparable size had not been previously. Additionally, it
is possible that control measures implemented to meet alternative NAAQS
may result in changes to soil Pb concentrations; these are not
reflected in the assessment.
Estimates of indoor dust Pb concentrations for the urban
case studies (application of the hybrid model): The hybrid mechanistic-
empirical model for estimating indoor dust Pb for the urban case
studies (as described in Section 3.1.4.1 of the Risk Assessment Report,
USEPA, 2007b) utilizes a mechanistic model to simulate the exchange of
outdoor ambient air Pb indoors and subsequent deposition (and buildup)
of Pb on indoor surfaces, which relies on a number of empirical
measurements for parameterization (e.g., infiltration rates, deposition
velocities, cleaning frequencies and efficiencies). There is
considerable uncertainty associated with these parameter estimates. In
addition, there is uncertainty associated with the partitioning of
total indoor dust Pb estimates between the infiltration-related
(``recent air'') component and other contributions (``other'' as
described in section II.C.2.e).
Estimates of indoor dust Pb concentrations for the primary
Pb smelter case study (application of the site-specific regression
model): There is uncertainty associated with the site-specific
regression model applied in the remediation zone (as described in
Section 3.1.4.2 of the Risk Assessment Report), and relatively greater
uncertainty associated with its application to air quality scenarios
that simulate notably lower air Pb levels (as is typically the case
when applying regression-based models beyond the bounds of the datasets
used in their derivation). The log-log form of the regression model
prevents the ready identification of an intercept term handicapping us
in partitioning estimates of air-related indoor dust (and consequently
exposure and risk estimates) between ``recent air'' and ``other''
components. In addition, limitations in the model-derived air estimates
used in deriving the regression model prevented effective consideration
for the role of ambient air Pb related to resuspension in influencing
indoor dust Pb levels. A public commenter suggested that indoor dust Pb
levels using this model may be overestimated due to factors associated
with the model's derivation. Factors identified by the commenter,
however, may contribute to a potential for either over- or
underestimation, and as noted by the commenter, additional research
might reduce this uncertainty.
Characterizing interindividual variability using a GSD:
There is uncertainty associated with the GSD specified for each case
study (as described in Sections 3.2.3 and 5.2.2.3 of the Risk
Assessment Report). Two factors are described here as contributors to
that uncertainty. Interindividual variability in blood Pb levels for
any study population (as described by the GSD) will reflect, to a
certain extent, spatial variation in media concentrations, including
outdoor ambient air Pb levels and indoor dust Pb levels, as well as
differences in physiological response to Pb exposure. For each case
study, there is significant uncertainty in the specification of spatial
variability in ambient air Pb levels and associated indoor dust Pb
levels, as noted above. In addition, there are a limited number of
datasets for different types of residential child populations from
which a GSD can be derived (e.g., NHANES datasets \94\ for more
heterogeneous populations and individual study datasets for likely more
homogeneous populations near specific industrial Pb sources). This
uncertainty associated with the GSDs introduces significant uncertainty
in exposure and risk estimates for the 95th population percentile.
---------------------------------------------------------------------------
\94\ The GSD for the urban case studies, in the risk assessment
described in this notice, was derived using NHANES data for the
years 1999-2000.
---------------------------------------------------------------------------
Exposure pathway apportionment for higher percentile blood
Pb level and IQ loss estimates: Apportionment of blood Pb levels for
higher population percentiles is assumed to be the same as that
estimated using the central tendency estimate of blood Pb in an
exposure zone. This introduces significant uncertainty into projections
of pathway apportionment for higher population percentiles of blood Pb
and IQ loss. In reality, pathway apportionment may differ in higher
exposure percentiles. For example, paint and/or drinking water
exposures may increase in importance, with air-related contributions
decreasing as an overall percentage of blood Pb levels and associated
risk. Because of this uncertainty related to pathway apportionment, as
mentioned earlier, greater confidence is placed in estimates of total
Pb exposure and risk in evaluating the impact of the current NAAQS and
alternative NAAQS relative to current conditions.
Relating blood Pb levels to IQ loss: Specification of the
quantitative relationship between blood Pb level and IQ loss is subject
to significant uncertainty at lower blood Pb levels (e.g., below 5
[mu]g/dL concurrent blood Pb). As discussed earlier, there are
limitations in the datasets and concentration-response analyses
available for characterizing the concentration-response relationship at
these lower blood Pb levels. For example, the pooled international
dataset analyzed by Lanphear and others (2005) includes relatively few
children with blood Pb levels below 5 [mu]g/dL and no children with
levels below 1 [mu]g/dL. In recognition of the uncertainty in
specifying a quantitative concentration-response relationship at such
levels, our core modeling approach involves the application of four
different functions to generate a range of risk estimates (as described
in Section 4.2.6 and Section 5.3.1 of the Risk Assessment Report,
USEPA, 2007b). The difference in absolute IQ loss estimates for the
four concentration-response functions for a given case study/air
quality scenario combination is typically close to a factor of 3.
Estimates of differences in IQ loss between air quality scenarios (in
terms of percent), however, are more similar across the four functions,
although the function producing higher overall risk estimates (the dual
linear function, stratified at 7.5 [mu]g/dL, peak blood Pb) also
produces larger absolute reductions in IQ loss compared with the other
three functions.
3. Summary of Estimates and Key Observations
This section presents blood Pb and IQ loss estimates generated in
the exposure and risk assessments. Blood Pb estimates (and air-to-blood
Pb ratios) are presented first, followed by IQ loss estimates.
[[Page 29215]]
a. Blood Pb Estimates
This section presents a summary of blood Pb modeling results for
concurrent blood Pb drawn from the more detailed presentation in the
Staff Paper and the Risk Assessment Report (USEPA, 2007a, 2007b,
2007c).
Blood Pb level estimates for the current conditions air quality
scenarios for these case studies differ somewhat from the national
values associated with recent NHANES information. For example, median
blood Pb levels for the current conditions scenario for the urban case
studies are somewhat larger than the national median from the NHANES
data for 2003-2004. Specifically, values for the three location-
specific urban case studies range from 1.7 to 1.8 [mu]g/dL with the
general urban case study having a value of 1.9 [mu]g/dL (current-
conditions mean) (presented in Risk Assessment Report, Volume I, Table
5-5), while the median value from NHANES (2003-2004) is 1.6 [mu]g/dL
(http://www.epa.gov/envirohealth/children/body_burdens/b1-table.htm).
Additionally, NHANES values for the 90th percentile (for 2003-2004)
were identified and these values can be compared against 90th
percentile estimates generated for the urban case studies (see Risk
Assessment Report, Appendix O, Section O.3.2 for the location-specific
urban case study and Appendix N, Section N.2.1.2 for the general urban
case study). The 90th percentile blood Pb levels for the current
conditions scenario, for the three location-specific urban case studies
range from 4.5 to 4.6 [mu]g/dL, while the estimate for the general
urban case study is 5.0 [mu]g/dL. These 90th percentile values for the
case study populations are larger than the 90th percentile value of 3.9
[mu]g/dL reported by NHANES for all children in 2003-2004. It is noted
that ambient air levels reflected in the urban case studies are likely
to differ from those underlying the NHANES data.\95\
---------------------------------------------------------------------------
\95\ The maximum quarterly mean Pb concentrations in the
location-specific case studies ranged from 0.09-0.36 [mu]g/m\3\,
which are higher levels than the maximum quarterly mean values in
most monitoring sites in the U.S. The median of the maximum
quarterly mean values across all sites in the 2003-05 national
dataset is 0.03 [mu]g/m\3\ (USEPA, 2007a, appendix A).
---------------------------------------------------------------------------
Table 2 presents total blood Pb estimates for alternative
standards, focusing on the median in the assessed population, and
associated estimates for the air-related percentage of total blood Pb
(i.e., bounded on the low end by the ``recent air'' contributions and
on the high end by the ``recent'' plus ``past air'' contribution to
total Pb exposure).
Generally, 95th percentile blood Pb estimates across air quality
scenarios for all case studies (not shown here) are 2-3 times higher
than the median estimates in Table 2. For example, 95th percentile
estimates of total blood Pb for the current NAAQS scenario are 10.6
[mu]g/dL for the general urban case study, 12.3 [mu]g/dL for the
primary Pb smelter subarea, and 7.4 to 10.2 [mu]g/dL for the three
location-specific urban case studies (Staff Paper, Table 4-2). While
the estimates indicate similar fractions of total blood Pb that is air-
related between the 95th percentile and median, there is greater
uncertainty in pathway apportionment among air-related and other
sources for higher percentiles, including the 95th percentile.
Table 2.--Summary of Median Blood Pb Estimates for Concurrent Blood Pb
[Total]
--------------------------------------------------------------------------------------------------------------------------------------------------------
Total blood Pb ([mu]g/dL) (air-related percentage) \A\
--------------------------------------------------------------------------------------------------------------------
NAAQS Level simulated ([mu]g/m\3\ Location-specific urban case studies
max monthly, except as noted below) General urban case Primary Pb smelter --------------------------------------------------------------------
study (subarea) case studyB Cleveland (0.56 [mu]g/ Chicago (0.31 [mu]g/ Los Angeles (0.17
C m\3\) m\3\) [mu]g/m\3\)
--------------------------------------------------------------------------------------------------------------------------------------------------------
1.5 max quarterly \D\.............. 3.1 (61 to 84%)....... 4.6 (up to 87%)....... 2.1 \D\ (57 to 86%).. 3.0 \E\ (63 to 83%).. 2.6E (50 to 81%).
0.50............................... 2.2 (41 to 73%)....... 3.2 (up to 81%)....... 1.8 (39 to 72%)...... (\F\)................ (\F\)
0.20............................... 1.9 (26 to 74%)....... 2.3 (up to 78%)....... 1.7 (6 to 65%)....... 1.8 (17 to 67%)...... 1.7 (\G\) (18 to
71%).
0.05............................... 1.7 (12 to 65%)....... 1.7 (up to 65%)....... 1.6 (1 to 63%)....... 1.6 (6 to 69%)....... 1.6 (13 to 69%).
0.02............................... 1.6 (6 to 69%)........ 1.6 (up to 69%)....... 1.6 (1 to 63%)....... 1.6 (1 to 63%)....... 1.6 (6 to 63%).
--------------------------------------------------------------------------------------------------------------------------------------------------------
\A\ --Blood Pb estimates are rounded to one decimal place. Air-related percentage is bracketed by ``recent air'' (lower bound of presented range) and
``recent'' plus ``past air'' (upper bound of presented range). The term ``past air'' includes contributions from the outdoor soil/dust contribution to
indoor dust, historical air contribution to indoor dust, and outdoor soil/dust pathways; ``recent air'' refers to contributions from inhalation of
ambient air Pb or ingestion of indoor dust Pb predicted to be associated with outdoor ambient air Pb levels, with outdoor ambient air also potentially
including resuspended, previously deposited Pb (see Section II.C.2.e).
\B\ --In the case of the primary Pb smelter subarea, only recent plus past air estimates are available.
\C\ --Median blood Pb levels for the primary smelter (full study area) are estimated at 1.5 [mu]g/dL (for the 1.5 [mu]g/m\3\ max quarterly level) and
1.4 [mu]g/dL for the remaining NAAQS levels simulated. The air-related percentages for these standard levels range from 36% to 79%.
\D\ --This corresponds to roughly 0.7-1.0 [mu]g/m\3\ maximum monthly mean, across the urban case studies.
\E\ --A ``roll-up'' was performed so that the highest monitor in the study area is increased to just meet this level.
\F\ --A ``roll-up'' to this level was not performed.
\G\ --A ``roll-up'' to this level was not performed; these estimates are based on current conditions in this area.
As described in section II.C.2.f, the risk assessment also
developed estimates for air-to-blood ratios, which are described in
section 5.2.5.2 of the Risk Assessment Report (USEPA, 2007b). These
ratios reflect a subset of air-related pathways related to inhalation
and ingestion of indoor dust; inclusion of the remaining pathways would
be expected to yield higher ratios. Additionally, these ratios are
based on blood Pb estimates for the 7th year of exposure (concurrent
blood Pb) which are lower than blood Pb estimates at younger ages (and
than the lifetime-averaged blood Pb metric). Ratios based on other
blood Pb estimates (e.g., lifetime-averaged or peak blood Pb) would be
higher.
For the general urban case study, estimates of air-to-
blood ratios, presented in section 5.2.5.2 of the Risk Assessment
Report (USEPA, 2007b) ranged from 1:2 to 1:9, with the majority of the
estimates ranging from 1:4 to 1:6.\96\ As noted in Section II.C.2.f,
[[Page 29216]]
because the risk assessment only reflects the impact of reductions on
recent air-related pathways in predicting changes in indoor dust Pb for
urban case studies, these ratios are lower than they would be if they
had also reflected potential reductions in other air-related pathways
(e.g., changes in outdoor surface soil/dust Pb levels and diet with
changes in ambient air Pb levels). We also note that the median blood
Pb levels associated with exposure pathways that were not varied in
this assessment (and consequently are not reflected in these ratios)
generally range from 1.3 to 1.5 [mu]g/dL for this case study.
---------------------------------------------------------------------------
\96\ The ratios increase as the level of the alternate standard
decreases. This reflects nonlinearity in the Pb response, which is
greater on a per-unit basis for lower ambient air Pb levels.
---------------------------------------------------------------------------
For the primary Pb smelter subarea, estimates of air-to-
blood ratios, presented in section 5.2.5.2 of the Risk Assessment
Report (USEPA, 2007b) ranged from 1:10 and higher.97 98 One
reason for these estimates being higher than those for the urban case
study is that the dust Pb model used may reflect somewhat ambient air-
related pathways other than that of ambient air infiltrating a home (as
described in Section II.C.2.f above).\99\
---------------------------------------------------------------------------
\97\ As with such estimates for the urban case study, ratios are
higher at lower ambient air Pb levels, reflecting the nonlinearity
of the dust Pb response with air concentration.
\98\ For the primary Pb smelter (full study area), for which
limitations are noted above in section II.C.2.c, the air-to-blood
ratio estimates, presented in section 5.2.5.2 of the Risk Assessment
Report (USEPA, 2007b), ranged from 1:3 to 1:7. As in the other case
studies, ratios are higher at lower ambient air Pb levels. It is
noted that the underlying changes in both ambient air Pb and blood
Pb across standard levels are extremely small, introducing
uncertainty into ratios derived using these data.
\99\ Also, as noted above (Section II.C.2.h), there is increased
uncertainty with application of this regression-based model in air
quality scenarios of notably lower air Pb levels than the data set
used in its derivation.
---------------------------------------------------------------------------
b. IQ Loss Estimates
The risk assessment estimated IQ loss associated with both total Pb
exposure and air-related Pb exposure. This section focuses on findings
in relation to air-related Pb exposure, since this is the category of
risk results considered most relevant to the review in considering
whether the current NAAQS and potential alternative NAAQS provide
protection of public health with an adequate margin of safety
(additional categories of risk results, including IQ loss estimates
based on total Pb exposure and population incidence results, are
presented at the end of the section).\100\
---------------------------------------------------------------------------
\100\ The detailed results are provided in the Risk Assessment
Report (USEPA, 2007b).
---------------------------------------------------------------------------
In considering air-related risk results, we note that IQ loss
associated with air-related exposure for each NAAQS scenario is bounded
by recent-air on the low-end and recent plus past air on the high-end
(as described in section II.C.2.e above). In considering differences in
these risk estimates (or in the total risk estimates presented in the
final Risk Assessment Report) for alternative NAAQS, we note that these
comparisons underestimate the true impacts of the alternate NAAQS and
accordingly, the benefit to public health that would result from lower
NAAQS levels. This is due to our inability to simulate in this
assessment reductions in several outdoor air deposition-related
pathways (e.g., diet, ingestion of outdoor surface soil). The magnitude
of this underestimation is unknown.
As with the discussion of blood Pb results, the IQ loss estimates
are summarized here according to air quality scenario and case study
category (Table 3). In presenting these results, we have focused this
presentation on estimates for the median in each case study population
of children because of the greater confidence associated with estimates
for the median as compared to those for 95th percentile.\101\
Generally, 95th percentile IQ loss estimates for all case studies are
80 to 100% higher than the median results in Table 3. The fraction of
total IQ loss that is air-related for the 95th percentile is generally
similar to that for the median (for a particular combination of case
study and air quality scenario).
---------------------------------------------------------------------------
\101\ A complete presentation of risk estimates is available in
the final Risk Assessment Report, including a presentation of
estimates for the 95th percentile in Table 5-10 of that report.
---------------------------------------------------------------------------
The risk estimates presented in boldface in Table 3 are those
derived using the log-linear with low-exposure linearization
concentration-response function, while the range of estimates
associated with all four concentration-response functions is presented
in parentheses. These functions are discussed above in section
II.C.2.b.
[[Page 29217]]
Table 3.--Summary of Risk Attributable to Air-Related Pb Exposure
----------------------------------------------------------------------------------------------------------------
Median air-related IQ loss \A\
-------------------------------------------------------------------------------
NAAQS level simulated ([mu]g/m Location-specific urban case studies
\3\ max monthly, except as noted Primary Pb -----------------------------------------------
below) General urban smelter Cleveland Los Angeles
case study (subarea) case (0.56 [mu]g/m Chicago (0.31 (0.17 [mu]g/m
study \B, C\ \3\) [mu]g/m \3\) \3\)
----------------------------------------------------------------------------------------------------------------
1.5 max quarterly \D\........... 3.5-4.8 < 6 2.8-3.9 \E\ 3.4-4.7 \E\ 2.7-4.2 \E\
(1.5-7.7) <(3.2-9.4) (0.6-4.6) (1.4-7.4) (1.1-6.2)
0.5............................. 1.9-3.6 < 4.5 0.6-2.9 \F\ \F\
(0.7-4.8) <(2.1-7.7) (0.2-3.9)
0.2............................. 1.2-3.2 < 3.7 0.6-2.8 0.6-2.9 0.7-2.9 \G\
(0.4-4.0) <(1.2-5.1) (0.1-3.2) (0.3-3.6) (0.2-3.5)
0.05............................ 0.5-2.8 < 2.8 0.1-2.6 0.2-2.6 0.3-2.7
(0.2-3.3) <(0.9-3.4) (<0.1-3.1) (0.1-3.2) (0.1-3.2)
0.02............................ 0.3-2.6 < 2.9 <0.1-2.6 0.1-2.6 0.1-2.6
(0.1-3.1) <(0.9-3.3) (<0.1-3.0) (<0.1-3.1) (<0.1-3.1)
----------------------------------------------------------------------------------------------------------------
\A\--Air-related risk is bracketed by ``recent air'' (lower bound of presented range) and ``recent'' plus ``past
air'' (upper bound of presented range). While differences between standard levels are better distinguished by
differences in the ``recent'' plus ``past air'' estimates (upper bounds shown here), these differences are
inherently underestimates. The term ``past air'' includes contributions from the outdoor soil/dust
contribution to indoor dust, historical air contribution to indoor dust, and outdoor soil/dust pathways;
``recent air'' refers to contributions from inhalation of ambient air Pb or ingestion of indoor dust Pb
predicted to be associated with outdoor ambient air Pb levels, with outdoor ambient air also potentially
including resuspended, previously deposited Pb (see Section II.C.2.e). Boldface values are estimates generated
using the log-linear with low-exposure linearization function. Values in parentheses reflect the range of
estimates associated with all four concentration-response functions.
\B\--In the case of the primary Pb smelter case study, only recent plus past air estimates are available.
\C\--Median air-related IQ loss estimates for the primary Pb smelter (full study area) range from <1.7 to <2.9
points, with no consistent pattern across simulated NAAQS levels. This lack of a pattern reflects inclusion of
a large fraction of the study population with relatively low ambient air impacts such that there is lower
variation (at the population median) across standard levels (see Section 4.2 of the Risk Assessment, Volume
1).
\D\--This corresponds to roughly 0.7--1.0 [mu]g/m3 maximum monthly mean, across the urban case studies
\E\--A ``roll-up'' was performed so that the highest monitor in the study area is increased to just meet this
level.
\F\--A ``roll-up'' to this level was not performed.
\G\--A ``roll-up'' to this level was not performed; these estimates are based on current conditions in this
area.
Key observations regarding the median estimates of air-related risk
for the current NAAQS and alternative standards presented in Table 3
include:
For the scenario for the current NAAQS (1.5 [mu]g/m\3\,
maximum quarterly average), air-related risk exceeds 2 points IQ loss
at the median and the upper bound of air-related risk is near or above
4 points IQ loss in all five case studies.\102\
---------------------------------------------------------------------------
\102\ As noted in Table 3 and section II.C.2.d above, and
discussed further, with regard to associated limitations and
uncertainties, in section II.C.2.h above, a proportional roll-up
procedure was used to estimate air Pb concentrations in this
scenario for the location-specific case studies.
---------------------------------------------------------------------------
Alternate standards provide substantial reduction in
estimates of air-related risk across the full set of alternative NAAQS
considered in this analysis (i.e., 0.5 to 0.02 [mu]g/m\3\ max monthly).
This is particularly the case for the lower bounds of the air-related
estimates presented in Table 3, which reflect the estimates for
``recent air''-related pathways, which are the pathways that were
varied with changes in air concentrations (as described above in
section II.C.2.e). There is less risk reduction associated with the
upper bounds of these estimates as the upper bound values are inclusive
of the exposure pathways categorized as ``past air'' which were not
varied with changes in air concentrations (as described in section
II.C.2.3). The upper bound estimates for the lowest level assessed
(0.02 [mu]g/m\3\) are 2.6-2.9 points IQ loss.
In the general urban case study, the lower bound of air-
related risk falls below 2 points IQ loss for an alternative NAAQS of
0.5 [mu]g/m\3\ max monthly, and below 1 point IQ loss somewhere between
an alternative NAAQS of 0.2 and 0.05 [mu]g/m\3\ max monthly.
The upper-bound of air-related risk for the primary Pb
smelter subarea is generally higher than that for the general urban
case study, likely due to the difference in indoor dust models used for
the two case studies. The indoor dust Pb model used for the primary Pb
smelter considered more completely, the impact of outdoor ambient air
Pb on indoor dust (compared to the hybrid indoor dust Pb model used in
the urban case studies). Specifically, the regression model used for
the primary Pb smelter included consideration for longer-term
relationships between outdoor ambient air and indoor dust (e.g.,
changes in outdoor soil and subsequent tracking in of soil Pb).
As noted above (section II.C.2.c), the three location-
specific urban case studies provide risk estimates for populations with
a broader range of air-related exposures. Accordingly, because of the
population distribution in these three case studies, the air-related
risk is smaller for them than for the other case studies, particularly
at the population median. Further, the majority of the population in
each case study resides in areas with ambient air Pb levels well below
each standard level assessed, particularly for levels above 0.05 [mu]g/
m\3\ max monthly. Consequently, risk estimates indicate little response
to alternative standard levels above 0.05 [mu]g/m\3\ max monthly.
In addition to the air-related risk results described above, we
present two additional categories of risk results, including (a)
estimates of median IQ loss based on total Pb exposure for each case
study (Table 4) and (b) IQ loss incidence estimates for each of the
location-specific case studies (Tables 4 and 5).\103\ Each of these
categories of risk results are described in creater detail below:
---------------------------------------------------------------------------
\103\ As recognized in section II.C.2.d above, to simulate air
concentrations associated with the current NAAQS, a proportional
roll-up of concentrations from those for current conditions was
performed for the location-specific urban case studies. This was not
necessary for the primary Pb smelter case study in which air
concentrations currently exceed the current standard.
---------------------------------------------------------------------------
Estimates of IQ loss for all air quality scenarios (based
on total Pb exposure): Table 4 presents median IQ loss estimates for
total Pb exposure for each of the air quality scenarios simulated for
each case study (as noted earlier in this section, there is greater
uncertainty associated with higher-end risk percentiles and therefore,
they are
[[Page 29218]]
not presented in tabular format here--see Table 5-10 of Risk Assessment
Volume 1 for 95th percentile total IQ loss estimates). As with the
incremental risk results presented in Table 3 above, in order to
reflect the variation in estimates derived from the four different
concentration-response functions included in the analysis, three
categories of estimates are presented in Table 4 including (a) IQ loss
estimates generated using the low concentration-response function (the
model that generated the lowest IQ loss estimates), (b) estimates
generated using the log-linear with low-exposure linearization (LLL)
model, and (c) IQ loss estimates generated using the high
concentration-response function (the model that generated the highest
IQ loss estimates). It is important to emphasize, that, as noted in
Section II.C.2.e, because of limitations in modeling methods, we were
only able to simulate reduction in recent air-related exposures in
considering alternate standard levels and could not simulate reduction
in past air-related exposures. This likely results in an underestimate
of the total degree of reduction in exposure and risk associated with
each standard level. Therefore, in comparing total risk estimates
between alternate NAAQS scenarios (i.e., considering incremental risk
reductions), this aspect of the analysis will tend to underestimate the
reductions in risk associated with alternative NAAQS.
IQ loss incidence estimates for the three location-
specific urban case studies: Estimates of the number of children for
each location-specific urban case study projected to have total Pb-
related IQ loss greater than one point are summarized in Table 5, and
similar estimates for IQ loss greater than 7 points are summarized in
Table 6. Also presented are the changes in incidence of the current
NAAQS and alternative NAAQS scenarios compared to current conditions,
with emphasis placed on estimates generated using the LLL
concentration-response function. Estimates are presented for each of
the four concentration-response functions used in the risk analysis.
This metric illustrates the overall number of children within a given
urban case study location projected to experience various levels of IQ
loss due to Pb exposure and how that distribution of incidence changes
with alternate standard levels. These incidence estimates were only
generated for the location-specific urban case studies, since these
have larger enumerated study populations (additional detail on the
derivation of these incidence estimates is presented in Section 5.3.1.2
of the Risk Assessment Report). The complete set of incidence results
is presented in Risk Assessment Report Appendix O, Section O.3.4.
Total IQ loss results presented in Table 4 for the primary Pb
smelter case study (full study area) illustrate the reason why these
results were not presented earlier in summarizing air-related IQ loss
estimates for the primary Pb smelter case study in Table 3 (and
instead, results for the subarea were presented). As mentioned earlier
in Section II.C.2.c, the full study area for the primary Pb smelter
case study incorporates a large number of simulated children with
relatively low air-related impacts, which results in little
differentiation between alternate standard levels in terms of total IQ
loss (as well as air-related IQ loss). This can be seen by considering
the results in Table 4 for the primary Pb smelter (full study area).
Those results suggest that total IQ loss varies little across alternate
standard levels for the full study area simulation, with the only
noticeable difference in total IQ loss resulting from analysis of the
current standard (when compared to alternate levels). By contrast,
there are notable differences in total IQ loss between alternative
standard levels for the sub-area of the primary Pb smelter case study.
Table 4.--Summary of Risk Estimates for Medians of Total-Exposure Risk Distributions
----------------------------------------------------------------------------------------------------------------
Points IQ loss (total Pb exposure) \a\
-----------------------------------------------
Case study and air quality scenario Low C-R High C-R
function LLL \b\ function
estimate estimate
----------------------------------------------------------------------------------------------------------------
Location-specific (Chicago)
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 2.4 5.6 8.8
Current conditions (0.14 [mu]g/m\3\ max quarterly; 0.31 [mu]g/ 1.4 4.2 5.2
m\3\ max monthly)..............................................
Alternative NAAQS (0.2 [mu]g/m\3\, max monthly)................. 1.4 4.2 5.2
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.8
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.7
----------------------------------------------------------------------------------------------------------------
Location-specific (Cleveland)
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 1.7 4.7 6.3
Current conditions (0.36 [mu]g/m\3\ max quarterly; 0.56 [mu]g/ 1.4 4.2 5.2
m\3\ max monthly)..............................................
Alternative NAAQS (0.5 [mu]g/m\3\, max monthly)................. 1.4 4.2 5.2
Alternative NAAQS (0.2 [mu]g/m\3\, max quarterly)............... 1.4 4.1 5.0
Alternative NAAQS (0.2 [mu]g/m\3\, max monthly)................. 1.3 4.1 4.9
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.7
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 1.2 3.9 4.6
----------------------------------------------------------------------------------------------------------------
Location-specific (Los Angeles)
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 2.1 5.3 7.7
Current conditions (0.09 [mu]g/m\3\ max quarterly; 0.17 [mu]g/ 1.4 4.2 5.1
m\3\ max monthly)..............................................
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.8
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.7
----------------------------------------------------------------------------------------------------------------
General Urban
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 2.5 5.8 9.2
[[Page 29219]]
Alternative NAAQS (0.5 [mu]g/m\3\, max monthly)................. 1.7 4.8 6.4
Current conditions--high-end (0.87 [mu]g/m\3\ max quarterly).... 1.7 4.7 6.3
Alternative NAAQS (0.2 [mu]g/m\3\, max quarterly)............... 1.6 4.6 5.9
Current conditions--mean (0.14 [mu]g/m\3\ max quarterly)........ 1.5 4.5 5.6
Alternative NAAQS (0.2 [mu]g/m\3\, max monthly)................. 1.5 4.4 5.6
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 1.3 4.1 5.0
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.8
----------------------------------------------------------------------------------------------------------------
Primary Pb smelter--full study area
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 1.2 3.8 4.4
Alternative NAAQS (0.5 [mu]g/m\3\, max monthly)................. 1.0 3.7 4.2
Alternative NAAQS (0.2 [mu]g/m\3\, max quarterly)............... 0.9 3.6 4.2
Alternative NAAQS (0.2 [mu]g/m\3\, max monthly)................. 0.9 3.6 4.1
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 0.9 3.6 4.0
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 0.9 3.6 4.1
----------------------------------------------------------------------------------------------------------------
Primary Pb smelter--1.5km subarea
----------------------------------------------------------------------------------------------------------------
Current NAAQS (1.5 [mu]g/m\3\, max quarterly)................... 3.7 6.8 11.2
Alternative NAAQS (0.5 [mu]g/m\3\, max monthly)................. 2.6 5.8 9.4
Alternative NAAQS (0.2 [mu]g/m\3\, max quarterly)............... 2.0 5.2 7.4
Alternative NAAQS (0.2 [mu]g/m\3\, max monthly)................. 1.9 5.0 6.9
Alternative NAAQS (0.05 [mu]g/m\3\, max monthly)................ 1.4 4.2 5.1
Alternative NAAQS (0.02 [mu]g/m\3\, max monthly)................ 1.3 4.0 4.8
----------------------------------------------------------------------------------------------------------------
\a\ --These columns present the estimates of total IQ loss resulting from total Pb exposure (policy-relevant
plus background). Estimates below 1.0 are rounded to one decimal place, all values below 0.05 are presented as
<0.1 and values between 0.05 and 0.1 as 0.1. All values above 1.0 are rounded to the nearest whole number.
\b\ --Log-linear with low-exposure linearization concentration-response function.
Table 5.--Incidence of Children With >1 Point Pb-Related IQ Loss
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dual linear--stratified Log-linear with Dual linear--stratified Log-linear with cutpoint
at 7.5 mg/dl peak blood linearization at 10 m/dL peak blood -------------------------
Pb -------------------------- Pb
-------------------------- -------------------------- Delta
Air quality scenario (for location-specific Delta Delta Delta (change in
urban case studies) (change Incidence (change in (change in Incidence incidence
Incidence inincidence of >1 point incidence Incidence incidence of >1 point compared to
of >1 point compared to IQ loss compared to of >1 point compared to IQ loss current
IQ loss current current IQ loss current conditions)
conditions) conditions) conditions)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Chicago (total modeled child population:
396,511):
Chicago Current Conditions.................. 391,602 ........... 389,754 ........... 271,031 ........... 236,257
Current NAAQS (1.5 mg/m\3\ Maximum 395,797 4,195 395,528 5,773 347,415 76,384 314,053 77,795
Quarterly).................................
Alternative NAAQS (0.2 mg/m\3\ Maximum 391,158 -444 389,461 -293 271,444 412 235,559 -698
Monthly)...................................
Alternative NAAQS (0.05 mg/m\3\ Maximum 389,572 -2,030 387,407 -2,347 253,775 -17,256 224,394 -11,864
Monthly)...................................
Alternative NAAQS (0.02 mg/m\3\ Maximum 389,176 -2,427 386,630 -3,125 249,865 -21,166 219,294 -16,963
Monthly)...................................
Cleveland (total modeled child population:
13,990):
Cleveland Current Conditions................ 13,809 ........... 13,745 ........... 9,526 ........... 8,515
Current NAAQS (1.5 mg/m\3\ Maximum 13,893 84 13,857 112 10,664 1,137 9,769 1,254
Quarterly).................................
Alternative NAAQS (0.2 mg/m\3\ Maximum 13,770 -38 13,703 -42 9,221 -305 8,160 -354
Quarterly).................................
Alternative NAAQS (0.5 mg/m\3\ Maximum 13,789 -20 13,720 -25 9,497 -29 8,464 -51
Monthly)...................................
Alternative NAAQS (0.2 mg/m\3\ Maximum 13,759 -50 13,694 -51 9,083 -443 8,010 -505
Monthly)...................................
Alternative NAAQS (0.05 mg/m\3\ Maximum 13,729 -80 13,642 -103 8,785 -741 7,720 -795
Monthly)...................................
Alternative NAAQS (0.02 mg/m\3\ Maximum 13,720 -88 13,628 -117 8,736 -790 7,668 -846
Monthly)...................................
Los Angeles (total modeled child population:
372,252):
Los Angeles Current Conditions.............. 282,216 ........... 280,711 ........... 191,675 ........... 170,474 ...........
Current NAAQS (1.5 mg/m\3\ Maximum, 285,272 3,056 284,945 4,234 240,988 49,313 226,608 56,134
Quarterly).................................
[[Page 29220]]
Alternative NAAQS (0.05 mg/m\3\ Maximum 281,112 -1,104 279,658 -1,053 183,395 -8,280 161,914 -8,560
Monthly)...................................
Alternative NAAQS (0.02 mg/m\3\ Maximum 280,740 -1,476 279,057 -1,654 180,745 -10,929 158,234 -12,240
Monthly)...................................
--------------------------------------------------------------------------------------------------------------------------------------------------------
Table 6.--Incidence of Children With >7 Points Pb-Related IQ Loss
--------------------------------------------------------------------------------------------------------------------------------------------------------
Dual linear--stratified Log-linear with Dual linear--stratified Log-linear with cutpoint
at 7.5 ug/dL peak blood linearization at 10 ug/dL peak blood -------------------------
Pb -------------------------- Pb
-------------------------- -------------------------- Delta
Air quality scenario (location-specific urban Delta Delta Delta Incidence (change in
case studies) Incidence (change in Incidence (change in Incidence (change in of > 7 incidence
of > 7 incidence of > 7 incidence of > 7 incidence points IQ compared to
points IQ compared to points IQ compared to points IQ compared to loss current
loss current loss current loss current conditions)
conditions) conditions) conditions)
--------------------------------------------------------------------------------------------------------------------------------------------------------
Chicago (total modeled child population:
396,511):
Chicago Current Conditions.................. 136,709 ........... 33,664 ........... 63 ........... 1,015 ...........
Current NAAQS (1.5 [mu]g/m\3\ Maximum 244,401 107,692 100,159 66,495 555 492 5,226 4,211
Quarterly).................................
Alternative NAAQS (0.2 [mu]g/\3\ Maximum 136,067 -642 32,546 -1,118 48 -16 1,007 -8
Monthly)...................................
Alternative NAAQS (0.05 [mu]g/\3\ Maximum 120,706 -16,003 27,367 -6,297 16 -48 864 -151
Monthly)...................................
Alternative NAAQS (0.02 [mu]g/\3\ Maximum 117,819 -18,890 26,027 -7,637 8 -56 690 -325
Monthly)...................................
Cleveland (total modeled child population:
13,990):
Cleveland Current Conditions................ 4,834 ........... 1,212 ........... 3 ........... 46 ...........
Current NAAQS (1.5 [mu]g/m\3\ Maximum 6,139 1,305 1,858 647 4 2 105 59
Quarterly).................................
Alternative NAAQS (0.2 [mu]g/m\3\ Maximum 4,525 -309 1,073 -139 1 -2 40 -6
Quarterly).................................
Alternative NAAQS (0.5 [mu]g/m\3\ Maximum 4,806 -28 1,180 -31 1 -2 43 -3
Monthly)...................................
Alternative NAAQS (0.2 [mu]g/m\3\ Maximum 4,424 -410 1,026 -186 1 -2 43 -3
Monthly)...................................
Alternative NAAQS (0.05 [mu]g/m\3\ Maximum 4,106 -728 886 -326 0 -3 24 -22
Monthly)...................................
Alternative NAAQS (0.02 [mu]g/m\3\ Maximum 4,051 -783 866 -345 0 -3 27 -18
Monthly)...................................
Los Angeles (total modeled child population:
372,252):
Los Angeles Current Conditions.............. 94,684 ........... 22,665 ........... 23 ........... 732 ...........
Current NAAQS (1.5 [mu]g/m\3\ Maximum, 158,171 63,487 57,834 35,168 183 160 3,771 3,038
Quarterly).................................
Alternative NAAQS (0.05 [mu]g/m\3\ Maximum, 87,303 -7,382 19,781 -2,884 11 -11 624 -109
Monthly)...................................
Alternative NAAQS (0.02 [mu]g/m\3\ Maximum, 83,909 -10,775 17,939 -4,726 17 -6 498 -235
Monthly)...................................
--------------------------------------------------------------------------------------------------------------------------------------------------------
D. Conclusions on Adequacy of the Current Primary Standard
The initial issue to be addressed in the current review of the
primary Pb standard is whether, in view of the advances in scientific
knowledge and additional information, the existing standard should be
retained or revised. In evaluating whether it is appropriate to retain
or revise the current standard, the Administrator builds on the general
approach used in the initial setting of the standard, as well as that
used in the last review, and reflects the broader body of evidence and
information now available.
The approach used is based on an integration of information on
health effects associated with exposure to ambient Pb; expert judgment
on the adversity of such effects on individuals; and policy judgments
as to when the standard is requisite to protect public health with an
adequate margin of safety, which are informed by air quality and
related analyses, quantitative exposure and risk assessments when
possible, and qualitative assessment of impacts that could not be
quantified.
The Administrator has taken into account both evidence-based \104\
and quantitative exposure- and risk-based considerations in developing
conclusions on the adequacy of the current primary Pb standard.
Evidence-based considerations include the assessment of evidence for a
variety of
[[Page 29221]]
Pb-related health endpoints from epidemiological, and animal
toxicological studies. Consideration of quantitative exposure- and
risk-based information draws from the results of the exposure and risk
assessments described above. More specifically, estimates of the
magnitude of Pb-related exposures and risks associated with air quality
levels associated with just meeting the current primary Pb NAAQS have
been considered.\105\
---------------------------------------------------------------------------
\104\ The term ``evidence-based'' as used here refers to the
drawing of information directly from published studies, with
specific attention to those reviewed and described in the Criteria
Document, and is distinct from considerations that draw from the
results of the quantitative exposure and risk assessement.
\105\ As described in seciton II.C.2.d above, levels in the
location-specific urban case studies were increased from current
conditions such that the portion of each case study with highest
concentrations would just meet the current NAAQS.
---------------------------------------------------------------------------
In this review, a series of general questions frames the approach
to reaching a decision on the adequacy of the current standard, such as
the following: (1) To what extent does newly available information
reinforce or call into question evidence of associations of Pb
exposures with effects identified when the standard was set?; (2) to
what extent has evidence of new effects or at-risk populations become
available since the time the standard was set?; (3) to what extent have
important uncertainties identified when the standard was set been
reduced and have new uncertainties emerged?; and (4) to what extent
does newly available information reinforce or call into question any of
the basic elements of the current standard?
The question of whether the available evidence supports
consideration of a standard that is more protective than the current
standard includes consideration of: (1) Whether there is evidence that
associations with blood Pb in epidemiological studies extend to ambient
Pb concentration levels that are as low as or lower than had previously
been observed, and the important uncertainties associated with that
evidence; (2) the extent to which exposures of potential concern and
health risks are estimated to occur in areas upon meeting the current
standard and the important uncertainties associated with the estimated
exposures and risks; and (3) the extent to which the Pb-related health
effects indicated by the evidence and the exposure and risk assessments
are considered important from a public health perspective, taking into
account the nature and severity of the health effects, the size of the
at-risk populations, and the kind and degree of the uncertainties
associated with these considerations.
This approach is consistent with the requirements of the NAAQS
provisions of the Act and with how EPA and the courts have historically
interpreted the Act. These provisions require the Administrator to
establish primary standards that, in the Administrator's judgment, are
requisite to protect public health with an adequate margin of safety.
In so doing, the Administrator seeks to establish standards that are
neither more nor less stringent than necessary for this purpose. The
Act does not require that primary standards be set at a zero-risk level
but rather at a level that avoids unacceptable risks to public health,
including the health of sensitive groups.
The following discussion starts with background information on the
current standard (section II.D.1), including both the basis for
derivation of the current standard and considerations and conclusions
from the 1990 Staff Paper (USEPA, 1990b). This is followed by a
discussion of the Agency's approach in this review for evaluating the
adequacy of the current standard, in section II.D.2, including both
evidence-based and exposure/risk-based considerations (sections
II.D.2.a and b, respectively). CASAC advice and recommendations
concerning adequacy of the current standard are summarized in section
II.D.3. Lastly, the Administrator's proposed conclusions with regard to
the adequacy of the current standard are presented in section II.D.4.
1. Background
a. The Current Standard
The current primary standard is set at a level of 1.5 [mu]g/m\3\,
measured as Pb-TSP, not to be exceeded by the maximum arithmetic mean
concentration averaged over a calendar quarter. The standard was set in
1978 to provide protection to the public, especially children as the
particularly sensitive population subgroup, against Pb-induced adverse
health effects (43 FR 46246). In setting the standard, EPA relied on
conclusions regarding sources of exposure, air-related exposure
pathways, variability and susceptibility of young children, the most
sensitive health endpoints, blood Pb level thresholds for various
health effects and the stability and distributional characteristics of
Pb (both in the human body and in the environment) (43 FR 46247). The
specific basis for selecting each of the elements of the standard is
described below.
i. Level
EPA's objective in selecting the level of the current standard was
``to estimate the concentration of Pb in the air to which all groups
within the general population can be exposed for protracted periods
without an unacceptable risk to health'' (43 FR 46252). As stated in
the notice of final rulemaking, ``This estimate was based on EPA's
judgment in four key areas:
(1) Determining the `sensitive population' as that group within the
general population which has the lowest threshold for adverse effects
or greatest potential for exposure. EPA concludes that young children,
aged 1 to 5, are the sensitive population.
(2) Determining the safe level of total lead exposure for the
sensitive population, indicated by the concentration of lead in the
blood. EPA concludes that the maximum safe level of blood lead for an
individual child is 30 [mu]g Pb/dl and that population blood lead,
measured as the geometric mean, must be 15 [mu]g Pb/dl in order to
place 99.5 percent of children in the United States below 30 [mu]g Pb/
dl.
(3) Attributing the contribution to blood lead from nonair
pollution sources. EPA concludes that 12 [mu]g Pb/dl of population
blood lead for children should be attributed to nonair exposure.
(4) Determining the air lead level which is consistent with
maintaining the mean population blood lead level at 15 [mu]g Pb/dl [the
maximum safe mean level]. Taking into account exposure from other
sources (12 [mu]g Pb/dl), EPA has designed the standard to limit air
contribution after achieving the standard to 3 [mu]g Pb/dl. On the
basis of an estimated relationship of air lead to blood lead of 1 to 2,
EPA concludes that the ambient air standard should be 1.5 [mu]g Pb/
m\3\.'' (43 FR 46252)
EPA's judgments in these key areas, as well as margin of safety
considerations, are discussed below.
The assessment of the science that was presented in the 1977
Criteria Document (USEPA, 1977), indicated young children, aged 1 to 5,
as the population group at particular risk from Pb exposure. Children
were recognized to have a greater physiological sensitivity than adults
to the effects of Pb and a greater exposure. In identifying young
children as the sensitive population, EPA also recognized the
occurrence of subgroups with enhanced risk due to genetic factors,
dietary deficiencies or residence in urban areas. Yet information was
not available to estimate a threshold for adverse effects for these
subgroups separate from that of all young children. Additionally, EPA
recognized both a concern regarding potential risk to pregnant women
and fetuses, and a lack of information to establish that these
subgroups are more at risk than young children. Accordingly, young
children, aged 1 to 5, were identified as the group which has the
lowest threshold for adverse
[[Page 29222]]
effects of greatest potential for exposure (i.e., the sensitive
population) (43 FR 46252).
In identifying the maximum safe exposure, EPA relied upon the
measurement of Pb in blood (43 FR 46252-46253). The physiological
effect of Pb that had been identified as occurring at the lowest blood
Pb level was inhibition of an enzyme integral to the pathway by which
heme (the oxygen carrying protein of human blood) is synthesized, i.e.,
delta-aminolevulinic acid dehydratase ([delta]-ALAD). The 1977 Criteria
Document reported a threshold for inhibition of this enzyme in children
at 10 [mu]g Pb/dL. The 1977 Criteria Document also reported a threshold
of 15-20 [mu]g/dL for elevation of erythrocyte protoporphyrin (EP),
which is an indication of some disruption of the heme synthesis
pathway. EPA concluded that this effect on the heme synthesis pathway
(indicated by EP) was potentially adverse. EPA further described a
range of blood levels associated with a progression in detrimental
impact on the heme synthesis pathway. At the low end of the range (15-
20 [mu]g/dL), the initial detection of EP associated with blood Pb was
not concluded to be associated with a significant risk to health. The
upper end of the range (40 [mu]g/dL), the threshold associated with
clear evidence of heme synthesis impairment and other effects
contributing to clinical symptoms of anemia, was regarded by EPA as
clearly adverse to health. EPA also noted that for some children with
blood Pb levels just above those for these effects (e.g., 50 [mu]g/dL),
there was risk for additional adverse effects (e.g., nervous system
deficits). Additionally, in the Agency's statement of factors on which
the conclusion as to the maximum safe blood Pb level for an individual
child was based, EPA stated that the maximum safe blood level should be
``no higher than the blood Pb range characterized as undue exposure by
the Center for Disease Control of the Public Health Service, as
endorsed by the American Academy of Pediatrics, because of elevation of
erythrocyte protoporphyrin (above 30 [mu]g Pb/dL)''.\106\
---------------------------------------------------------------------------
\106\ The CDC subsequently revised their advisory level for
children's blood Pb to 25 [mu]g/dL in 1985, and to 10 [mu]g/dL in
1991. In 2005, with consideration of a review of the evidence by
their advisory committee, CDC revised their statement on Preventing
Lead Poisoning in Young Children, specifically recognizing the
evidence of adverse health effects in children with blood Pb levels
below 10 [mu]g/dL and the data demonstrating that no ``safe''
threshold for blood Pb in children had been identified, and
emphasizing the importance of preventative measures (CDC, 2005a).
Recently, CDC's Advisory Committee on Childhood Lead Poisoning
Prevention noted the 2005 CDC statements and reported on a review of
the clinical interpretation and management of blood Pb levels below
10 [mu]g/dL (ACCLPP, 2007). More details on this level are provided
in Section II.B.1.
---------------------------------------------------------------------------
Having identified the maximum safe blood level in individual
children, EPA next made a public health policy judgment regarding the
target mean blood level for the U.S. population of young children (43
FR 46252-46253). With this judgment, EPA identified a target of 99.5
percent of this population to be brought below the maximum safe blood
Pb level. This judgment was based on consideration of the size of the
sensitive subpopulation, and the recognition that there are special
high-risk groups of children within the general population. The
population statistics available at the time (the 1970 U.S. Census)
indicated a total of 20 million children younger than 5 years of age,
with 15 million residing in urban areas and 5 million in center cities
where Pb exposure was thought likely to be ``high''. Concern about
these high-risk groups influenced EPA's determination of 99.5 percent,
deterring EPA from selecting a population percentage lower than 99.5
(43 FR 46253). EPA then used standard statistical techniques to
calculate the population mean blood Pb level that would place 99.5
percent of the population below the maximum safe level. Based on the
then available data, EPA concluded that blood Pb levels in the
population of U.S. children were normally distributed with a GSD of
1.3. Based on standard statistical techniques, EPA determined that a
thus described population in which 99.5 percent of the population has
blood Pb levels below 30 [mu]g/dL would have a geometric mean blood
level of 15 [mu]g/dL. EPA described 15 [mu]g/dL as ``the maximum safe
blood lead level (geometric mean) for a population of young children''
(43 FR 46247).
When setting the current NAAQS, EPA recognized that the air
standard needed to take into account the contribution to blood Pb
levels from Pb sources unrelated to air pollution. Consequently, the
calculation of the current NAAQS included the subtraction of Pb
contributed to blood Pb from nonair sources, from the estimate of a
safe mean population blood Pb level. Without this subtraction, EPA
recognized that the combined exposure to Pb from air and nonair sources
would result in a blood Pb concentration exceeding the safe level (43
FR 46253). In developing an estimate of this nonair contribution, EPA
recognized the lack of detailed or widespread information about the
relative contribution of various sources to children's blood Pb levels,
such that an estimate could only be made by inference from other
empirical or theoretical studies, often involving adults. Additionally,
EPA recognized the expectation that the contribution to blood Pb levels
from nonair sources would vary widely, was probably not in constant
proportion to air Pb contribution, and in some cases may alone exceed
the target mean population blood Pb level (43 FR 46253-46254). The
amount of blood Pb attributed to nonair sources was selected based
primarily on findings in studies of blood Pb levels in areas where air
Pb levels were low relative to other locations in U.S. The air Pb
levels in these areas ranged from 0.1 to 0.7 [mu]g/m\3\. The average of
the reported blood Pb levels for children of various ages in these
areas was on the order of 12 [mu]g/dL. Thus, 12 [mu]g/dL was identified
as the nonair contribution, and subtracted from the population mean
target level of 15 [mu]g/dL to yield a value of 3 [mu]g/dL as the limit
on the air contribution to blood Pb.
In determining the air Pb level consistent with an air contribution
of 3 [mu]g Pb/dL, EPA reviewed studies assessed in the 1977 Criteria
Document that reported changes in blood Pb with different air Pb
levels. These studies included a study of children exposed to Pb from a
primary Pb smelter, controlled exposures of adult men to Pb in fine
particulate matter, and a personal exposure study involving several
male cohorts exposed to Pb in a large urban area in the early 1970s (43
FR 46254).\107\ Using all three studies, EPA calculated an average
slope or ratio over the entire range of data. That value was 1.95
(rounded to 2 [mu]g/dL blood Pb concentration to 1 [mu]g/m\3\ air Pb
concentration), and is recognized to fall within the range of values
reported in the 1977 Criteria Document. On the basis of this 2-to-1
relationship, EPA concluded that the ambient air standard should be 1.5
[mu]g Pb/m\3\ (43 FR 46254).
---------------------------------------------------------------------------
\107\ Mean blood Pb levels in the adult study groups ranged from
10 [mu]g/dL to approximately 30 [mu]g/dL and in the child groups
they ranged from approximately 20 [mu]g/dL up to 65 [mu]g/dL (USEPA,
1986a, section 11.4.1).
---------------------------------------------------------------------------
In consideration of the appropriate margin of safety during the
development of the current NAAQS, EPA identified the following factors:
(1) The 1977 Criteria Document reported multiple biological effects of
Pb in practically all cell types, tissues and organ systems, of which
the significance for health had not yet been fully studied; (2) no
beneficial effects of Pb at then current environmental levels were
recognized;
[[Page 29223]]
(3) data were incomplete as to the extent to which children are
indirectly exposed to air Pb that has moved to other environmental
media, such as water, soil and dirt, and food; (4) Pb is chemically
persistent and with continued uncontrolled emissions would continue to
accumulate in human tissue and the environment; and (5) the possibility
that exposure associated with blood Pb levels previously considered
safe might influence neurological development and learning abilities of
the young child (43 FR 46255). Recognizing that estimating an
appropriate margin of safety for the air Pb standard was complicated by
the multiple sources and media involved in Pb exposure, EPA chose to
use margin of safety considerations principally in establishing a
maximum safe blood Pb level for individual children (30 [mu]g Pb/dL)
and in determining the percentage of children to be placed below this
maximum level (about 99.5 percent). Additionally, in establishing other
factors used in calculating the standard, EPA used margin of safety
considerations in the sense of making careful judgment based on
available data, but these judgments were not considered to be at the
precautionary extreme of the range of data available at the time (43 FR
46251).
EPA further recognized that, because of the variability between
individuals in a population experiencing a given level of Pb exposure,
it was considered impossible to provide the same margin of safety for
all members in the sensitive population or to define the margin of
safety in the standard as a simple percentage. EPA believed that the
factors it used in designing the standards provided an adequate margin
of safety for a large proportion of the sensitive population. The
Agency did not believe that the margin was excessively large or on the
other hand that the air standard could protect everyone from elevated
blood Pb levels (43 FR 46251).
ii. Averaging Time, Form, and Indicator
The averaging time for the current standard is a calendar quarter.
In the decision for this aspect of the standard, the Agency also
considered a monthly averaging period, but concluded that ``a
requirement for the averaging of air quality data over calendar quarter
will improve the validity of air quality data gathered without a
significant reduction in the protectiveness of the standards.'' As
described in the notice for this decision (43 FR 46250), this
conclusion was based on several points, including the following:
An analysis of ambient measurements available at the time
indicated that the distribution of air Pb levels was such that there
was little possibility that there could be sustained periods greatly
above the average value in situations where the quarterly standard was
achieved.
A recognition that the monitoring network may not actually
represent the exposure situation for young children, such that it
seemed likely that elevated air Pb levels when occurring would be close
to Pb air pollution sources where young children would typically not
encounter them for the full 24-hour period reported by the monitor.
Medical evidence available at the time indicated that
blood Pb levels re-equilibrate slowly to changes in air exposure, a
finding that would serve to dampen the impact of short-term period of
exposure to elevated air Pb.
Direct exposure to air is only one of several routes of
total exposure, thus lessening the impact of a change in air Pb on
blood Pb levels.
The statistical form of the current standard is a not-to-be-
exceeded or maximum value. EPA set the standard as a ceiling value with
the conclusion that this air level would be safe for indefinite
exposure for young children (43 FR 46250).
The indicator is total airborne Pb collected by a high volume
sampler (43 FR 46258). EPA's selection of Pb-TSP as the indicator for
the standard was based on explicit recognition both of the significance
of ingestion as an exposure pathway for Pb that had deposited from the
air and of the potential for Pb deposited from the air to become re-
suspended in respirable size particles in the air and available for
human inhalation exposure. As stated in the final rule, ``a significant
component of exposure can be ingestion of materials contaminated by
deposition of lead from the air,'' and that, ``in addition to the
indirect route of ingestion and absorption from the gastrointestinal
tract, non-respirable Pb in the environment may, at some point become
respirable through weathering or mechanical action'' (43 FR 46251).
b. Policy Options Considered in the Last Review
During the 1980s, EPA initiated a review of the air quality
criteria and NAAQS for Pb. CASAC and the public were fully involved in
this review, which led to the publication of a criteria document with
associated addendum and a supplement (USEPA, 1986a, 1986b, 1990a), an
exposure analysis methods document (USEPA, 1989), and a staff paper
(USEPA, 1990b).
Total emissions to air were estimated to have dropped by 94 percent
between 1978 and 1987, with the vast majority of it attributed to the
reduction of Pb in gasoline. Accordingly, the focus of the last review
was on areas near stationary sources of Pb emissions. Although such
sources were not considered to have made a significant contribution (as
compared to Pb in gasoline) to the overall Pb pollution across large-
urban or regional areas, Pb emissions from such sources were considered
to have the potential for a significant impact on a local scale. Air Pb
concentrations, and especially soil and dust Pb concentrations, had
been associated with elevated levels of Pb absorption in children and
adults in numerous Pb point source community studies. Exceedances of
the current NAAQS were found at that time only in the vicinity of
nonferrous smelters or other point sources of Pb.
In summarizing and interpreting the health evidence presented in
the 1986 Criteria Document and associated documents, the 1990 Staff
Paper described the collective impact on children of the effects at
blood Pb levels above 15 [mu]g/dL as representing a clear pattern of
adverse effects worthy of avoiding. This is in contrast to EPA's
identification of 30 [mu]g/dL as a safe blood Pb level for individual
children when the NAAQS was set in 1978. The Staff Paper further stated
that at blood Pb levels of 10-15 [mu]g/dL, there was a convergence of
evidence of Pb-induced interference with a diverse set of physiological
functions and processes, particularly evident in several independent
studies showing impaired neurobehavioral function and development.
Further, the available data did not indicate a clear threshold in this
blood Pb range. Rather, it suggested a continuum of health risks down
to the lowest levels measured.\108\
---------------------------------------------------------------------------
\108\ In 1991, the CDC reduced their advisory level for
children's blood Pb from 25 [mu]g/dL to 10 [mu]g/dL.
---------------------------------------------------------------------------
For the purposes of comparing the relative protectiveness of
alternative Pb NAAQS, the staff conducted analyses to estimate the
percentages of children with blood Pb levels above 10 [mu]g/dL and
above 15 [mu]g/dL for several air quality scenarios developed for a
small set of stationary source exposure case studies. The results of
the analyses of child populations living near two Pb smelters indicated
that substantial reductions in Pb exposure could be achieved through
just meeting the current Pb NAAQS. According to the best estimate
analyses, over 99.5% of children living in areas significantly affected
by the smelters would have blood Pb levels below 15
[[Page 29224]]
[mu]g/dL if the current standard was achieved. Progressive changes in
this number were estimated for the alternative monthly Pb NAAQS levels
evaluated in those analyses, which ranged from 1.5 [mu]g/m\3\ to 0.5
[mu]g/m\3\.
In light of the health effects evidence available at the time, the
1990 Staff Paper presented air quality, exposure, and risk analyses,
and other policy considerations, as well as the following staff
conclusions with regard to the primary Pb NAAQS (USEPA, 1990b, pp. xii
to xiv):
(1) ``The range of standards * * * should be from 0.5 to 1.5 [mu]g/
m\3\.''
(2) ``A monthly averaging period would better capture short-term
increases in lead exposure and would more fully protect children's
health than the current quarterly average.''
(3) ``The most appropriate form of the standard appears to be the
second highest monthly averages {sic{time} in a 3-year span. This form
would be nearly as stringent as a form that does not permit any
exceedances and allows for discounting of one `bad' month in 3 years
which may be caused, for example, by unusual meteorology.''
(4) ``With a revision to a monthly averaging time more frequent
sampling is needed, except in areas, like roadways remote from lead
point sources, where the standard is not expected to be violated. In
those situations, the current 1-in-6 day sampling schedule would
sufficiently reflect air quality and trends.''
(5) ``Because exposure to atmospheric lead particles occurs not
only via direct inhalation, but via ingestion of deposited particles as
well, especially among young children, the hi-volume sampler provides a
reasonable indicator for determining compliance with a monthly standard
and should be retained as the instrument to monitor compliance with the
lead NAAQS until more refined instruments can be developed.''
Based on its review of a draft Staff Paper, which contained the
above recommendations, the CASAC strongly recommended to the
Administrator that EPA should actively pursue a public health goal of
minimizing the Pb content of blood to the extent possible, and that the
Pb NAAQS is an important component of a multimedia strategy for
achieving that goal (CASAC, 1990, p. 4). In noting the range of levels
recommended by staff, CASAC recommended consideration of a revised
standard that incorporates a ``wide margin of safety, because of the
risk posed by Pb exposures, particularly to the very young whose
developing nervous system may be compromised by even low level
exposures'' (id., p. 3). More specifically, CASAC judged that a
standard within the range of 1.0 to 1.5 [mu]g/m\3\ would have
``relatively little, if any, margin of safety;'' that greater
consideration should be given to a standard set below 1.0 [mu]g/m\3\;
and, to provide perspective in setting the standard, it would be
appropriate to consider the distribution of blood Pb levels associated
with meeting a monthly standard of 0.25 [mu]g/m\3\, a level below the
range considered by staff (id.).
After consideration of the documents developed during the review,
EPA chose not to propose revision of the NAAQS for Pb. During the same
time period, the Agency published and embarked on the implementation of
a broad, multi-program, multi-media, integrated national strategy to
reduce Pb exposures (USEPA, 1991). As discussed above in section I.C.,
as part of implementing this integrated Pb strategy, the Agency focused
efforts primarily on regulatory and remedial clean-up actions aimed at
reducing Pb exposures from a variety of nonair sources judged to pose
more extensive public health risks to U.S. populations, as well as on
actions to reduce Pb emissions to air, particularly near stationary
sources.\109\
---------------------------------------------------------------------------
\109\ A description of the various programs implemented since
1990 to reduce Pb exposures, including the recent RRP rule, is
provided in section I.C.
---------------------------------------------------------------------------
2. Considerations in the Current Review
a. Evidence-Based Considerations
In considering the broad array of health effects evidence assessed
in the Criteria Document with respect to the adequacy of the current
standard, the discussion here, like that in the Staff Paper and ANPR,
focuses on those health endpoints associated with the Pb exposure and
blood levels most pertinent to ambient exposures. In so doing, EPA
gives particular weight to evidence available today that differs from
that available at the time the standard was set with regard to its
support of the current standard.
First, with regard to the sensitive population, the susceptibility
of young children to the effects of Pb is well recognized, in addition
to more recent recognition of effects of chronic or cumulative Pb
exposure with advancing age (CD, Sections 5.3.7 and pp. 8-73 to 8-75).
The prenatal period and early childhood are periods of increased
susceptibility to Pb exposures, with evidence of adverse effects on the
developing nervous system that generally appear to persist into later
childhood and adolescence (CD, Section 6.2).\110\ Thus, while the
sensitivity of the elderly and other particular subgroups is
recognized, as at the time the standard was set, young children
continue to be recognized as a key sensitive population for Pb
exposures.
---------------------------------------------------------------------------
\110\ For example, the following statement is made in the
Criteria Document ``Negative Pb impacts on neurocognitive ability
and other neurobehavioral outcomes are robust in most recent studies
even after adjustment for numerous potentially confounding factors
(including quality of care giving, parental intelligence, and
socioeconomic status). These effects generally appear to persist
into adolescence and young adulthood.'' (CD, p.E-9)
---------------------------------------------------------------------------
With regard to the exposure levels at which adverse health effects
occur, the current evidence demonstrates the occurrence of adverse
health effects at appreciably lower blood Pb levels than those
demonstrated by the evidence at the time the standard was set, at which
time the Agency identified 30 [mu]g/dL as the maximum safe blood Pb
level for individual children and 15 [mu]g/dL as the maximum safe
geometric mean blood Pb level for a population of children (as
described in section II.D.1.a above). This change in the evidence since
the time the standard was set is reflected in changes made by the CDC
in their advisory level for Pb in children's blood, and changes they
have made in their characterization of that level (as described in
section II.B.1.b). Although CDC recognized a level of 30 [mu]g/dL blood
Pb as warranting individual intervention in 1978 when the Pb NAAQS was
set, in 2005 they recognized the evidence of adverse health effects in
children with blood Pb levels below 10 [mu]g/dL and the data
demonstrating that no ``safe'' threshold for blood Pb had been
identified (CDC, 1991; CDC, 2005).
As summarized in section II.B above, the Criteria Document
describes current evidence regarding the occurrence of a variety of
health effects, including neurological effects in children associated
with blood Pb levels extending well below 10 [mu]g/dL (CD, Sections
6.2, 8.4 and 8.5).\111\ As stated
[[Page 29225]]
in the Criteria Document, ``The overall weight of the available
evidence provides clear substantiation of neurocognitive decrements
being associated in young children with blood-Pb concentrations in the
range of 5-10 [mu]g/dL, and possibly somewhat lower. Some newly
available analyses appear to show Pb effects on the intellectual
attainment of preschool and school age children at population mean
concurrent blood-Pb levels ranging down to as low as 2 to 8 [mu]g/dL''
(CD, p. E-9). With regard to the evidence of neurological effects at
these low levels, EPA notes, in particular (and discusses more
completely in section II.B.2.b above), the international pooled
analysis by Lanphear and others (2005), studies of individual cohorts
such as the Rochester, Boston, and Mexico City cohorts (Canfield et
al., 2003a; Canfield et al., 2003b; Bellinger and Needleman, 2003;
Tellez-Rojo et al., 2006), the study of African-American inner-city
children from Detroit (Chiodo et al., 2004), the cross-sectional study
of young children in three German cities (Walkowiak et al., 1998) and
the cross-sectional analysis of a nationally representative sample from
the NHANES III (collected from 1988-1994) (Lanphear et al., 2000). In
the study by Lanphear et al (2000), the mean blood Pb for the full
study group was 1.9 [mu]g/dL and the mean blood Pb level in the lowest
blood Pb subgroup with which a statistically significant association
with neurocognitive effects was found (individual blood Pb values <5
[mu]g/dL) was 1.7 [mu]g/dL (CD, pp. 6-31 to 6-32; Lanphear et al.,
2000; Auinger, 2008).\112\ These studies and associated limitations are
discussed above in section II.B.2.b.
---------------------------------------------------------------------------
\111\ For context, it is noted that the 2001-2004 median blood
level for children aged 1-5 of all races and ethnic groups is 1.6
[mu]g/dL, the median for the subset living below the poverty level
is 2.3 [mu]g/dL and 90th percentile values for these two groups are
4.0 [mu]g/dL and 5.4 [mu]g/dL, respectively. Similarly, the 2001-
2004 median blood level for black, non-hispanic children aged 1-5 is
2.5 [mu]g/dL, while the median level for the subset of that group
living below the poverty level is 2.9 [mu]g/dL and the median level
for the subset living in a household with income more than 200% of
the poverty level is 1.9 [mu]g/dL. Associated 90th percentile values
for 2001-2004 are 6.4 [mu]g/dL (for black, non-hispanic children
aged 1-5), 7.7 [mu]g/dL (for the subset of that group living below
the poverty level) and 4.1 [mu]g/dL (for the subset living in a
household with income more than 200% of the poverty level). (http://
www.epa.gov/envirohealth/children/body_burdens/b1-table.htm--then
click on ``Download a universal spreadsheet file of the Body Burdens
data tables'').
\112\ These findings include significant associations in some of
the study sample subsets of children, namely those with blood Pb
levels less than 10 [mu]g/dL, less than 7.5 [mu]g/dL, and less than
5 [mu]g/dL. The mean blood Pb level in the third subset was 1.7
[mu]g/dL (Auinger, 2008). A positive, but not statistically
significant association, was observed in the less than 2.5 [mu]g/dL
subset (mean blood Pb of 1.2 [mu]g/dL [Auinger, 2008]), although the
effect estimate for this subset was largest among all the subsets
(Lanphear et al., 2000). The lack of statistical significance for
this subset may be due to the smaller sample size of this subset
which would lead to lower statistical power.
---------------------------------------------------------------------------
As stated in the Criteria Document with regard to the
neurocognitive effects in children, the ``weight of overall evidence
strongly substantiates likely occurrence of type of effect in
association with blood-Pb concentrations in range of 5-10 [mu]g/dL, or
possibly lower, as implied by (???) [in associated Table 8-5 of
Criteria Document]. Although no evident threshold has yet been clearly
established for those effects, the existence of such effects at still
lower blood-Pb levels cannot be ruled out based on available data.''
(CD, p. 8-61). The Criteria Document further notes that any such
threshold may exist ``at levels distinctly lower than the lowest
exposures examined in these epidemiological studies'' (CD, p. 8-67).
i. Evidence-Based Framework Considered in the Staff Paper
In considering the adequacy of the current standard, the Staff
Paper considered the evidence in the context of the framework used to
determine the standard in 1978, as adapted to reflect the current
evidence. In so doing, the Staff Paper recognized that the health
effects evidence with regard to characterization of a threshold for
adverse effects has changed since the standard was set in 1978, as have
the Agency's views on the characterization of a safe blood Pb level. As
described in section II.D.1.a, parameters for this framework include
estimates for average nonair blood Pb level, and air-to-blood ratio, as
well as a maximum safe individual and/or geometric mean blood Pb level.
For this last parameter, the Staff Paper for the purposes of this
evaluation considered the lowest population mean blood Pb levels with
which some neurocognitive effects have been associated in the evidence.
As when the standard was set in 1978, there remain today
contributions to blood Pb levels from nonair sources. In 1978, the
Agency estimated the average blood Pb level for young children
associated with nonair sources to be 12 [mu]g/dL (as described in
section II.D.1.a). However, consistent with reductions since that time
in air Pb concentrations \113\ which contribute to blood Pb, nonair
contributions have also been reduced (as described in section II.A.4
above). The Staff Paper noted that the current evidence is limited with
regard to estimates of the aggregate reduction since 1978 of all nonair
sources to blood Pb and with regard to an estimate of current nonair
blood Pb levels (discussed in sections II.A.4). In recognition of
temporal reductions in nonair sources discussed in section II.A.4 and
in the context of estimates pertinent to an application of the 1978
framework, the CASAC Pb Panel recommended consideration of 1.0-1.4
[mu]g/dL or lower as an estimate of the nonair component of blood Pb
pertinent to average blood Pb levels (as more fully described in
section II.A.4 above; Henderson, 2007b).
---------------------------------------------------------------------------
\113\ Air Pb concentrations nationally are estimated to have
declined more than 90% since the early 1980s, in locations not known
to be directly influenced by stationary sources (Staff Paper, pp. 2-
22 to 2-23).
---------------------------------------------------------------------------
As in 1978, the evidence demonstrates that Pb in ambient air
contributes to Pb in blood, with the pertinent exposure routes
including both inhalation and ingestion (CD, Sections 3.1, 3.2, 4.2 and
4.4). In 1978, the evidence indicated a quantitative relationship
between ambient air Pb and blood Pb in terms of an air-to-blood ratio
that ranged from 1:1 to 1:2 (USEPA, 1977). In setting the standard, the
Agency relied on a ratio of 1:2, i.e., 2 [mu]g/dL blood Pb per 1 [mu]g/
m\3\ air Pb (as described in section II.D.1.a above). The Staff Paper
observed that ``[W]hile there is uncertainty and variability in the
absolute value of an air-to-blood relationship, the current evidence
indicates a notably greater ratio * * * e.g., on the order of 1:3 to
1:10'' (USEPA, 2007c).
Based on the information described above, the Staff Paper concluded
that young children remain the sensitive population of primary focus in
this review, ``there is now no recognized safe level of Pb in
children's blood and studies appear to show adverse effects at
population mean concurrent blood Pb levels as low as approximately 2
[mu]g/dL (CD, pp. 6-31 to 6-32; Lanphear et al., 2000)'' (USEPA,
2007c). The Staff Paper further stated that ``while the nonair
contribution to blood Pb has declined, perhaps to a range of 1.0-1.4
[mu]g/dL, the air-to-blood ratio appears to be higher at today's lower
blood Pb levels than the estimates at the time the standard was set,
with current estimates on the order of 1:3 to 1:5 and perhaps up to
1:10'' (USEPA, 2007c). Adapting the framework employed in setting the
standard in 1978, the Staff Paper concluded that ``the more recently
available evidence suggests a level for the standard that is lower by
an order of magnitude or more'' (USEPA, 2007c).
ii. Air-Related IQ Loss Evidence-Based Framework
Since completion of the Staff Paper and ANPR, the Agency has
further considered the evidence with regard to adequacy of the current
standard using an approach other than the adapted 1978 framework
considered in the Staff Paper. This alternative evidence-based
framework, referred to as the air-related IQ loss framework, shifts
focus from identifying an appropriate target population mean blood lead
level and instead focuses on the magnitude of effects of air-related Pb
on neurocognitive functions. This framework builds on a recommendation
by the CASAC Pb Panel to consider the evidence in a more quantitative
manner,
[[Page 29226]]
and is discussed in more detail below in section II.E.3.a, concerning
the level of the standard.
In this air-related IQ loss framework, we have drawn from the
entire body of evidence as a basis for concluding that there are causal
associations between air-related Pb exposures and population IQ
loss.\114\ We have also drawn more quantitatively from the evidence by
using evidence-based C-R functions to quantify the association between
air Pb concentrations and air-related population mean IQ loss. Thus,
this framework more fully considers the evidence with regard to the
concentration-response relationship for the effect of Pb on IQ, and it
also draws from estimates for air-to-blood ratios.
---------------------------------------------------------------------------
\114\ For example, as stated in the Criteria Document,
``Fortunately, there exists a large database of high quality studies
on which to base inferences regarding the relationship between Pb
exposure and neurodevelopment. In addition, Pb has been extensively
studied in animal models at doses that closely approximate the human
situation. Experimental animal studies are not compromised by the
possibility of confounding by such factors as social class and
correlated environmental factors. The enormous experimental animal
literature that proves that Pb at low levels causes neurobehavioral
deficits and provides insights into mechanisms must be considered
when drawing causal inferences (Bellinger, 2004; Davis et al., 1990;
U.S. Environmental Protection Agency, 1986a, 1990).'' (CD, p. 6-75)
---------------------------------------------------------------------------
While we note the evidence of steeper slope for the C-R
relationship for blood Pb concentration and IQ loss at lower blood Pb
levels (described in sections II.B.2.b and II.E.3.a), for purposes of
consideration of the adequacy of the current standard we are concerned
with the C-R relationship for blood Pb levels that would be associated
with exposure to air-related Pb at the level of the current standard.
For this purpose, we have focused on a median linear estimate of the
slope of the C-R function for blood Pb levels up to, but no higher
than, 10 [mu]g/dL (described in section II.B.2.b above). The median
slope estimate is -0.9 IQ points per [mu]g/dL blood Pb \115\ (CD, p. 8-
80).
---------------------------------------------------------------------------
\115\ As noted above (in section II.B.2.b), this slope is
similar to the slope for the below 10 [mu]g/dL piece of the
piecewise model used in the RRP rule economic analysis.
---------------------------------------------------------------------------
Applying estimates of air-to-blood ratios ranging from 1:3 to 1:5,
drawing from the discussion of air-to-blood ratios in section II.B.1.c
above, a population of children exposed at the current level of the
standard might be expected to result in an average air-related blood Pb
level above 4 [mu]g/dL.\116\ Multiplying these blood Pb levels by the
slope estimate, identified above, for blood Pb levels extending up to
10 [mu]g/dL (-0.9 IQ points per [mu]g/dL), would imply an average air-
related IQ loss for such a group of children on the order of 4 or more
IQ points.
---------------------------------------------------------------------------
\116\ This is based on the calculation in which 1.5 [mu]g/m\3\
is multiplied by a ratio of 3 [mu]g blood Pb per 1 [mu]g/m\3\ air Pb
to yield an air-related blood Pb estimate of 4.5 [mu]g/dL; using a
1:5 ratio yields an estimate of 7.5 [mu]g/dL. As with the 1978
framework considered in the Staff Paper, the context for use of the
air-to-blood ratio here is a population being exposed at the level
of the standard.
---------------------------------------------------------------------------
b. Exposure- and Risk-Based Considerations
As discussed above in section II.C, we have estimated exposures and
health risks associated with air quality that just meets the current
standard to help inform judgments about whether or not the current
standard provides adequate protection of public health, taking into
account key uncertainties associated with the estimated exposures and
risks (summarized above in section II.C and more fully in the Risk
Assessment Report).
As discussed above, children are the sensitive population of
primary focus in this review. The exposure and risk assessment
estimates Pb exposure for children (less than 7 years of age), and
associated risk of neurocognitive effects in terms of IQ loss. In
addition to the risks (IQ loss) that were quantitatively estimated, EPA
recognizes that there may be long-term adverse consequences of such
deficits over a lifetime, and there are other, unquantified adverse
neurocognitive effects that may occur at similarly low exposures which
might additionally contribute to reduced academic performance, which
may have adverse consequences over a lifetime (CD, pp. 8-29 to 8-
30).\117\ Other impacts at low levels of childhood exposure that were
not quantified in the risk assessment include: other neurological
effects (sensory, motor, cognitive and behavioral), immune system
effects (including some related to allergic responses and asthma), and
early effects related to anemia. Additionally, as noted in section
II.B.2, other health effects evidence demonstrates associations between
Pb exposure and adverse health effects in adults (e.g., cardiovascular
and renal effects).\118\
---------------------------------------------------------------------------
\117\ For example, the Criteria Document notes particular
findings with regard to academic achievement as ``suggesting that
Pb-sensitive neuropsychological processing and learning factors not
reflected by global intelligence indices might contribute to reduced
performance on academic tasks'' (CD, pp. 8-29 to 8-30).
\118\ The weight of the evidence differs for the different
endpoints.
---------------------------------------------------------------------------
As noted in the Criteria Document, a modest change in the
population mean of a health index, that is quantified for each
individual, can have substantial implications at the population level
(CD, p. 8-77, Sections 8.6.1 and 8.6.2; Bellinger, 2004; Needleman et
al., 1982; Weiss, 1988; Weiss, 1990)). For example, for an individual
functioning in the low range of IQ due to the influence of risk factors
other than Pb, a Pb-associated IQ loss of a few points might be
sufficient to drop that individual into the range associated with
increased risk of educational, vocational, and social handicap (CD, p.
8-77), while such a decline might create less significant impacts for
the individual near the mean of the population. Further, given a
uniform manifestation of Pb-related decrements across the range of IQ
scores in a population, a downward shift in the mean IQ value is
associated not only with a substantial increase in the percentage of
individuals achieving very low scores, but also with substantial
decreases in percentages achieving very high scores (CD, p. 8-81). The
CASAC Pb Panel has advised on this point that ``a population loss of 1-
2 IQ points is highly significant from a public health perspective''
(Henderson, 2007a, p. 6).
In considering exposure and risk estimates with regard to adequacy
of the current standard, EPA has focused on IQ loss for air-related
exposure pathways. As described in section II.C.2.e above, limitations
in our data and modeling tools have resulted in an inability to develop
specific estimates such that we have approximated estimates for the
air-related pathways, bounded on the low end by exposure/risk estimated
for the ``recent air'' category and on the upper end by the exposure/
risk estimated for the ``recent air'' plus ``past air'' categories.
Thus, the following discussion presents air-related IQ loss estimates
in terms of upper and lower bounds. In addition, as noted above
(section II.C.3.b), this discussion focuses predominantly on risk
estimates derived using the log-linear with low-exposure linearization
(LLL) C-R function, with the range associated with the other three
functions used in the assessment also being noted. Further, air-related
risk estimates are presented for the median and for an upper percentile
(i.e., the 95th percentile of the population assessed).
EPA and CASAC recognize uncertainties in the risk estimates in the
tails of the distribution and consequently the 95th percentile is
reported as the estimate of the high end of the risk distribution
(Henderson, 2007b, p. 3). In so doing, however, EPA notes that it is
important to consider that there are individuals in the population
expected to have higher risk, particularly in light of the risk
management objectives for the current standard which was set in 1978 to
[[Page 29227]]
protect the 99.5th percentile. Further, we note an increased
uncertainty in our estimates of air-related risk for the upper
percentiles, such as the 95th percentile, due to limitations in the
data and tools available to us to estimate pathway contributions to
blood Pb and associated risk for individuals at the upper ends of the
distribution.
In order to consider exposure and risk associated with the current
standard, EPA developed estimates for a case study based on air quality
projected to just meet the standard in a location of the country where
air concentrations currently do not meet the current standard (the
primary Pb smelter case study). Estimates of median air-related IQ loss
associated with just meeting the current NAAQS in the primary Pb
smelter case study subarea had a lower bound estimate of <3.2 points IQ
loss (``recent air'' category of Pb exposures) and an upper bound
estimate of <9.4 points IQ loss (``recent air'' plus ``past air''
category) for the range of C-R functions (Table 3). This estimate
(recent air plus past air) for the subarea based on the LLL C-R
function is 6.0 points IQ loss for the median and 8.0 points IQ loss
for the 95th percentile, with which we note a greater uncertainty than
for the median estimate (as discussed above).\119\ Modeling limitations
have affected our ability to derive lower bound estimates for this case
study (as described above in section II.C.2.c).
---------------------------------------------------------------------------
\119\ We note that while we have termed risk estimates derived
for the sum of ``recent air'' plus ``past air'' exposure pathways as
``upper bound'' estimates of air-related risk, the primary Pb
smelter subarea is an area where soil has been remediated and thus
does not reflect any historical deposition. Further, soil Pb
concentrations in this area are not stable and may be increasing,
seeming to indicate ongoing response to current atmospheric
depositon in the area. Thus, for this case study, the ``recent air''
plus ``past air'' estimates are less of an ``upper bound'' for air-
related risk than in other case studies where historical Pb
deposition may have some representation in the ``past air'' soil
ingestion pathway.
---------------------------------------------------------------------------
Additionally, we developed estimates of blood Pb and associated IQ
loss associated with the current standard for the urban case studies.
We note that we consider it extremely unlikely that air concentrations
in urban areas across the U.S. that are currently well below the
current standard would increase to just meet the standard. However, we
recognize the potential, although not the likelihood, for air Pb
concentrations in some limited areas currently well below the standard
to increase to just meet the standard by way of, for example, expansion
of existing sources (e.g., facilities operating as secondary smelters
may exercise previously used capabilities as primary smelters) or by
the congregation of multiple Pb sources in adjacent locations. We have
simulated this scenario (increased Pb concentrations to just meet the
current standard) in a general urban case study and three location-
specific urban case studies. For the location-specific urban case
studies, we note substantial uncertainty in simulating how the profile
of Pb concentrations might change in the hypothetical case where
concentrations increase to just meet the current standard.
Turning first to the exposure/risk estimates for the current NAAQS
scenario simulated for the general urban case study, which is a
simplified representation of a location within an urban area (described
in section II.C.2.h above), median estimates of air-related IQ loss
range from 1.5 to 7.7 points (across all four C-R functions), with an
estimate based on the LLL function bounded at the low end by 3.4 points
and at the high end by 4.8 points (Table 3). At the 95th percentile for
total IQ loss (LLL estimate), IQ loss associated with air-related Pb is
estimated to fall somewhere between 5.5 and 7.6 points (Staff Paper,
Table 4-6).
In considering the estimates for the three location-specific urban
case studies, we first note the extent to which exposures associated
with increased air Pb concentrations that simulate just meeting the
current standard are estimated to increase blood Pb levels in young
children. The magnitude of this for the median total blood Pb ranges
from 0.3 [mu]g/dL (an increase of 20 percent) in the case of the
Cleveland study area (where the highest monitor is estimated to be
approximately one fourth of the current NAAQS), up to approximately 1
[mu]g/dL (an increase of 50 to 70%) for the Chicago and Los Angeles
study areas, where the highest monitor is estimated to be at or below
one tenth of the current NAAQS (Table 1). Median estimates of air-
related risk for these case studies range from 0.6 points IQ loss
(recent air estimate using low-end C-R function) to 7.4 points IQ loss
(recent plus past air estimate using the high-end C-R function). The
corresponding estimates based on the LLL C-R function range from 2.7
points (lowest location-specific recent air estimate) to 4.7 points IQ
loss (highest location-specific recent plus past air estimate). The
comparable estimates of air-related risk for children at the 95th
percentile in these three case studies range from 2.6 to 7.6 points IQ
loss for the LLL C-R function (Staff paper, Table 4-6), although we
note increased uncertainty in the magnitude of these 95th percentile
air-related estimates.
Another way in which the risk assessment results might be
considered is by comparing current NAAQS scenario estimates to current
conditions, although in so doing, it is important to recognize that, as
stated below and described in section II.C., this will underestimate
air-related impacts associated with the current NAAQS. In making such a
comparison of estimates for the three location-specific urban case
studies, the estimated difference in total Pb-related IQ loss for the
median child is about 0.5 to 1.4 points using the LLL C-R function and
a similar magnitude of difference is estimated for the 95th percentile.
The corresponding comparison for the general urban case study indicates
the current NAAQS scenario median total Pb-related IQ loss is 1.1 to
1.3 points higher than the two current conditions scenarios. As
described in section II.C, such comparisons are underestimates of air-
related impacts brought about as a result of increased air Pb
concentrations, and consequently they are inherently underestimates of
the true impact of an increased NAAQS level on public health.
In considering the exposure/risk information with regard to
adequacy of the current standard, the Staff Paper first considered the
estimates described above, particularly those associated with air-
related risk.\120\ The Staff Paper described these estimates for the
current NAAQS as being indicative of levels of IQ loss associated with
air-related risk that may ``reasonably be judged to be highly
significant from a public health perspective'' (USEPA, 2007c).
---------------------------------------------------------------------------
\120\ As recognized in section III.B.2.d above, to simulate air
concentrations associated with the current NAAQS, a proportional
roll-up of concentrations from those for current conditions was
performed for the location-specific urban case studies. This was not
necessary for the primary Pb smelter case study in which air
concentrations currently exceed the current standard, nor for the
general urban case study.
---------------------------------------------------------------------------
The Staff Paper also describes a different risk metric that
estimated differences in the numbers of children with different amounts
of Pb-related IQ loss between air quality scenarios for current
conditions and for the current NAAQS in the three location-specific
urban case studies. For example, estimates of the additional number of
children with IQ loss greater than one point (based on the LLL C-R
function) in these three study areas, for the current NAAQS scenario as
compared to current conditions, range from 100 to 6,000 across the
three locations (as shown above in Table 5). The corresponding
estimates for the additional number of children with IQ
[[Page 29228]]
loss greater than seven points, for the current NAAQS as compared to
current conditions, range from 600 to 66,000 (as shown above in Table
6). These latter values for the change in incidence of children with
greater than seven points Pb-related IQ loss represent 5 to 17 percent
of the children (aged less than 7 years of age) in these study areas.
This increase corresponds to approximately a doubling in the number of
children with this magnitude of Pb-related IQ loss in the study area
most affected. The Staff Paper concluded that these estimates indicate
the potential for significant numbers of children to be negatively
affected if air Pb concentrations increased to levels just meeting the
current standard.
Beyond the findings related to quantified IQ loss, the Staff Paper
recognized the potential for other, unquantified adverse effects that
may occur at similarly low exposures. In summary, the Staff Paper
concluded that taken together, ``the quantified IQ effects associated
with the current NAAQS and other, nonquantified effects are important
from a public health perspective, indicating a need for consideration
of revision of the standard to provide an appreciable increase in
public health protection'' (USEPA, 2007c).
3. CASAC Advice and Recommendations and Public Comment
CASAC's recommendations in this review builds upon the CASAC
recommendations during the 1990 review, which also advised on
consideration of more health protective NAAQS. In CASAC's review of the
1990 Staff Paper, as discussed in Section II.D.1.b, they generally
recommended consideration of levels below 1.0 [mu]g/m\3\, specifically
recommended analyses of a standard set at 0.25 [mu]g/m\3\, and also
recommended a revision to a monthly averaging time (CASAC, 1990).
In its letter to the Administrator subsequent to consideration of
the ANPR, the final Staff Paper and the final Risk Assessment Report,
the CASAC Pb Panel unanimously and fully supported ``Agency staff's
scientific analyses in recommending the need to substantially lower the
level of the primary (public-health based) Lead NAAQS, to an upper
bound of no higher than 0.2 [mu]g/m\3\ with a monthly averaging time''
(Henderson, 2008, p. 1). This recommendation is consistent with their
recommendations conveyed in two earlier letters in the course of this
review (Henderson, 2007a, 2007b). Further, in their advice to the
Agency over the course of this review, CASAC has provided rationale for
their conclusions that has included their statement that the current Pb
NAAQS ``are totally inadequate for assuring the necessary decreases of
lead exposures in sensitive U.S. populations below those current health
hazard markers identified by a wealth of new epidemiological,
experimental and mechanistic studies'', and stated that ``Consequently,
it is the CASAC Lead Review Panel's considered judgment that the NAAQS
for Lead must be decreased to fully-protect both the health of children
and adult populations'' (Henderson, 2007a, p. 5). CASAC drew support
for their recommendation from the current evidence, described in the
Criteria Document, of health effects occurring at dramatically lower
blood Pb levels than those indicated by the evidence available when the
standard was set and of a recognition of effects that extend beyond
children to adults.
The Agency has also received comments from the public on drafts of
the Staff Paper and related technical support document, as well as on
the ANPR.\121\ Public comments received to date that have addressed
adequacy of the current standard overwhelmingly concluded that the
current standard is inadequate and should be substantially revised, in
many cases suggesting specific reductions to a level at or below 0.2
[mu]g/m\3\. Two comments were received from specific industries
expressing the view that the current standard might need little or no
adjustment. One comment received early in the review stated that
current conditions justified revocation of the standard.
---------------------------------------------------------------------------
\121\ All written comments submitted to the Agency are available
in the docket for this rulemaking, are transcripts of the public
meetings held in conjunction with CASAC's review of the Staff Paper,
the Risk Assessment Report, the Criteria Document and the ANPR.
---------------------------------------------------------------------------
4. Administrator's Proposed Conclusions Concerning Adequacy
Based on the large body of evidence concerning the public health
impacts of Pb, including significant new evidence concerning effects at
blood Pb concentrations substantially below those identified when the
current standard was set, the Administrator proposes that the current
standard does not protect public health with an adequate margin of
safety and should be revised to provide additional public health
protection.
In considering the adequacy of the current standard, the
Administrator has carefully considered the conclusions contained in the
Criteria Document, the information, exposure/risk assessments,
conclusions, and recommendations presented in the Staff Paper, the
advice and recommendations from CASAC, and public comments received on
the ANPR and other documents to date.
The Administrator notes that the body of available evidence,
summarized above in section III.B and discussed in the Criteria
Document, is substantially expanded from that available when the
current standard was set three decades ago. The Criteria Document
presents evidence of the occurrence of health effects at appreciably
lower blood Pb levels than those demonstrated by the evidence at the
time the standard was set. Subsequent to the setting of the standard,
the Pb NAAQS criteria review during the 1980s and the current review
have provided (a) expanded and strengthened evidence of still lower Pb
exposure levels associated with slowed physical and neurobehavioral
development, lower IQ, impaired learning, and other indicators of
adverse neurological impacts; and (b) other effects of Pb on
cardiovascular function, immune system components, calcium and vitamin
D metabolism and other health endpoints (discussed fully in the
Criteria Document).
The Administrator notes particularly the robust evidence of
neurotoxic effects of Pb exposure in children, both with regard to
epidemiological and toxicological studies. While blood Pb levels in
U.S. children have decreased notably since the late 1970s, newer
studies have investigated and reported associations of effects on the
neurodevelopment of children with these more recent blood Pb levels.
The toxicological evidence includes extensive experimental laboratory
animal evidence that substantiates well the plausibility of the
epidemiologic findings observed in human children and expands our
understanding of likely mechanisms underlying the neurotoxic effects.
Further, the Administrator notes the current evidence that suggests a
steeper dose-response relationship at these lower blood Pb levels than
at higher blood Pb levels, indicating the potential for greater
incremental impact associated with exposure at these lower levels.
In addition to the evidence of health effects occurring at
significantly lower blood Pb levels, the Administrator recognizes that
the current health effects evidence together with findings from the
exposure and risk assessments (summarized above in section III.B), like
the information available at the time the standard was set, supports
our finding that air-related Pb exposure pathways contribute to blood
Pb levels in young children, by inhalation and ingestion. Furthermore,
the Administrator takes
[[Page 29229]]
note of the information that suggests that the air-to-blood ratio
(i.e., the quantitative relationship between air concentrations and
blood concentrations) is now likely larger, when air inhalation and
ingestion are considered, than that estimated when the standard was
set.
Based on evidence discussed above, the Administrator first
considered the evidence in the context of an adaptation of the 1978
framework, as presented in the Staff Paper, recognizing that the health
effects evidence with regard to characterization of a threshold for
adverse effects has changed dramatically since the standard was set in
1978. As discussed above, however, the 1978 framework was premised on
an evidentiary basis that clearly identified an adverse health effect
and a health-based policy judgment that identified a level that would
be safe for an individual child with respect to this adverse health
effect. The adaptation to the 1978 framework applies this framework to
a situation where there is no longer an evidentiary basis to determine
a safe level for individual children. In addition, this approach does
not address explicitly what magnitude of effect should be considered
adverse. Given these two limitations, the Administrator has focused
primarily instead on the air-related IQ loss evidence-based framework
described above in considering the adequacy of the current standard.
In considering the application the air-related IQ loss framework to
the current evidence as discussed above in section II.D.2.a, the
Administrator notes that this framework suggests an average air-related
IQ loss for a population of children exposed at the level of the
current standard on the order of 4 or more IQ points. The Administrator
judges that an air-related IQ loss of this magnitude is large from a
public health perspective and that this evidence-based framework
supports a conclusion that the current standard does not protect public
health with an adequate margin of safety. Further, the Administrator
believes that the current evidence indicates the need for a standard
level that is substantially lower than the current level to provide
increased public health protection, especially for at-risk groups,
including most notably children, against an array of effects, most
importantly including effects on the developing nervous system.
The Administrator has also considered the results of the exposure
and risk assessments conducted for this review, which provides some
further perspective on the potential magnitude of air-related IQ loss.
However, taking into consideration the uncertainties and limitations in
the assessments, notably including questions as to whether the
assessment scenarios that roll up current air quality to simulate just
meeting the current standard are realistic in wide areas across the
U.S., the Administrator has not placed primary reliance on the exposure
and risk assessments. Nonetheless, the Administrator observes that in
areas projected to just meet the current standard, the quantitative
estimates of IQ loss associated with air-related Pb, as summarized
above in section II.D.2.b, indicate risk of a magnitude that in his
judgment is significant from a public health perspective. Further,
although the current monitoring data indicate few areas with airborne
Pb near or just exceeding the current standard, the Administrator
recognizes significant limitations with the current monitoring network
and thus the potential that the prevalence of such levels of Pb
concentrations may be underestimated by currently available data.
The Administrator believes that the air-related blood Pb and IQ
loss estimates discussed in the Staff Paper and Risk Assessment Report,
summarized above, as well as the estimates of air-related IQ loss
suggested by this evidence-based framework, are important from a public
health perspective and are indicative of potential risks to susceptible
and vulnerable groups. In reaching this proposed judgment, the
Administrator considered the following factors: (1) The estimates of
blood Pb and IQ loss for children from air-related Pb exposures
associated with the current standard, (2) the estimates of numbers of
children with different amounts of increased Pb-related IQ loss
associated with the current standard, (3) the variability within and
among areas in both the exposure and risk estimates, (4) the
uncertainties in these estimates, and (5) the recognition that there is
a broader array of Pb-related adverse health outcomes for which risk
estimates could not be quantified and that the scope of the assessment
was limited to a sample of case studies and to some but not all at-risk
populations, leading to an incomplete estimation of public health
impacts associated with Pb exposures across the country.\122\ In
addition to the evidence-based and risk-based conclusions described
above, the Administrator also notes that it was the unanimous
conclusion of the CASAC Panel that EPA needed to ``substantially
lower'' the level of the primary Pb NAAQS to fully protect the health
of children and adult populations (Henderson, 2007a, 2007b, 2008).
---------------------------------------------------------------------------
\122\ While recognizing that there are significant uncertainties
associated with the risk estimates from the case studies, EPA places
an appropriate weight on the risk assessment results for purposes of
evaluating the adequacy of the current standard, given the strength
of the evidence of the existence of effects at blood Pb levels
associated with exposures at the level of the current standard, the
magnitude of the IQ losses that are estimated, and the consistency
of these IQ losses with the estimates of IQ loss derived from the
alternative evidence-based framework. The weight to place on the
risk assessment results for purposes of evaluating alterative levels
of the standard is discussed later in the discussion on the level of
the standard.
---------------------------------------------------------------------------
Based on all of these considerations, the Administrator proposes
that the current Pb standard is not requisite to protect public health
with an adequate margin of safety because it does not provide
sufficient protection, and that the standard should be revised to
provide increased public health protection, especially for members of
at-risk groups.
E. Conclusions on the Elements of the Standard
The four elements of the standard--indicator, averaging time, form,
and level--serve to define the standard and must be considered
collectively in evaluating the health and welfare protection afforded
by the standard. In considering revisions to the current primary Pb
standard, as discussed in the following sections, EPA considers each of
the four elements of the standard as to how they might be revised to
provide a primary standard for Pb that is requisite to protect public
health with an adequate margin of safety. Considerations and proposed
conclusions on indicator are discussed in section II.E.1, and on
averaging time and form in section II.E.2. Considerations and proposed
conclusions on a level for a Pb NAAQS with a Pb-TSP indicator are
discussed in section II.E.3, and considerations on a level for a Pb
NAAQS with a Pb-PM10 indicator are discussed in section
II.E.4.
1. Indicator
The indicator for the current standard is Pb-TSP (as described in
section II.D.1.a above).\123\ When the standard was set in 1978, the
Agency proposed Pb-TSP as the indicator, but considered identifying Pb
in particulate matter less than or equal to 10 [mu]m in diameter (Pb-
PM10) as the indicator. EPA had received comments expressing
concern
[[Page 29230]]
that because only a fraction of airborne particulate matter is
respirable, an air standard based on total air Pb would be
unnecessarily stringent. The Agency responded that while it agreed that
some Pb particles are too small or too large to be deposited in the
respiratory system, a significant component of exposures can be
ingestion of materials contaminated by deposition of Pb from the air.
In addition to the route of ingestion and absorption from the
gastrointestinal tract, nonrespirable Pb in the environment may, at
some point, become respirable through weathering or mechanical action.
EPA concluded that total airborne Pb, both respirable and nonrespirable
fractions, should be addressed by the air standard (43 FR 46251). The
federal reference method (FRM) for Pb-TSP specifies the use of the
high-volume FRM sampler for TSP.
---------------------------------------------------------------------------
\123\ The current standard specifies the measurement of airborne
Pb with a high-volume TSP federal reference method (FRM) sampler
with atomic absorption spectrometry of a nitric acid extract from
the filter for Pb, or with an approved equivalent method.
---------------------------------------------------------------------------
In the 1990 Staff Paper, this issue was reconsidered in light of
information regarding limitations of the high-volume sampler used for
the Pb-TSP measurements, and the continued use of Pb-TSP as the
indicator was recommended in the Staff Paper (USEPA, 1990):
Given that exposure to lead occurs not only via direct
inhalation, but via ingestion of deposited particles as well,
especially among young children, the hi-vol provides a more complete
measure of the total impact of ambient air lead. * * * Despite its
shortcomings, the staff believes the high-volume sampler will
provide a reasonable indicator for determination of compliance * * *
In the current review, the Staff Paper evaluated the evidence with
regard to the indicator for a revised primary standard. This evaluation
included consideration of the basis for using Pb-TSP as the current
indicator, information regarding the sampling methodology for the
current indicator, and CASAC advice with regard to indicator (described
below). Based on this evaluation, the Staff Paper recommended retaining
Pb-TSP as the indicator for the primary standard. The Staff Paper also
recommended activities intended to encourage collection and development
of datasets that will improve our understanding of national and site-
specific relationships between Pb-PM10 (collected by low-
volume sampler) and Pb-TSP to support a more informed consideration of
indicator during the next review. The Staff Paper suggested that such
activities might include describing a federal equivalence method (FEM)
in terms of PM10 and allowing its use for a TSP-based
standard in certain situations, such as where sufficient data are
available to adequately demonstrate a relationship between Pb-TSP and
Pb-PM10 or, in combination with more limited Pb-TSP
monitoring, in areas where Pb-TSP data indicate Pb levels well below
the NAAQS level.
The ANPR further identified issues and options associated with
consideration of the potential use of Pb-PM10 data for
judging attainment or nonattainment with a Pb-TSP NAAQS. These issues
included the impact of controlling Pb-PM10 for sources
predominantly emitting Pb in particles larger than those captured by
PM10 monitors \124\ (i.e., ultra-coarse), \125\ and the
options included potential application of Pb-PM10 FRM/FEMs
at sites with established relationships between Pb-TSP and Pb-
PM10, and use of Pb-PM10 data, with adjustment,
as a surrogate for Pb-TSP data. The ANPR broadly solicited comment in
these areas.
---------------------------------------------------------------------------
\124\ For simplicity, the discussion here and below speaks as if
PM10 samplers have a sharp size cut-off. In reality, they
have a size selection behavior in which 50% of particles 10 microns
in size are captured, with a progressively higher capture rate for
smaller particles and a progressively lower capture rate for larger
particles. The ideal capture efficiency curve for PM10
samplers specifies that particles above 15 microns not be captured
at all, although real samplers may capture a very small percentage
of particles above 15 microns. TSP samplers have 50% capture points
in the range of 25 to 50 microns, which is broad enough to include
virtually all particles capable of being transported any significant
distance from their source except under extreme wind events. As
explained below, the capture efficiency of a high-volume TSP sampler
for any given size particle is affected by wind speed and wind
direction.
\125\ In this notice, we use ``ultra-coarse'' to refer to
particles collected by a TSP sampler but not by a PM10
sampler (we note that CASAC has variously also referred to these
particles as ``very coarse'' or ``larger coarse-mode'' particles),
``fine'' to refer to particles collected by a PM2.5
sampler, and ``coarse'' to refer to particles collected by a
PM10 sampler but not by a PM2.5 sampler,
recognizing that there will be some overlap in the particle sizes in
the three types of collected material.
---------------------------------------------------------------------------
In the current review, both the CASAC Pb Panel and members of the
CASAC Ambient Air Monitoring and Methods (AAMM) Subcommittee have
recommended that EPA consider a change in the indicator to
PM10, utilizing low-volume PM10 sampling
(Henderson, 2007a, 2007b, 2008; Russell, 2008). \126\ In their January
2008 letter, the CASAC Lead Panel unanimously recommended that EPA
revise the Pb NAAQS indicator to rely on low-volume PM10
sampling (Henderson, 2008). They indicated support for their
recommendation in a range of areas. First, they noted poor precision in
high-volume TSP sampling, wide variation in the upper particle size-cut
as a function of wind speed and direction, and greater difficulties in
capturing the spatial non-homogeneity of ultra-coarse particles with a
national monitoring network. They stated that the low-volume
PM10 collection method is a much more accurate and precise
collection method, and would provide a more representative
characterization on a large spatial scale of monitored particles which
remain airborne longer, thus providing a characterization that is more
broadly representative of ambient exposures over large spatial scales.
They also noted the automated sequential sampling capability of low-
volume PM10 monitors which would be particularly useful if
the averaging time is revised (i.e., to a monthly averaging time, as
recommended by CASAC), which, in CASAC's view would necessitate an
increased monitoring frequency. Further, they noted the potential for
utilization of the more widespread PM10 sampling network
(Henderson, 2007a, 2007b, 2008).\127\ In their advice, CASAC also
stated that they ``recognize the importance of coarse dust
contributions to total Pb ingestion and acknowledge that TSP sampling
is likely to capture additional very coarse particles which are
excluded by PM10 samplers'' (Henderson 2007b). They
suggested that an adjustment of the NAAQS level would accommodate the
loss of these ultra-coarse Pb particles, and that development of such a
quantitative adjustment might appropriately be based on concurrent Pb-
PM10 and Pb-TSP sampling data \128\ (Henderson, 2007a,
2007b, 2008).
---------------------------------------------------------------------------
\126\ ``Low-volume PM10 sampling'' refers to sampling
using any of a number of monitor models that draw 16.67 liters/
minute (1 m3/hour) of air through the filter, in contrast
to ``high-volume'' sampling of either TSP or PM10 in
which the monitor draws 1500 liters/minute (90 m3/hour).
All commercial TSP FRM samplers at this time are high-volume
samplers; both high-volume and low-volume PM10 FRM
samplers are available. Low-volume sampling is the more recently
introduced method. Low-volume and high-volume samplers differ in
many other ways also, including filter size, accuracy of the flow
control, and degree of computerization.
\127\ EPA notes that costs, including those of operating a
monitoring network, may not be considered in establishing or
revising the NAAQS.
\128\ In their advice, CASAC recognized the potential for site-
to-site variability in the relationship between Pb-TSP and Pb-
PM10 (Henderson, 2007a, 2007b). They also stated in their
September 2007 letter, ``The Panel urges that PM10
monitors, with appropriate adjustments, be used to supplement the
data. * * * A single quantitative adjustment factor could be
developed from a short period of collocated sampling at multiple
sites; or a PM10 Pb/TSP Pb 'equivalency ratio' could be
determined on a regional or site-specific basis.''
---------------------------------------------------------------------------
The Agency received comments on the discussion of the indicator in
the ANPR from several state and local agencies and national/regional
air pollution control organizations, as well as a national
environmental organization. These public comments
[[Page 29231]]
were somewhat mixed. Most of these commenters recommended maintaining
Pb-TSP as the indicator to ensure that Pb emitted in larger particles
is not overlooked by the Pb NAAQS. Some of those comments and others
suggested keeping TSP as the indicator but revising the FRM to a low-
volume TSP method \129\ and considering tighter sampling height
criteria to reduce variability.\130\ Others, in considering a potential
PM10-based indicator or the use of PM10 data as a
surrogate for Pb-TSP, noted the need for characterization of the
relationship between Pb-PM10 and Pb-TSP, which varies with
proximity to some sources. One state agency and a national organization
of regulatory air agencies expressed clear support for revising the
indicator to Pb-PM10, predominantly citing advantages
associated with improved technology and efficiency in data collection.
---------------------------------------------------------------------------
\129\ The Pb-TSP FRM specification, 40 CFR 50 appendix G,
currently explicitly requires the use of the high-volume TSP FRM
sampler which is required by appendix B for the mass of TSP.
Therefore it would require amendments to 40 CFR 50 appendix B and/or
G (or a new dedicated appendix) to establish a low-volume TSP
sampler as the only FRM, or as an alternative FRM, for TSP and/or
Pb-TSP measurement. A number of researchers have utilized both self-
built and commercially available low-volume TSP samplers in ambient
air studies. Typically, these samplers are identical to low-volume
PM10 FRM samplers with the exception that their inlets
and other size separation devices (or lack thereof) are aimed at
collecting TSP. EPA is not aware of any rigorous evaluation of the
performance of these available, non-designated low-volume TSP
samplers or their equivalence to the TSP FRM. No one has applied to
date for designation of a low-volume TSP sampler as a FEM, either
for TSP measurement per se or for purposes of Pb-TSP measurement.
\130\ Currently, probe heights for Pb-TSP and PM10
sampling are allowed to be between 2 and 15 meters above ground
level for neighborhood-scale monitoring sites (those intended to
represent concentrations over a relatively large area around the
site) and between 2 and 7 meters for microscale sites. Near very
low-height sources of TSP, including fugitive dust sources at ground
level, concentrations of TSP, especially the concentrations of
particles larger than 10 microns, can vary substantially across this
height range with higher concentrations closer to the ground; near-
ground concentrations can also vary more in time than concentrations
higher up.
---------------------------------------------------------------------------
In considering these issues concerning the appropriate indicator,
EPA takes note of previous Agency conclusions that the health evidence
indicates that Pb in all particle size fractions, not just respirable
Pb, contributes to Pb in blood and to associated health effects.
Further, the evidence and exposure/risk estimates in the current review
indicate that ingestion pathways dominate air-related exposure. Lead is
unlike other criteria pollutants, where inhalation of the airborne
pollutant is the key contributor to exposure. For Pb it is the quantity
of Pb in ambient particles with the potential to deposit indoors or
outdoors, thereby leading to a role in ingestion pathways, that is the
key contributor to air-related exposure. As recognized by the Agency in
setting the standard, and as noted by CASAC in their advice during this
review, these particles include ultra-coarse particles. Thus, choosing
the appropriate indicator requires consideration of the impact of the
indicator on protection from both the inhalation and ingestion pathways
of exposure and Pb in all particle sizes, including ultra-coarse
particles.
As discussed in section V.A., the Agency recognizes the body of
evidence indicating that the high-volume Pb-TSP sampling methodology
contributes to imprecision in resultant Pb measurements due to
variability in the efficiency of capture of particles of different
sizes and thus, in the mass of Pb measured. For example, the measured
values from a high-volume TSP sampler may differ substantially,
depending on wind speed and direction, for the same actual ambient
concentration of Pb-TSP.\131\ Variability is most substantial in
samples with a large portion of Pb particles greater than 10 microns,
such as those samples collected near sources with emissions of ultra-
coarse particles. The result is a clear risk of error from
underestimating the ambient level of total Pb in the air, especially in
areas near sources of ultra-coarse particles, by underestimating the
amount of the ultra-coarse particles. There is also the potential for
overestimation of individual sampling period measurements associated
with high wind events.\132\
---------------------------------------------------------------------------
\131\ As noted in section V, the collection efficiency (over the
24-hour collection period) of particles larger than approximately 10
microns in a high-volume TSP FRM sampler varies with wind speed due
to aerodynamic effects, with a lower collection efficiency under
high winds. The collection efficiency also varies with wind
direction due to the non-cylindrical shape of the TSP sampler inlet.
These characteristics tend in the direction of reporting less than
the true TSP concentration over the 24-hour collection period.
\132\ We note that it is possible for high winds to blow Pb
particles onto a high-volume TSP sampler's filter after the end of
its 24-hour collection period before the filter is retrieved,
causing the reported concentration for the 24-hour period to be
higher than the actual 24-hour concentration.
---------------------------------------------------------------------------
The low-volume PM10 sampling methodology does not
exhibit such variability \133\ due both to increased precision of the
monitor and decreased spatial variation of Pb-PM10
concentrations. As a result, greater precision is associated with
sample measurements for Pb collected using the PM10 sampling
methodology. The result is a lower risk of error in measuring the
ambient Pb in the PM10 size class than there is risk of
error in measuring the ambient Pb in the TSP size class using Pb TSP
samplers. On the other hand, PM10 samplers do not include
the Pb in particles greater than PM10 that also contributes
to the health risks posed by air-related Pb, especially in areas
influenced by sources of ultra-coarse particles. There are also
concerns over whether control strategies put in place to meet a NAAQS
with a Pb-PM10 indicator will be effective in controlling
ultra-coarse Pb-containing particles. In evaluating these two
indicators, the differences in the nature and degree of these sources
of error between Pb-TSP and Pb-PM10 need to be considered
and weighed, to determine the appropriate way to protect the public
from exposure to air-related Pb.
---------------------------------------------------------------------------
\133\ Low-volume PM10 samplers are equipped with an
omni-directional (cylindrical) inlet, which reduces the effect of
wind direction, and a sharp particle separator which excludes most
of the particles greater than 10-15 microns in diameter whose
collection efficiency is most sensitive to wind speed. Also, in low-
volume samplers, the filter is protected from post-sampling
contamination.
---------------------------------------------------------------------------
As noted above, EPA is concerned about the total mass of all Pb
particles emitted into the air and subsequently inhaled or ingested.
Measurements of Pb-TSP address a greater fraction of the particles of
concern from a public health perspective than measurements of Pb-
PM10, but limitations with regard to the sampler mean that
these data are less precise. EPA recognizes substantial variability in
the high-volume Pb-TSP method, meaning there is a risk of not
consistently identifying sites that fail to achieve the standard, both
across sites and across time periods for the same site.
Alternatively, using low-volume Pb-PM10 as the indicator
would allow the use of a technology that has better precision in
measuring PM10. In addition, since Pb-PM10
concentrations have less spatial variability, such monitoring data may
be representative of Pb-PM10 air quality conditions over a
larger geographic area (and larger populations) than would Pb-TSP
measurements. The larger scale of representation for Pb-PM10
would mean that reported measurements of this indicator, and hence
designation outcomes, would be less sensitive to exact monitor siting
than with Pb-TSP as the indicator.\134\ However, there would be a
different source of error, in that larger Pb particles not captured by
PM10 samplers would not be measured.
[[Page 29232]]
The fraction of Pb collected with a TSP sampler that would not be
collected by a PM10 sampler varies depending on proximity to
sources of ultra-coarse Pb particles and the size mix of the particles
they emit (as well as the sampling variability inherent in the method
discussed above). This means that this error is of most concern in
locations in closer proximity to such sources, which may also be
locations with some of the higher ambient air levels. As discussed
below, such variability would be a consideration in determining the
appropriate level for a standard based on a Pb-PM10
indicator.
---------------------------------------------------------------------------
\134\ The larger scale would also make comparisons between two
or more monitoring sites more indicative of the true comparison
between the areas surrounding the monitoring sites, with regard to
the Pb captured by Pb-PM10 monitors, which could be
informative in studies of Pb uptake and health effects in
populations.
---------------------------------------------------------------------------
Accordingly, we believe it is reasonable to consider continued use
of a Pb-TSP indicator, focusing on the fact that it specifically
includes the ultra-coarse Pb particles in the air that are of concern
and need to be addressed in protecting public health from air-related
exposures. In considering the option of retaining Pb-TSP as the
indicator, EPA recognizes that high-volume FRM TSP samplers would
continue to be used at many monitoring sites operated by State and
local agencies. In addition, it is possible that one or more low-volume
TSP monitors would be approved as FEM, under the provisions of 40 CFR
53, Ambient Air Monitoring Reference and Equivalent Methods. EPA
believes, along with some commenters as noted above, that low-volume
Pb-TSP sampling would have important advantages over high-volume Pb-TSP
sampling.\135\ To facilitate the ability of monitor vendors and
monitoring agencies to gain FEM status for low-volume Pb-TSP monitors,
EPA is proposing certain revisions to the side-by-side equivalence
testing requirements in 40 CFR 53 regarding the ambient Pb
concentrations required during testing so that testing is more
practical for a monitor vendor to conduct, as described in more detail
in section V below. We note that 40 CFR 53.7, Testing of Methods at the
Initiative of the Administrator, allows EPA itself to conduct the
required equivalence testing for a method and then determine whether
the requirements for equivalence are met. It would also be possible for
EPA to promulgate amendments to 40 CFR 50 establishing one or more
particular designs of a low-volume sampler as a Pb-TSP FRM, or to
establish performance specifications that would facilitate the approval
of low-volume samplers as FRM on a performance basis rather than a
design basis; this could be done as a replacement for the high-volume
TSP and Pb-TSP FRM or as an alternative TSP and/or Pb-TSP FRM. Either
path to FRM status would avoid the need for the side-by-side testing,
prescribed by 40 CFR 53, of low-volume samplers to demonstrate
equivalence to the high-volume FRM sampler, although some amount and
type of new testing in the field or in a wind tunnel may be appropriate
before such changes should be made. EPA invites comments on the low-
volume TSP sampler concept.
---------------------------------------------------------------------------
\135\ Low-volume Pb-TSP samplers could be assembled by making
low-cost parts substitution to either low-volume PM10 or
low-volume PM2.5 samplers; some models would have the
same sequential sampling ability as CASAC has noted for low-volume
Pb-PM10 samplers; sensitivity to wind direction would be
eliminated; and their flow control and data processing and reporting
abilities would be substantially better than high-volume Pb-TSP
samplers. Low-volume Pb-TSP sampling data would have the same
geographic variability as high-volume Pb-TSP sampling data, however.
The size-specific capture efficiency curves of currently available
commercial low-volume sampling systems are not well characterized,
nor their sensitivity to wind speed. EPA therefore recognizes some
uncertainty about their equivalence to high-volume samplers in terms
of the capture of ultra-coarse particles.
---------------------------------------------------------------------------
Within the option of continued use of a Pb-TSP indicator, EPA
recognizes that some State, local, or tribal monitoring agencies, or
other organizations, for the sake of the advantages noted above, may
wish to deploy low-volume Pb-PM10 samplers rather than Pb-
TSP samplers. In anticipation of this, we have also considered an
approach within the option of retaining Pb-TSP as the indicator that
would allow the use of Pb-PM10 data (when and if low-volume
Pb-PM10 samplers have been approved by EPA as either FRM or
FEM), with adjustment(s), for monitoring for compliance with the Pb-TSP
NAAQS. This approach would have five components: (1) The establishment
of a FRM specification for low-volume Pb-PM10 monitoring
including both a PM10 sampler specification and a reference
chemical analysis method for determination of Pb in the collected
particulate matter; (2) the establishment of a path to FEM designation
for Pb-PM10 monitoring methods that differ from the FRM in
either the sampler or the analytical method; (3) flexibility for
monitoring agencies to deploy low-volume Pb-PM10 monitors
anywhere that Pb monitoring is required by the revised Pb monitoring
requirements to help implement the revised NAAQS; (4) specific steps
for applying an adjustment to low-volume Pb-PM10 data for
purposes of making comparisons to the level of the NAAQS specified in
terms of Pb-TSP, and (5) a provision in the data interpretation
guidelines that, whenever and wherever Pb-TSP data from a monitoring
site is available and sufficient for determining whether or not the Pb-
TSP standard has been exceeded, any collocated Pb-PM10 data
from that site for the associated time period will not be considered.
The first three and the last components are discussed in depth in
sections IV and V below. Because the issue of adjustment to low-volume
Pb-PM10 data is linked closely to considerations of the
advantages of one indicator option versus another, it is discussed
here.
In considering how to identify the appropriate adjustment(s) to be
made to Pb-PM10 data for purposes of making comparisons to
the level of the NAAQS specified in terms of Pb-TSP, we recognize the
importance to protecting public health of taking into account the
ultra-coarse particles that are not included in Pb-PM10
measurement. As discussed below, one approach to doing so would be to
adjust or scale Pb-PM10 data upwards before comparison to a
Pb-TSP NAAQS level where the data are collected in an area that can be
expected to have ultra-coarse particles present.
Pb-PM10/Pb-TSP relationships vary from site to site and
time to time. These Pb-PM10/Pb-TSP relationships have a
systematic variation with distance from emissions sources emitting
particles larger than would be captured by Pb-PM10 samplers,
such that generally there are larger differences between Pb-
PM10 and Pb-TSP near sources. This is due to the faster
deposition of the ultra-coarse particles (as described in section
II.A.1). The exact size mix of particles at the point(s) of emissions
release and the height of the release point(s) also affect the
relationship. Accordingly, EPA is proposing to require the one-time
development and the continued use of site-specific adjustments for Pb-
PM10 data, for those sites for which a State prefers to
conduct Pb-PM10 monitoring rather than Pb-TSP monitoring.
Site-specific studies to establish the relationships between Pb-TSP and
Pb-PM10, conducted using side-by-side paired samplers, would
allow Pb-PM10 monitoring using locally determined factors
based on local study data to determine compliance with a NAAQS based on
Pb-TSP.
In addition, EPA invites comment on also providing in the final
rule default scaling factor(s) for use of Pb-PM10 data in conjunction
with a Pb-TSP indicator, as an alternative for States which wish to
conduct Pb-PM10 monitoring rather than Pb-TSP monitoring near Pb
sources but prefer not to conduct a site-specific scaling factor study.
EPA has identified and analyzed available collocated Pb-PM10 and Pb-TSP
data from 23 monitoring sites in seven States. (Schmidt and Cavender,
2008). This analysis considered both source-
[[Continued on page 29233]]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
]
[[pp. 29233-29282]] National Ambient Air Quality Standards for Lead
[[Continued from page 29232]]
[[Page 29233]]
oriented and nonsource-oriented sites. In this analysis, EPA identified
only three of the 23 monitoring sites with collocated data as being
source-oriented. One of these sites was near an operating Pb smelter at
the time of the collocated monitoring; Pb emissions from smelters
typically contain both ultra-coarse particles from materials handling
and resuspension of contaminated dust, and fine and coarse particles
from the high temperature smelting operation itself. However, since
this study was conducted, EPA has promulgated a Maximum Achievable
Control Technology (MACT) standard for primary lead smelting that
controls process and fugitive dust emissions. (64 FR 30194, June 4,
1999). The other two source-oriented sites include one located near a
battery manufacturer, and one located near an automobile plant. The
data for the smelter site was collected in 1988 and indicate an average
Pb-TSP concentration of about 2.5 [mu]g/m\3\. The data for the battery
manufacturer site were collected in the mid-1990s and indicate an
average Pb-TSP concentration of about 0.09 [mu]g/m\3\; data for the
third site, located near an automotive plant, collected within the past
5 years, indicate an average Pb-TSP concentration at that site of about
0.03 [mu]g/m\3\. As discussed in Schmidt and Cavender (2008), ratios
between Pb-TSP and Pb-PM10 concentrations varied somewhat
within the data for each site, but the ratios between the Pb-TSP and
Pb-PM10 concentration averages were 2.0 for the smelter site
(based on 20 data pairs), 1.6 at the site near the battery manufacturer
(based on 107 data pairs), and 1.1 at the site located near an
automotive plant (based on 167 data pairs).
Collectively, these three monitoring sites suggest that site-
specific scaling factors for source-oriented monitoring sites may vary
between 1.1 and 2.0; the range may also be greater. EPA notes that in
selecting a default factor for source-oriented monitoring sites, if
that approach is taken in the final rule, it may be appropriate to
consider default adjustment factors from within the mid to upper part
of this range rather than the lower end to avoid the possibility of
underestimating the appropriate scaling factor for a large proportion
of the source-oriented sites for which States might choose the default
factor rather than conduct a local study. On this basis, EPA invites
comment on the possibility of providing a default factor(s) for source-
oriented sites and on the selection of a value(s) from within this
range for all source-oriented monitoring sites, as an option to the
proposed requirement for development a site-specific factor through
analysis of paired monitoring data. EPA invites comment on the
selection of a single or multiple default factors for source-oriented
sites from within this range. While the selection of the scaling factor
in concept could depend on a characterization of the particle size mix
emitted by the Pb source, we note that reliable information on the mix
of coarse and ultra-coarse particles may often be unavailable. For
example, EPA could select a default factor that is at or near the upper
end of the range, 2.0, to avoid the risk of underprotection in
situations in which there is as high or nearly as high a proportion of
ultra-coarse Pb as at the smelter site. Alternatively, EPA could
discount the smelter data set on the basis that the 1988 data set does
not reasonably represent any likely current or future smelter
situation. Similarly, EPA could rely on the data taken near the
automotive plant since it is the most recent and largest dataset. EPA
also invites comment on other sets of paired data from near Pb sources
of which we may be unaware, and comment on other approaches of
selecting a default factor for the final rule based on paired data,
including approaches that might use more than one default factor for
source-oriented monitoring sites with the selection of the factor for a
given monitoring site depending on the characteristics of the nearby
sources, the ambient concentration of Pb-PM10, or other
factors.
EPA also invites comment on whether and what default scaling
factor(s) should be established for monitoring sites which, as far as
is known, are not influenced by nearby emission sources. We have
reviewed paired data from the 20 monitoring sites that appear to fit
this description (Schmidt and Cavender, 2008). Average Pb-TSP
concentrations at nearly all these sites were near to or below the
lowest concentration on which comments are invited as to the NAAQS
level. Judging from ratios at these 20 sites, it appears that site-
specific factors generally range from 1.0 to 1.4 (with the factors for
three sites ranging from 1.8 to 1.9), and the ratios may be influenced
by measurement variability in both samplers as well as by actual air
concentrations. Given the relatively low ambient concentrations that we
believe currently prevail at nonsource-oriented sites, the value of a
default scaling factor selected within the range of 1.0 to 1.4 would
have little effect on the NAAQS compliance determination at such sites.
EPA invites comment on the approach of requiring use of a default
factor(s) for adjusting Pb-PM10 data at nonsource-oriented
sites and on the selection of a value(s) from within the range of 1.0
to 1.4 and also solicits comment on selection of a default scaling
factor from within the broader range of 1.0 to 1.9. We note that
allowing the use of a default scaling factor of 1.0 for nonsource-
oriented sites would in effect allow a State the option of comparing
Pb-PM10 data directly to the level of the Pb-TSP standard at
nonsource-oriented monitoring sites, without conducting a site-specific
study. Below, and in section II.E.4, EPA discusses the possibility of
revising the indicator to Pb-PM10, which would result in
such unadjusted comparisons of Pb-PM10 data to the standard
at all monitoring sites.
EPA recognizes that the available data from collocated monitoring
of Pb-TSP and Pb-PM10, described above, have limitations
which make their interpretation and use in selecting default scaling
factors subject to considerable uncertainty. All of the Pb-
PM10 measurements at these sites were made with high-volume
PM10 samplers, which are more variable than the low-volume
samplers for which scaling factors would actually be applied after the
final rule; this greater variability no doubt has added to the
variation in ratios discussed above. Only three source-oriented sites
have collocated data; with such a small sample of sites both the range
of ratios and the distribution of ratios among all current and future
source-oriented sites remains uncertain. There were many more
nonsource-oriented sites which tended to show notably lower ratios,
implying lower scaling factors, but all had relatively low
concentrations; these ratios may or may not be representative of
monitoring sites near well controlled Pb sources. In many cases, the
period of collocated testing was only a few months; ratios observed in
such a short period may not be representative of ratios that occur at
other times of the year that may be more critical to attainment status.
Also, EPA has not yet had the benefit of CASAC review of the detailed
compilation of these data, as (Schmidt and Cavender, 2008) was prepared
subsequent to the most recent consultation with CASAC's AAMM
Subcommittee. Because of these uncertainties, EPA is proposing to
require States that wish to use Pb-PM10 data for a Pb-TSP
standard to develop site-specific scaling factors based on their own
collocated monitoring using paired Pb-TSP and low-volume Pb-
PM10 samplers over at least a one-year period, as described
in section IV. EPA intends
[[Page 29234]]
to encourage States to consider conducting local studies, even if the
final rule allows the use of default factors. Also, EPA invites comment
on whether to provide for the use of default scaling factors, and the
values of those factors.
As a possible second option, taking into consideration the advice
of the CASAC Pb Panel and members of the CASAC AAMM Subcommittee, EPA
has also considered potential revision of the indicator to Pb-
PM10. In so doing, we recognize several potential important
benefits of such a revision, as well as the need to reflect such a
revision in the selection of level of the standard.\136\ We recognize
that the low volume PM10 sampler provides better precision
and size selection characteristics which would make the associated data
more comparable across sites.
---------------------------------------------------------------------------
\136\ EPA recognizes and has specifically considered that such a
decision would affect the selection of the level of the standard,
recognizing that it is the combination of indicator and level (with
averaging and time and form) that determine the degree of protection
afforded by the standard. Section II.E.4 further considers the
impact of adoption of a Pb-PM10 indicator on the
selection of a level for the standard.
---------------------------------------------------------------------------
In considering a potential revision of the indicator to Pb-
PM10, we recognize that an important issue is whether
regulating concentrations of Pb-PM10 will lead to
appropriate controls on all particle size Pb emissions from sources.
For example, it would be of concern if a NAAQS based on a Pb-
PM10 indicator resulted in different emissions control
decisions at sources with a large percentage of Pb in the size range
not substantially captured by PM10 sampling (e.g., fugitive
dust emissions from Pb smelters) than the emission control decisions
that would be made if the NAAQS was based on Pb-TSP. In that case, a
PM10-based NAAQS might not yield emissions changes by some
Pb sources which under a Pb-TSP indicator would have contributed to
NAAQS exceedances and subsequent emissions changes. Alternatively,
while collocated Pb-TSP and Pb-PM10 data are lacking for a
broad range of source types, there are likely many sources (e.g., high
temperature combustion processes) for which virtually all of the
emitted particles represented in a Pb-TSP measurement would be captured
by a Pb-PM10 measurement. Further, there are likely other
source types with a range of particle sizes extending beyond Pb-
PM10, for which controls adopted to meet a Pb-
PM10 requirement would also achieve a proportional reduction
in ultra-coarse particles. In these situations, one might not expect
any difference in emissions control decisions whether the NAAQS is Pb-
PM10-based or Pb-TSP-based.
If the indicator were to be revised to Pb-PM10, low-
volume Pb-PM10 samplers would become the required approach
to Pb monitoring at required monitoring sites and would be a logical
choice wherever else NAAQS-oriented Pb monitoring is undertaken.
Nonetheless EPA notes that retaining Pb-TSP monitors at some relatively
small subset of the Pb-PM10 monitoring sites would be
beneficial for purposes of scientific understanding of both ambient
conditions and the performance of the two types of measurement systems.
For reasons discussed here, and taking into account information and
assessments presented in the Criteria Document, Staff Paper, and ANPR,
the advice and recommendations of CASAC and of members of the CASAC
AAMM Subcommittee, and public comments to date, the Administrator
proposes to retain the current indicator of Pb-TSP, measured by the
current FRM, a current FEM, or an FEM approved under the proposed
revisions to 40 CFR part 53, but with expansion of the measurements
accepted for determining attainment or nonattainment of the Pb NAAQS to
provide an allowance for use of Pb-PM10 data, measured by
the new low-volume Pb-PM10 FRM specified in the proposed
appendix Q to 40 CFR part 50 or by a FEM approved under the proposed
revisions to 40 CFR part 53, with site-specific scaling factors as
described above and more specifically below in section IV. The
Administrator invites comment on also providing States the option of
using default scaling factors instead of conducting the testing that
would be needed to develop the site-specific scaling factors. In
consideration of all of the issues discussed above, the Administrator
also invites comment on a second option, a revision of the current
indicator to Pb-PM10. (Considerations related to the level
of a standard based on a PM10 indicator are discussed below
in section II.E.4.) The Administrator solicits comment on all of the
issues discussed above, and specifically with regard to the potential
for a Pb-PM10 indicator to influence implementation of
controls in ways that would lead to less control associated with larger
particles than might be achieved with a Pb-TSP-based NAAQS, taking into
account the variability noted above for TSP sampling.
2. Averaging Time and Form
The statistical form of the current standard is a not-to-be-
exceeded or maximum value, averaged over a calendar quarter. This might
also be described as requiring that no average air Pb concentration
representing a time period of duration as long as calendar quarter (or
longer) may exceed the level of the standard. As noted in section
II.D.1.a, EPA set the standard in 1978 as a ceiling value with the
conclusion that this air level would be safe for indefinite exposure
for young children (43 FR 46250).
The basis for selection of the current standard's averaging time of
calendar quarter reflects consideration of the evidence available when
the Pb NAAQS were promulgated in 1978. At that time, the Agency had
concluded that the level of the standard, 1.5 [mu]g/m\3\, would be a
``safe ceiling for indefinite exposure of young children'' (43 FR
46250), and that the slightly greater possibility of elevated air Pb
levels for shorter periods within the quarterly averaging period as
contrasted to the monthly averaging period proposed in 1977 (43 FR
63076), was not significant for health. These conclusions were based in
part on the Agency's interpretation of the health effects evidence as
indicating that 30 [mu]g/dL was the maximum safe level of blood Pb for
an individual child.
With regard to averaging time, after consideration of the evidence
available at that time, the 1990 Staff Paper concluded that ``[a]
monthly averaging period would better capture short-term increases in
lead exposure and would more fully protect children's health than the
current quarterly average'' (USEPA, 1990b). The 1990 Staff Paper
further concluded that ``[t]he most appropriate form of the standard
appears to be the second highest monthly average in a 3-year span. This
form would be nearly as stringent as a form that does not permit any
exceedances and allows for discounting of one `bad' month in 3 years
which may be caused, for example, by unusual meteorology.'' In their
review of the 1990 Staff Paper, the CASAC Pb Panel concurred with the
staff recommendation to express the lead NAAQS as a monthly standard
not to be exceeded more than once in three years.
As summarized in section II.B above and discussed in detail in the
Criteria Document, the currently available health effects evidence
\137\ indicates a wider variety of neurological effects, as well as
immune system and hematological effects, associated with substantially
lower blood Pb levels in children than were recognized when the
standard was set in 1978. Further, the health effects evidence with
regard to characterization of a threshold for
[[Page 29235]]
adverse effects has changed since the standard was set in 1978, as have
the Agency's views on the characterization of a safe blood Pb
level.\138\ In consideration of averaging time for the Pb NAAQS, we
note the following aspects of the current health effects evidence.
---------------------------------------------------------------------------
\137\ The differing evidence and associated strength of the
evidence for these different effects is described in detail in the
Criteria Document.
\138\ For example, EPA recognizes today that ``there is no level
of Pb exposure that can yet be identified, with confidence, as
clearly not being associated with some risk of deleterious health
effects'' (CD, p. 8-63).
---------------------------------------------------------------------------
Children are exposed to ambient Pb via inhalation and
ingestion, with Pb that is taken into the body absorbed through the
lungs and through the gastrointestinal tract. Studies on Pb uptake,
elimination, and distribution show that Pb is absorbed into peripheral
tissues in adults within a few days (USEPA 1986a; USEPA 1990b, p. IV-
2). Absorption of Pb from the gastrointestinal tract appears to be
greater and faster in children as compared to adults (CD, Section
4.2.1). Once absorbed, it is quickly distributed from plasma to red
blood cells and throughout the body.
Lead accumulates in the body and is only slowly removed,
with bone Pb serving as a blood Pb source for years after exposure and
as a source of fetal Pb exposure during pregnancy (CD, Sections 4.3.1.4
and 4.3.1.5).
Blood Pb levels, including levels of the toxicologically
active fraction, respond quickly to increased Pb exposure, such that an
abrupt increase in Pb uptake rapidly changes blood Pb levels. The
associated time to reach a new quasi-steady state with the total body
burden after such an occurrence is projected to be approximately 75 to
100 days (CD, p. 4-27).
The elimination half-life, which describes the time for
blood Pb levels to stabilize after a reduction in exposure, for the
dominant phase for blood Pb responses to changes in exposure is on the
order of 20 to 30 days for adults (CD, p. 4-25). Blood elimination
half-lives are influenced by contributions from bone. Given the tighter
coupling in children of bone stores with blood levels, children's blood
Pb is expected to respond more quickly than adults (CD, pp. 4-20 and 4-
27).
Data from NHANES II and an analysis of the temporal
relationship between gasoline consumption data and blood lead data
generally support the inference of a prompt response of children's
blood Pb levels to changes in exposure. Children's blood Pb levels and
the number of children with elevated blood Pb levels appear to respond
to monthly variations in Pb emissions from Pb in gasoline (EPA, 1986a,
p. 11-39; Rabinowitz and Needleman, 1983; Schwartz and Pitcher, 1989;
USEPA, 1990b).
The evidence with regard to sensitive neurological effects
is limited in what it indicates regarding the specific duration of
exposure associated with effect, although it indicates both the
sensitivity of the first 3 years of life and a sustained sensitivity
throughout the lifespan as the human central nervous system continues
to mature and be vulnerable to neurotoxicants (CD, Section 8.4.2.7).
The animal evidence supports our understanding of periods of
development with increased vulnerability to specific types of effect
(CD, Section 5.3), and indicates a potential importance of exposures on
the order of months.
Evidence of a differing sensitivity of the immune system
to Pb across and within different periods of life stages indicates a
potential importance of exposures as short as weeks to months duration.
For example, the animal evidence suggests that the gestation period is
the most sensitive life stage followed by early neonatal stage, and
within these life stages, critical windows of vulnerability are likely
to exist (CD, Section 5.9 and p. 5-245).
Evidence described in the Criteria Document and the risk assessment
indicate that ingestion of dust can be a predominant exposure pathway
for young children to air-related Pb, and that there is a strong
association between indoor dust Pb levels and children's blood Pb
levels. As stated in the Criteria Document, ``given the large amount of
time people spend indoors, exposure to Pb in dusts and indoor air can
be significant'' (CD, p. 3-27). The Criteria Document further describes
studies that evaluated the influence of dust Pb exposure on children's
blood Pb: ``Using a structural equation model, Lanphear and Roghmann
(1997) also found the exposure pathway most influential on blood Pb was
interior dust Pb loading, directly or through its influence on hand Pb.
Both soil and paint Pb influenced interior dust Pb; with the influence
of paint Pb greater than that of soil Pb. Interior dust Pb loading also
showed the strongest influence on blood Pb in a pooled multivariate
regression analysis (Lanphear et al., 1998).'' (CD, p. 4-134). Further,
a recent study of dustfall near an open window in New York City
indicates the potential for a relatively rapid response of indoor dust
Pb loading to ambient airborne Pb, on the order of weeks (CD, p. 3-28;
Caravanos et al., 2006a).
We note that the health effects evidence identifies varying length
durations in exposure that may be relevant and important. In light of
uncertainties in aspects such as response times of children's exposure
to airborne Pb, we recognize, as in the past, that this evidence
provides a basis for consideration of both calendar quarter and
calendar month as averaging times.
In considering averaging time and form, EPA has combined the
current quarterly averaging time with the current not-to-be exceeded
(maximum) form and has also combined a monthly averaging time with a
second maximum form, so as to provide an appropriate degree of year-to-
year stability that a maximum monthly form would not afford. We also
note that, as discussed below, the second maximum monthly form provides
a roughly comparable degree of protection on a broad national scale.
In this consideration of averaging time and form, EPA has taken
into account analyses using air quality data for 2003-2005 that are
presented in the Staff Paper (chapter 2). These analyses consider both
a period of three calendar years and a period of one calendar year
(with the form of the current standard being the maximum quarterly
mean). These analyses indicate that, with regard to either single-year
or 3-year statistics for the 2003-2005 dataset, a second maximum
monthly mean yields very similar, although just slightly greater,
numbers of sites exceeding various alternative levels as a maximum
quarterly mean, with both yielding fewer exceedances than a maximum
monthly mean.\139\ That is, these two averaging time and form
combinations resulted in roughly the same number of areas that would
not attain a standard at any given level on a broad national scale,
suggesting roughly comparable public health protection. However, the
relative protection provided by these two forms may differ from area to
area. For example, some of the areas meeting a maximum quarterly mean
standard over the 2003-2005 period at a given level did not meet a
second maximum monthly mean standard at the same level because there
were at least two months with high monthly concentrations which were
averaged with a lower concentration month in the same quarter. On the
other hand,
[[Page 29236]]
theoretically it is possible for an area to meet a given standard level
with a second maximum monthly mean averaging time and form and not meet
it for a maximum quarterly mean (e.g., the second highest monthly
average may be below the standard level while the quarterly average may
exceed it). Moreover, control programs to reduce quarterly mean
concentrations may not have the same protective effect as control
programs aimed at reducing concentrations in every individual month.
Given the limited scope of the current monitoring network which lacks
monitors near many significant Pb sources and uncertainty about Pb
source emissions and possible controls, it is difficult to more
quantitatively compare the protectiveness of the quarterly mean versus
the second maximum monthly mean approaches.
---------------------------------------------------------------------------
\139\ For example, 49 sites (of 189) exceed a standard level of
0.10 [mu]g/m\3\ based on a form of maximum quarterly mean while 54
sites exceed based on a form of second maximum monthly mean.
Further, 25 sites exceed a standard level of 0.30 [mu]g/m\3\ based
on a form of maximum quarterly mean while 29 sites exceed based on a
form of second maximum monthly mean (Staff Paper, Table 2-6).
---------------------------------------------------------------------------
In their advice to the Agency in this review, CASAC has recommended
that consideration be given to changing from a calendar quarter to a
monthly averaging time (Henderson, 2007a, 2007b, 2008). In making that
recommendation, CASAC emphasizes support from studies that suggest that
blood Pb concentrations respond at shorter time scales than would be
captured completely by quarterly values, as indicated by their
description of their recommendation for adoption of a monthly averaging
time as ``more protective of human health in light of the response of
blood lead concentrations that occur at sub-quarterly time scales''
(Henderson, 2007a). With regard to form of the standard, CASAC has
stated that one could ``consider having the lead standards based on the
second highest monthly average, a form that appears to correlate well
with using the maximum quarterly value'', while also indicating that
``the most protective form would be the highest monthly average in a
year'' (Henderson, 2007a).
Among the public comments the Agency received on the discussion of
averaging time and form in the ANPR, the majority concurred with the
CASAC recommendation for a revision of the averaging time to a calendar
month.
The 1990 Staff Paper and the Staff Paper for this review both
recommended that the Administrator consider specifying, in the form of
the NAAQS, that compliance with the NAAQS will be evaluated over a 3-
year period. The Administrator has considered this recommendation and
is proposing to adopt it. In the 3-year approach, a monitor would be
considered to be in violation of the NAAQS as of a certain date if in
any of the three previous calendar years with sufficiently complete
data (as explained in detail in section IV below), the value of the
selected form of the indicator (e.g., second maximum monthly average or
maximum quarterly average) exceeded the level of the NAAQS. A monitor,
initially or after once having violated the NAAQS, would not be
considered to have attained the NAAQS until three years have passed
without the form and level of the standard being violated. Many types
of Pb sources have variable emissions from day-to-day and year-to-year
due to market conditions for their products and/or weather variations
that can affect the generation of fugitive dust from contaminated
roadways and grounds. In addition, variations in wind patterns from
year to year can cause a near-source Pb monitor to be exposed to high
concentrations on more days in one year than in another, even if source
emissions are constant, especially if it operates on only some days.
Thus, it is possible for a monitor to indicate a violation of a
hypothetical form and level in one period but not in another, even if
no permanent controls have been applied at nearby source(s). Analysis
of historical Pb air concentration data has confirmed that this pattern
of fluctuating monitoring results can happen at the levels and forms
being proposed. It would potentially reduce the public health
protection afforded by the standard if areas fluctuated in and out of
formal nonattainment status so frequently that states do not have
opportunity and incentive to identify sources in need of more emission
control and to require those controls to be put in place. The 3-year
approach would help ensure that areas initially found to be violating
the NAAQS have effectively controlled the contributing lead emissions
before being redesignated to attainment/maintenance.
In considering averaging time and form for the standard, the
Administrator has considered the information summarized above
(described in more detail in Criteria Document and Staff Paper), as
well as the advice from CASAC and public comments. The Administrator
recognizes that there is support in the evidence for a monthly
averaging time consistent with the following observations: (1) The
health evidence indicates that very short exposures can lead to
increases in blood Pb levels, (2) the time period of response of indoor
dust Pb to airborne Pb can be on the order of weeks, and (3) the health
evidence indicates that adverse effects may occur with exposures during
relatively short windows of susceptibility, such as prenatally and in
developing infants.\140\ The Administrator also recognizes limitations
and uncertainties in the evidence including the limited available
evidence specific to the consideration of the particular duration of
sustained airborne Pb levels having the potential to contribute to the
adverse health effects identified as most relevant to this review, as
well as variability in the response time of indoor dust Pb loading to
ambient airborne Pb.
---------------------------------------------------------------------------
\140\ The health evidence with regard to the susceptibility of
the developing fetus and infants is well documented in the evidence
as described in the 1986 Criteria Document, the 1990 Supplement
(e.g., chapter III) and the 2006 Criteria Document. For example,
``[n]eurobehavioral Neurobehavioral effects of Pb-exposure early in
development (during fetal, neonatal, and later postnatal periods) in
young infants and children (7 years old) have been observed
with remarkable consistency across numerous studies involving
varying study designs, different developmental assessment protocols,
and diverse populations.'' (CD, p. E-9)
---------------------------------------------------------------------------
Based on these considerations and the air quality analyses
summarized above, the Administrator concludes that this information
provides support for an averaging time no longer than a calendar
quarter. Further, the Administrator recognizes that if substantial
weight is given to the evidence of even shorter times for response of
dust Pb, blood Pb, and associated effects to airborne Pb, a monthly
averaging time may be appropriate. Accordingly, the Administrator is
proposing two options with regard to the form and averaging time for
the standard, and with both he proposes making the time period
evaluated in considering attainment be 3 years. One option is to retain
the current not-to-be-exceeded form with an averaging time of a
calendar quarter, such that the form would be maximum quarterly average
across a 3-year span. The second option is to revise the averaging time
to a calendar month and the form to be the second highest monthly
average across a 3-year span. Based on the considerations discussed
above, EPA requests comment on whether a level for a NAAQS with a
monthly averaging time and a second-highest monthly average form should
be based on an adjustment to a higher level than the level for a NAAQS
with a quarterly averaging time and a not-to-be-exceeded form, and, if
so, on the magnitude of the adjustment that would be appropriate.
3. Level for a Pb NAAQS With a Pb-TSP Indicator
With regard to level of the standard, for a standard using a Pb-TSP
indicator, we first discuss evidence-based and exposure/risk-based
considerations, including considerations and
[[Page 29237]]
conclusions of the Staff Paper, in sections II.E.3.a and II.E.3.b
below. This is followed by a summary of CASAC advice and
recommendations and public comments (section II.E.3.c) and the
Administrator's proposed conclusions (section II.E.3.d). In addition,
we discuss considerations and solicit comment with regard to a level of
a standard using a Pb-PM10 indicator in section II.E.4
below.
a. Evidence-Based Considerations
As a general matter, EPA recognizes that in the case of Pb there
are several aspects to the body of epidemiological evidence that add
complexity to the selection of an appropriate level for the primary
standard. As summarized above and discussed in greater depth in the
Criteria Document (CD, Sections 4.3 and 6.1.3), the epidemiological
evidence that associates Pb exposures with health effects generally
focuses on blood Pb for the dose metric.\141\ In addition, exposure to
Pb comes from various media, only some of which are air-related. This
presents a more complex situation than does evidence of associations
between occurrences of health effects and ambient air concentrations of
an air pollutant, such as is the case for particulate matter and ozone.
Further, for the health effects receiving greatest emphasis in this
review (neurological effects, particularly neurocognitive and
neurobehavioral effects, in children), no threshold levels can be
discerned from the evidence. As was recognized at the time of the last
review, estimating a threshold for toxic effects of Pb on the central
nervous system entails a number of difficulties (CD, pp. 6-10 to 6-11).
The task is made still more complex by support in the evidence for a
nonlinear rather than linear relationship of blood Pb with
neurocognitive decrement, with greater risk of decrement-associated
changes in blood Pb at the lower levels of blood Pb in the exposed
population (Section 3.3.7; CD, Section 6.2.13). In this context EPA
notes that the health effects evidence most useful in determining the
appropriate level of the NAAQS is this large body of epidemiological
studies. Unlike the recent review of the NAAQS for ozone, there are no
clinical studies useful for informing a determination of the
appropriate level for a standard.\142\ The discussion below therefore
focuses on the epidemiological studies, recognizing and taking into
consideration the complexity and resulting uncertainty in using this
body of evidence to determine the appropriate level for the NAAQS.
---------------------------------------------------------------------------
\141\ Among the studies of Pb health effects, in which blood Pb
level is generally used as an index of exposure, the sources of
exposure vary and are inclusive of air-related sources of Pb such as
smelters (e.g., CD, chapter 6).
\142\ See, e.g., 72 FR 37878-9 (July 11, 2007) (Ozone NAAQS
Notice of Proposed Rulemaking).
---------------------------------------------------------------------------
In considering the evidence with regard to selection of the level
of the standard, the Agency has considered the same evidence-based
frameworks discussed above in section II.D.2.a on the adequacy of the
current standard. That is, the Staff Paper considered how to apply an
adapted 1978 framework to the much expanded body of evidence that is
now available, and the Agency has further considered this evidence in
the context of the air-related IQ loss evidence-based framework that
builds on a recommendation by the CASAC Pb Panel. These evidence-based
approaches are discussed below in considering the appropriate standard
levels to propose.
As noted in section II.D.2.a above, this review focuses on young
children as a key sensitive population for Pb exposures. In this
sensitive population, the current evidence demonstrates the occurrence
of health effects, including neurological effects, associated with
blood Pb levels extending well below 10 [mu]g/dL (CD, sections 6.2, 8.4
and 8.5). As further described in section II.D.2.a above, some studies
indicate Pb effects on intellectual attainment of children for which
population mean blood Pb levels in the analysis ranged from
approximately 2 to 8 [mu]g/dL (CD, Sections 6.2, 8.4.2 and 8.4.2.6).
Further, as noted above, the current evidence does not indicate a
threshold for the more sensitive health endpoints such as neurological
effects in children (CD, pp. 5-71 to 5-74 and Section 6.2.13).\143\
---------------------------------------------------------------------------
\143\ This differs from the Agency's recognition in the 1978
rulemaking of a threshold of 40 [mu]g/dL blood Pb for an individual
child for effects of Pb considered clearly adverse to health at that
time, i.e., impairment of heme synthesis and other effects which
result in anemia.
---------------------------------------------------------------------------
As when the standard was set in 1978, there remain today
contributions to blood Pb levels from nonair sources. As discussed
above (section II.D.2), current evidence is limited with regard to
estimates of the aggregate reduction since 1978 of all nonair sources
to blood Pb and with regard to an estimate of current nonair blood Pb
levels (discussed more fully in sections II.A.4) In recognition of
temporal reductions in nonair sources discussed in section II.A.4 and
in the context of estimates pertinent to an application of the 1978
framework, the CASAC Pb Panel recommended consideration of 1.0 to 1.4
[mu]g/dL or lower as an estimate of the nonair component of blood Pb
pertinent to average blood Pb levels in children (as described in
section II.A.4 above; Henderson, 2007a). The Staff Paper considered
this range of 1.0 to 1.4 [mu]g/dL for the nonair component of blood Pb
in its application of the adapted 1978 evidence-based framework.
As discussed in section II.B.1.c, the current evidence in
conjunction with the results and observations drawn from the exposure
assessment support a focus on air-to-blood ratios for children in the
range of 1:3 to 1:7, based on consideration of both inhalation and
ingestion exposure pathways and on the lower air and blood Pb levels
pertinent to this review. In considerations here, we have described the
value of 1:5 as falling somewhat central within the range supported by
the evidence.
i. Evidence-Based Framework Considered in the Staff Paper
Recommendations in the Staff Paper on standard levels were based
upon an approach that built upon and adapted the general approach used
by EPA in setting the standard in 1978. In adapting this approach to
the currently available information, the Staff Paper recognized the
more extensive and stronger body of evidence now available on a broader
range of health effects associated with exposure to Pb. For example,
EPA recognizes that today ``there is no level of Pb exposure that can
yet be identified, with confidence, as clearly not being associated
with some risk of deleterious health effects'' (CD, p. 8-63). This is
in contrast to the situation in 1978 when the Agency judged that the
maximum safe individual and geometric mean blood Pb levels for a
population of young children were 30 [mu]g/dL and 15 [mu]g/dL,
respectively.\144\
---------------------------------------------------------------------------
\144\ More specifically, when the standard was set in 1978, the
Agency stated that the population mean, measured as the geometric
mean, must be 15 [mu]g/dL in order to ensure that 99.5 percent of
children in the United States would have a blood Pb level below 30
[mu]g/dL, which was identified as the maximum safe blood Pb level
for individual children based on the information available at that
time (43 FR 46247-46252).
---------------------------------------------------------------------------
In the Staff Paper application of an adapted 1978 framework, the
focus shifted away from identifying a safe blood Pb level for an
individual child (and then determining an ambient air level that would
keep a very high percentage of children at or below that safe level),
because information was no longer available to identify such a level.
Rather, the Staff Paper approach focused on identifying an appropriate
population mean blood Pb level, and then identifying an ambient air
level that would keep the mean blood Pb levels of children exposed at
that air level below the target population mean blood Pb level. Based
on the review of
[[Page 29238]]
the evidence, the Staff Paper approach substituted a level of 2 [mu]g/
dL for the target population geometric mean blood Pb of 15 [mu]g/dL
used in 1978. In the absence of a demonstrated safe level, at either an
individual or a population level, the Staff Paper used 2 [mu]g/dL as
representative of the lowest population mean level for which there is
evidence of a statistically significant association between blood lead
levels and health effects (e.g., CD, p. E-9; Lanphear et al., 2000).
This approach does not evaluate the magnitude or degree of health
effects occurring across the population at that mean blood lead level.
In this adaptation of the 1978 approach the focus is solely on the
existence of a relationship between blood lead levels and
neurocognitive effects. The approach takes as the public health goal
the identification of an ambient air lead level that can be expected to
keep the mean blood lead level of an exposed population of children at
or below the lowest level at which a statistically significant
association has been demonstrated between blood lead level and
neurocognitive effects.\145\
---------------------------------------------------------------------------
\145\ There are some similarities between this approach and the
approach employed in determining the levels for the daily and annual
PM standards in the latest PM review, where EPA determined an
ambient PM level based on the ambient levels in the epidemiology
studies that found statistically significant associations between
changes in ambient PM levels and changes in occurrences of health
effects. See 71 FR 61144 (October 17, 2006). However, there are
several important differences in this adaptation to the 1978
approach for lead. For example, the health effects evaluated in the
PM epidemiological studies were clearly adverse health effects,
ranging from hospital admissions to premature mortality. In
addition, the studies looked directly at the association between
ambient level and occurrences of health effects. Here the
epidemiology studies look at the association between blood lead
level and neurocognitive effect, and there is an additional step to
link the blood lead level to air-related lead. In addition, at a
population level there is a less clear delineation of when the
neurocognitive effect is adverse to public health. This is discussed
below in this section with respect to the impact on public health of
a shift in the mean IQ of a population of children.
---------------------------------------------------------------------------
Starting with a target population geometric mean blood lead level
of 2 [mu]g/dL for the population of exposed children, then subtracting
1 to 1.4 [mu]g/dL for the nonair component of blood Pb, yields 0.6 to 1
[mu]g/dL as a target for the geometric mean air contribution to blood
Pb. The adapted 1978 approach divides the air-related target by 5, an
air-to-blood ratio somewhat central within the range of air Pb to blood
Pb ratios generally supported by the currently available evidence. This
resulted in a potential standard level of 0.1 to 0.2 [mu]g/m\3\.
The Staff Paper conclusions on level for the primary Pb standard
built on the staff's conclusion that the overall body of evidence
clearly calls into question the adequacy of the current standard with
regard to health protection afforded to at-risk populations. Based on
consideration of the health effects evidence, as described above, the
Staff Paper concluded that it is reasonable to consider a range for the
level of the standard, for which the upper part is represented by 0.1
to 0.2 [mu]g/m\3\.
ii. Air-related IQ Loss Evidence-Based Framework
As mentioned above, in analyses subsequent to the Staff Paper and
ANPR, the Agency has primarily considered the evidence in the context
of an alternative evidence-based framework, referred to as the air-
related IQ loss framework. This framework focuses on the contribution
of air-related Pb to neurocognitive effects, with a public health goal
of identifying the appropriate ambient air level of Pb to protect
exposed children from health effects that are considered adverse, and
are associated with their exposure to air-related Pb. This framework
does not focus on overall blood lead levels or on nonair contribution
to blood lead levels. While this avoids some of the limitations noted
above with the adapted 1978 approach, EPA recognizes that looking at
air-related Pb in isolation from other sources of Pb could be
considered a limitation for this framework. The different limitations
of each of these frameworks derive from the limitations in the
underlying body of evidence available for this review.
In this air-related IQ loss evidence-based framework, we have drawn
from the entire body of evidence as a basis for concluding that there
are causal associations between air-related Pb exposures and population
IQ loss. We have drawn more quantitatively from the evidence by
combining air-to-blood ratios with evidence-based C-R functions from
the epidemiological studies to quantify the association between air Pb
concentrations and air-related population mean IQ loss in exposed
children. This air-related IQ loss framework focuses on selecting a
standard that would prevent air-related IQ loss (and related effects)
of a magnitude judged by the Administrator to be of concern in
populations of children exposed to the level of the standard, taking
into consideration such factors as the uncertainties inherent in such
estimates. In addition to this judgment by the Administrator, this
framework is also based on specifying an air-to-blood ratio (also used
in the adapted 1978 framework) and a C-R function(s) for population
mean IQ response associated with blood Pb level.
In considering the evidence with regard to C-R functions, and in
recognition of the finding in the evidence of a steeper slope at lower
blood Pb levels (i.e., the nonlinear relationship), we have identified
two sets of C-R functions (discussed more fully above in section
II.B.2.b). The first set focuses on C-R functions reflecting population
mean concurrent blood Pb levels of approximately 3 [mu]g/dL.\146\ The
second set (CD, pp. 8-78 to 8-80) considers functions descriptive of
the C-R relationship from a larger set of studies that include
population mean blood Pb levels ranging from a mean of 3.3 up to a
median of 9.7 [mu]g/dL (see Table 1).\147\
---------------------------------------------------------------------------
\146\ As noted above in section II.B.2.b, the log-linear C-R
function with low-exposure linearization (LLL) used in the
quantitative risk assessment, based on log-linear model in Lanphear
et al 2005), has a slope that falls intermediate within this first
set of functions at low blood Pb levels. The log-linear model by
Lanphear et al (2005) is derived from the pooled International
dataset for which the median blood Pb is 9.7 [mu]g/dL.
\147\ For context, it is noted that the 2001-2004 median blood
level for children aged 1-5 of all races and ethnic groups is 1.6
[mu]g/dL, the median for the subset living below the poverty level
is 2.3 [mu]g/dL and 90th percentile values for these two groups are
4.0 [mu]g/dL and 5.4 [mu]g/dL, respectively. Similarly, the 2001-
2004 median blood level for black, non-hispanic children aged 1-5 is
2.5 [mu]g/dL, while the median level for the subset of that group
living below the poverty level is 2.9 [mu]g/dL and the median level
for the subset living in a household with income more than 200% of
the poverty level is 1.9 [mu]g/dL. Associated 90th percentile values
for 2001-2004 are 6.4 [mu]g/dL (for black, non-hispanic children
aged 1-5), 7.7 [mu]g/dL (for the subset of that group living below
the poverty level) and 4.1 [mu]g/dL (for the subset living in a
household with income more than 200% of the poverty level). (http://
www.epa.gov/envirohealth/children/body_burdens/b1-table.htm--then
click on ``Download a universal spreadsheet file of the Body Burdens
data tables'').
---------------------------------------------------------------------------
As discussed above in section II.B.2.b, the C-R functions from
analyses involving the lower mean blood Pb levels, that are closer to
current mean blood Pb levels in U.S. children, provide slopes of IQ
loss with increasing blood Pb that range from -1.71 to -2.94 IQ points
per [mu]g/dL blood Pb. These include C-R function from Lanphear et al.
(2005) recommended for consideration by CASAC, in light of the current
blood Pb levels of U.S. children (Henderson, 2008),\148\ and also the
C-R function
[[Page 29239]]
given greatest weight in the risk assessment (discussed above in
section II.C.2.b), the loglinear function with low-exposure
linearization (the LLL function). The function yielding the lowest
slope in this range is from the analysis by Tellez-Rojo and others
(2006) of very young children with blood Pb levels below 5 [mu]g/dL,
with a group mean blood Pb level of 2.9 [mu]g/dL. The function yielding
the highest slope in this range is from the analysis by Lanphear and
others (2005) of children whose blood Pb levels never exceeded 7.5
[mu]g/dL, with a group mean blood Pb level of 3.24 [mu]g/dL. The LLL
function falls within the range of the other two functions at lower
blood Pb levels, with an average slope of -2.29 IQ points per [mu]g/dL
across blood Pb levels extending below 2 [mu]g/dL.
---------------------------------------------------------------------------
\148\ In their September 2007 letter, the CASAC Pb Panel
``recommends using the two-piece linear function for relating IQ
alterations to current blood lead levels with a slope change or
``hinge'' point closer to 7.5 [mu]g/dL than 10.82 [mu]g/dL as used
by EPA staff in the second draft exposure/risk assessments document.
The higher value used by staff underestimates risk at lower blood Pb
levels, where most of the population will be located. Epidemiologic
data indicate that the slope of the line below 7.5 [mu]g/dL is
approximately minus three (-3) IQ decrements per 1 [mu]g/dL blood
lead and the vast majority of children in the U.S. have maximal
baseline Pb blood levels below 7.5 [mu]g/dL (Lanphear et al., EHP
2005; MMWR 2005). On a population level, the mean increase in blood
lead concentration from airborne lead would generally be up to, but
not exceeding, a blood lead concentration of 7.5 [mu]g/dL. This
approach should also account for sensitive subpopulations of
children.'' In in their January 2008 letter, the Panel also points
to several other studies ``confirming that the relationship of lead
exposure is non-linear and per-sists at blood lead levels
considerably lower than 5 [mu]g/dL (Lanphear, 2000; Wasserman, 2003;
Kordas, 2006; Tellez-Rojo, 2006). In particular, Tellez-Rojo and co-
workers reported that the slope of the association between 24-month
blood lead and the 24-month Mental Development Index (MDI) for 294
children who had peak blood lead levels below 5 [mu]g/dL was
negative (-1.7 points for each 1 [mu]g/dL increase in blood lead
concentration, p=0.01). Collectively, these studies indicate that
there is sufficient evidence to support the use of the dose-response
relationship from the pooled analysis at blood lead levels < 5
[mu]g/dL (Lanphear, 2005), as described in the Final Lead Staff
Paper and previously recommended by CASAC.''
---------------------------------------------------------------------------
The second set of C-R functions discussed in section II.B.2.b is
drawn from a larger group of studies, although these studies include
groups of children with higher blood Pb levels (CD, pp. 8-78 to 8-80)
such that the population mean levels for these studies include
population mean blood Pb levels ranging from a mean of 3.3 up to a
median of 9.7 [mu]g/dL (see Table 1). This second set of C-R functions
is represented by a median of -0.9 IQ points per [mu]g/dL blood Pb (CD,
p. 8-80).\149\
---------------------------------------------------------------------------
\149\ As noted above (in section II.B.2.b), this slope is
similar to the slope for the below 10 [mu]g/dL piece of the
piecewise model used in the RRP rule economics analysis.
---------------------------------------------------------------------------
In applying the air-related IQ loss evidence-based framework, as
with the adapted 1978 framework, we recognize uncertainty in our
estimates for the two input parameters (air-to-blood ratio and C-R
function slope). Accordingly, in associating various standard levels
with the estimated magnitudes of air-related mean IQ loss that would
likely be prevented by keeping exposed populations below such standard
levels, we have considered combinations of parameter estimates that are
potentially supportable within this framework. With regard to the C-R
functions we have drawn estimates from both sets of functions. For the
first set of C-R functions, we have relied on the upper and lower-end
values to provide a range at lower blood Pb levels, and have focused on
the LLL function for blood Pb levels above approximately 2.5 to 3.0
[mu]g/dL, as shown in Table 7.\150\ From the second set of C-R
functions, we have relied on the median estimate across the range of
blood Pb levels considered. For air-to-blood ratios, we have focused on
the estimate of 1:5 as above, and also provided IQ loss estimates using
higher and lower estimates of air-to-blood ratio (i.e., 1:3 and 1:7)
within the range supported by the evidence. These estimates are
presented in Table 7 below.
---------------------------------------------------------------------------
\150\ We derived estimates of air-related IQ loss using the LLL
(nonlinear) function giving equal weight to all contributions of Pb
to total blood Pb as illustrated by the following example. For a
level of 0.30 [mu]g/m\3\, and an air-to-blood ratio of 1:5, the
resultant estimate of air-related blood Pb is 1.5 [mu]g/dL. Using
estimates for nonair blood Pb levels of 1 and 1.4 [mu]g/dL, the
estimates of total blood Pb are 2.5 and 2.9 [mu]g/dL. The
corresponding total Pb-related IQ loss estimates based on the LLL
function are 5.2 and 5.6 points IQ loss. These estimates are then
multiplied by the fraction of total Pb that is air-related (i.e.,
1.5/2.5 and 1.5/2.9) to derive the estimated range of air-related IQ
loss (2.9-3.1 points).
Table 7.--Estimates of Air-Related Population Mean IQ Loss for Children Exposed at the Level of the Standard
--------------------------------------------------------------------------------------------------------------------------------------------------------
Air-related population mean IQ loss (points) for children exposed at level of the standard
---------------------------------------------------------------------------------------------------------------------------
Air-to-blood ratio of Air-to-blood ratio of Air-to-blood ratio of Air-to-blood ratio of Air-to-blood ratio of
Potential level for standard 1:3 1:4 1:5 1:6 1:7
([mu]g/m\3\) ---------------------------------------------------------------------------------------------------------------------------
1st group 2nd group 1st group 2nd group 1st group 2nd group 1st group 2nd group 1st group 2nd group
of C-R of C-R of C-R of C-R of C-R of C-R of C-R of C-R of C-R of C-R
functions functions functions functions functions functions functions functions functions functions
--------------------------------------------------------------------------------------------------------------------------------------------------------
0.50........................ * 2.9-3.1 1.4 * 3.5-3.8 1.8 * 4.1-4.3 2.3 * 4.6-4.8 2.7 * 5.0-5.3 3.2
0.40........................ * 2.4-2.6 1.1 * 3.0-3.2 1.4 * 3.5-3.8 1.8 * 4.0-4.2 2.2 * 4.4-4.6 2.5
0.30........................ 1.5-2.6 0.8 * 2.4-2.6 1.1 * 2.9-3.1 1.4 * 3.3-3.5 1.6 * 3.6-3.9 1.9
0.20........................ 1.0-1.8 0.5 1.4-2.4 0.7 1.7-2.9 0.9 * 2.4-2.6 1.1 * 2.7-3.0 1.3
0.10........................ 0.5-0.9 0.3 0.7-1.2 0.4 0.9-1.5 0.5 1.0-1.8 0.5 1.2-2.1 0.6
0.05........................ 0.3-0.4 0.14 0.3-0.6 0.18 0.4-0.7 0.2 0.5-0.9 0.27 0.6-1.0 0.3
0.02........................ 0.1-0.2 0.05 0.1-0.2 0.07 0.2-0.3 0.09 0.2-0.4 0.1 0.2-0.4 0.1
--------------------------------------------------------------------------------------------------------------------------------------------------------
* These estimates were derived using only the nonlinear C-R function from the risk assessment which, given its nonlinearity, EPA considers to better
assess risk across the range that includes extending into these higher standard levels (and the associated higher blood Pb levels). That is, the upper
and lower values presented in the asterisked cells are both derived using the LLL function, as described in the text and associated footnote above,
rather than using the two linear functions of -1.71 from Tellez-Rojo, 2005 (<5 [mu]g/dL subgroup) and -2.94 from Lanphear, 2005 (<7.5 [mu]g/dL peak
blood Pb subgroup) as is the case in the cells without asterisks.
Using the air-to-blood ratio of 1:5 with the range of slopes from
the first set of C-R functions indicates an air-related mean IQ loss
estimate of 0.9 to 1.5 points for a population of children exposed at
the standard level of 0.10 [mu]g/m\3\. Similarly, the air-related mean
IQ loss estimate for a standard level of 0.20 [mu]g/m\3\ is 1.7 to 2.9
points. Using the air-to-blood ratio of 1:5 and the slope from the
second set of C-R functions (from blood Pb levels extending up to 10
[mu]g/dL) in the calculation indicates an air-related mean IQ loss of
0.5 points for a population of children exposed at the standard level
of 0.10 [mu]g/m\3\; the corresponding air-related mean IQ loss estimate
for a standard level of 0.20 [mu]g/m\3\ is 0.9 points. Using the 1:5
air-to-blood ratio with first set of C-R functions indicates an air-
related mean IQ loss estimate of approximately 3 points for a
population of children exposed at the standard level of 0.30 [mu]g/
m\3\. Using the slope from the second set of C-R functions indicates an
air-related mean IQ loss estimate of 1.4 points for a population of
children exposed at the standard level of 0.30 [mu]g/m\3\.
[[Page 29240]]
As mentioned above, we recognize uncertainty in the air-to-blood
values, and have accordingly also considered estimates of air-to-blood
ratio that are lower and higher than the 1:5 value used above.
Accordingly, we note that using a lower air-to-blood ratio, such as 1:3
(low end of range from evidence) generally results in lower air-related
IQ loss estimates with either set of C-R functions (approximately 40%
lower than those using a ratio of 1:5). Similarly, use of a higher air-
to-blood ratio, such as 1:7, yields higher air-related mean IQ loss
estimates with either set of C-R functions (approximately 40% higher
than those using a ratio of 1:5).
In applying this framework, we have also considered higher standard
levels, above 0.30 [mu]g/m\3\ up to the highest alternative level
included in the risk assessment (e.g., up to 0.50 [mu]g/m\3\). Using
the 1:5 air-to-blood ratio with the first set of C-R functions, the
air-related mean IQ loss estimate for a standard level of 0.50 [mu]g/
m\3\ is approximately 4 points. Using the slope from the second set of
C-R functions indicates an air-related mean IQ loss estimate of 2.3
points for a population of children exposed at the standard level of
0.50 [mu]g/m\3\. Using the 1:3 air-to-blood ratio with the first set of
C-R functions indicates an air-related mean IQ loss estimate of
approximately 3 points for a population of children exposed at the
standard level of 0.50 [mu]g/m\3\. Using the 1:3 air-to-blood ratio and
the slope for the second set of C-R functions indicates an air-related
mean IQ loss estimate of 1.4 points for a population of children
exposed at the standard level of 0.50 [mu]g/m\3\.
Further, we have also considered lower standard levels, down to the
lowest alternative levels included in the risk assessment (e.g., 0.05
to 0.02 [mu]g/m\3\). For example, across both sets of C-R functions and
the range of air-to-blood ratios considered above (1:3 to 1:7), a
standard level of 0.05 [mu]g/m\3\ indicates an air-related mean IQ loss
of approximately 0.1 to 1 point. The estimates for either set of C-R
functions are approximately 50% lower at the standard level of 0.02
[mu]g/m\3\.
b. Exposure- and Risk-Based Considerations
To inform judgments about a range of levels for the standard that
could provide an appropriate degree of public health protection, in
addition to considering the health effects evidence, EPA also
considered the quantitative estimates of exposure and health risks
attributable to air-related Pb upon meeting specific alternative levels
of alternative Pb standards and the uncertainties in the estimated
exposures and risks, as discussed above in Section III.B. As discussed
above, the risk assessment conducted by EPA is based on exposures that
have been estimated for children of less than 7 years of age in several
case studies. The assessment estimated the risk of adverse
neurocognitive effects in terms of IQ loss associated with total and
air-related Pb exposures, including incidence of different magnitudes
of IQ loss in the three location-specific case studies. In so doing,
EPA is mindful of the important uncertainties and limitations that are
associated with the exposure and risk assessments. For example, with
regard to the risk assessment important uncertainties include those
related to estimation of blood Pb C-R functions, particularly for blood
Pb concentrations at and below the lower end of those represented in
the epidemiological studies characterized in the Criteria Document.
EPA also recognizes important limitations in the design of, and
data and methods employed in, the exposure and risk analyses. For
example, the available monitoring data for Pb relied upon for
estimating current conditions for the urban case studies are quite
limited, in that monitors are not located near many of the larger known
Pb sources, which results in potential underestimation of current
conditions, and there is uncertainty about the proximity of existing
monitors to other Pb sources potentially influencing exposures, such as
old urban roadways and areas where housing with Pb paint has been
demolished or has undergone extensive exterior renovation. All of these
limitations raise uncertainty as to whether these data adequately
capture the magnitude of ambient Pb concentrations to which the target
population is currently exposed. Additionally, EPA recognizes that
there is not sufficient information available to evaluate all relevant
sensitive groups (e.g., adults with chronic kidney disease) or all Pb-
related health effects (e.g., neurological effects other than IQ loss,
immune system effects, adult cardiovascular or renal effects), and the
scope of our analyses was generally limited to estimating exposures and
risks in case studies intended to illustrate a variety of Pb exposure
situations across the U.S., with three of them focused on specific
areas in three cities. As noted above, however, coordinated intensive
efforts over the last 20 years have yielded a substantial decline in
blood Pb levels in the United States. Recent NHANES data (2003-2004)
yield blood lead level estimates for children age 1 to 5 years of 1.6
[mu]g/dL (median) and 3.9 [mu]g/dL (90th percentile). These median and
90th percentile national-level data are lower than modeled values
generated for the three location-specific urban case studies current
conditions scenarios (described in section II.C.3.a above). As noted in
section II.C.3.a, however, the urban case studies and the NHANES study
are likely to differ with regard to factors related to Pb exposure,
including ambient air levels (e.g., the national median ambient air Pb
concentrations are generally lower than those in the location-specific
case studies).
As described in section II.C.2.e, we also recognize limitations in
our ability to characterize the contribution of air-related Pb to total
Pb exposure and Pb-related health risk. As a result, we have
approximated estimates for the air-related pathways, bounded on the low
end by exposure/risk estimated for the ``recent air'' category and on
the upper end by the exposure/risk estimated for the ``recent air''
plus ``past air'' categories.\151\
---------------------------------------------------------------------------
\151\ As noted in section II.C.2.e above, the recent air
category does not include a variety of air-related categories
(including some associated with air deposition to outdoor surfaces
and diet) and both the recent air and past categories may include
some Pb in soil or dust from the historical use of Pb in paint.
---------------------------------------------------------------------------
We generally focus in this discussion on risk estimates derived
using the LLL (log-linear with low exposure linearization) C-R
function. Further, in considering the risk estimates in light of IQ
loss estimates (described in section II.E.3.a) of the air-related IQ
loss evidence-based framework, we focus here on risk estimates for the
general urban and primary Pb smelter subarea case studies as these
cases studies generally represent population exposures for more highly
air-pathway exposed children residing in small neighborhoods or
lozalized residential areas with air concentrations nearer the standard
level being evaluated than do the location-specific case studies in
which populations have a broader range of air-related exposures
including many well below the standard level being evaluated.
In considering the results of the risk assessment for the
alternative standard levels assessed, we note that the risk estimates
are roughly consistent with and generally supportive of the evidence-
based mean air-related IQ loss estimates described above (section
II.E.3.a). For example, at a standard level of 0.20 [mu]g/m\3\, the
evidence-based approach indicates estimates of mean air-related IQ loss
ranging from less than
[[Page 29241]]
1 to approximately 3 points IQ loss, while the median air-related risk
estimates for this level in the general urban case study are
represented by a lower bound near 1 point IQ loss and an upper bound
near 3 points IQ loss. The corresponding upper bound air-related IQ
loss estimate for the primary Pb smelter case study subarea is 3.7
points. Alternatively, at a standard level of 0.50 [mu]g/m\3\, the
evidence-based approach indicates estimates of mean air-related IQ loss
ranging from approximately 1.5 points to greater than 4 points, while
the median air-related risk estimates for this level for the general
urban case study are represented by a lower bound near 2 points IQ loss
and an upper bound just below 4 points IQ loss (section II.C.3.b). The
corresponding upper bound air-related IQ loss estimate for the primary
Pb smelter case study subarea is 4.5 points. Also, while the risk
assessment did not specifically assess the standard levels of 0.10 and
0.30 [mu]g/m\3\, we note that estimates for these levels based on
interpolation from the estimates described above are also roughly
consistent with and generally supportive of the evidence-based mean
air-related IQ loss estimates described in section II.E.3.a above
(Murphy and Pekar, 2008).
As mentioned above (section II.E.3.a), the Staff Paper conclusions
on level for the primary Pb standard built on staff 's conclusion that
the overall body of evidence clearly calls into question the adequacy
of the current standard with regard to health protection afforded to
at-risk populations. Drawing from both consideration of the evidence
and consideration of the quantitative risk and exposure information
(described in section II.E.3.b), staff concluded that the available
information provides strong support for consideration of a range of
standard levels that are appreciably below the level of the current
standard in order to provide increased public health protection for
these populations, with support for this conclusion. With regard to the
risk estimates, the Staff Paper recognized that, to the extent one
places weight on risk estimates for the lower standard levels, those
estimates may suggest consideration of a range of levels that extend
down to the lowest levels assessed in the risk assessment, 0.02 to 0.05
[mu]g/m\3\. In summary, the Staff Paper concluded that ``a level for
the standard set in the upper part of [the staff] recommended range
(0.1-0.2 [mu]g/m\3\, particularly with a monthly averaging time) is
well supported by the evidence and also supported by estimates of risk
associated with policy-relevant Pb that overlap with the range of IQ
loss that may reasonably be judged to be highly significant from a
public health perspective, and is judged to be so by CASAC'' (USEPA,
2007c). Further, the Staff Paper concluded that ``a standard set in the
lower part of the range would be more precautionary and would place
weight on the more highly uncertain range of estimates from the risk
assessment'' (USEPA, 2007c).
c. CASAC Advice and Recommendations and Public Comments
Beyond the evidence- and risk/exposure-based information discussed
above, EPA's consideration of the level for the TSP-based standard also
takes into account the advice and recommendations of CASAC, based on
their review of the Criteria Document, the Staff Paper and the related
technical support document, and the ANPR, as well as comments from the
public on drafts of the Staff Paper and related technical support
document and the ANPR.
In their advice to the Agency during this review CASAC has
recognized the importance of both the health effects evidence and the
exposure and risk information in selecting the level for the TSP-based
standard (Henderson, 2007a, 2007b, 2008). In two separate letters sent
prior to publication of the ANPR, CASAC stated that it is the unanimous
judgment of the CASAC Lead Panel that the primary NAAQS should be
``substantially lowered'' to ``a level of about 0.2 [mu]g/m\3\ or
less,'' reflecting their view of the health effects evidence
(Henderson, 2007a,b). In their most recent letter, reflecting their
review of the ANPR and Staff Paper, the Panel reiterated their earlier
judgment, stating that ``[t]he Committee unanimously and fully supports
Agency staff's scientific analyses in recommending the need to
substantially lower the level of the primary (public-health based) Lead
NAAQS, to an upper bound of no higher than 0.2 [mu]g/m\3\ with a
monthly averaging time.''
The CASAC Pb Panel also provided advice regarding how the Agency
should consider IQ loss estimates derived from the risk assessment in
selecting a level for the standard (Henderson, 2007a). The Panel stated
that they consider a population loss of 1-2 IQ points to be ``highly
significant from a public health perspective''.
Among the many public comments the Agency has received in this
review regarding the level of the standard, the overwhelming majority
recommended appreciable reductions in the level, e.g., setting it at
0.2 [mu]g/m\3\ or less, while only a few recommended that the Agency
make no or only a modest adjustment. Among the comments recommending
appreciable reduction, many noted the importance of considering
exposures and risks to vulnerable and susceptible populations. Some
recognized that blood Pb levels are disproportionately elevated among
minority and low-income children, and recommended more explicit
consideration of issues of environmental justice. And some comments
also noted the need for the standard to provide an adequate margin of
safety, indicating that such a need might provide support for
consideration of much lower levels. The American Academy of Pediatrics
recommended that EPA set the level at 0.2 or lower, and also
recommended that EPA consider the approach developed by the State of
California Environmental Protection Agency (Cal-EPA) for the purposes
of school site assessment, which has at its goal prevention of a rise
in blood Pb level that Cal-EPA has predicted to be associated with an
incremental increase estimated to decrease IQ by 1 point.
d. Administrator's Proposed Conclusion Concerning Level
For the reasons discussed below, and taking into account
information and assessments presented in the Criteria Document and
Staff Paper, the advice and recommendations of CASAC, and the public
comments to date, the Administrator proposes to revise the existing
primary Pb standard. Specifically, the Administrator proposes to revise
the level of the primary Pb standard, defined in terms of the current
Pb-TSP indicator, to within the range of 0.10 to 0.30 [mu]g/m\3\,
conditional on judgments as to the appropriate values of key parameters
to use in the context of the air-related IQ loss evidence-based
framework discussed below.
Further, in recognition of alternative views of the science, the
exposure and risk assessments, the uncertainties inherent in the
science and these assessments, and the appropriate public health policy
responses based on the currently available information, the
Administrator also solicits comments on whether to proceed instead with
alternative levels of a primary Pb-TSP standard within ranges from
above 0.30 [mu]g/m\3\ up to 0.50 [mu]g/m\3\ and below 0.10 [mu]g/m\3\.
Based on the comments received and the accompanying rationales, the
Administrator may adopt other standards within the range of the
alternative levels identified above in lieu of the standards he is
proposing today. In addition, as discussed below, the Administrator
also solicits comments on when, if ever, it would be
[[Page 29242]]
appropriate to set a NAAQS for Pb at a level of zero.
The Administrator's consideration of alternative levels of the
primary Pb-TSP standard builds on his proposed conclusion, discussed
above in section II.D.4, that the overall body of evidence indicates
that the current Pb standard is not requisite to protect public health
with an adequate margin of safety and that the standard should be
revised to provide increased public health protection, especially for
members of at-risk groups, notably including children, against an array
of adverse health effects. These effects range from IQ loss, a health
outcome that could be quantified in the risk assessment, to health
outcomes that could not be directly estimated, including decrements in
other neurocognitive functions, other neurological effects and immune
system effects, as well as cardiovascular and renal effects in adults.
In reaching a proposed decision about the level of the Pb primary
standard, the Administrator has considered: the evidence-based
considerations from the Criteria Document and the Staff Paper and those
based on the air-related IQ loss evidence-based framework discussed
above; the results of the exposure and risk assessments discussed above
and in the Staff Paper, giving weight to the exposure and risk
assessments as judged appropriate; CASAC advice and recommendations, as
reflected in discussions of the Criteria Document, Staff Paper, and
ANPR at public meetings, in separate written comments, and in CASAC's
letters to the Administrator; EPA staff recommendations; and public
comments received during the development of these documents, either in
connection with CASAC meetings or separately. In considering what
standard is requisite to protect public health with an adequate margin
of safety, the Administrator is mindful that this choice requires
judgment based on an interpretation of the evidence and other
information that neither overstates nor understates the strength and
limitations of the evidence and information nor the appropriate
inferences to be drawn.
In reaching a proposed decision on a range of levels for a revised
standard, as in reaching a proposed decision on the adequacy of the
current standard, the Administrator primarily considered the evidence
in the context of the air-related IQ loss evidence-based framework
described above in section II.E.3.a.ii. As a general matter, in
considering this evidence-based framework, the Administrator recognizes
that in the case of Pb there are several aspects to the body of
epidemiological evidence that add complexity to the selection of an
appropriate level for the primary standard. As discussed above, these
complexities include evidence based on blood Pb as the dose metric,
exposure pathways that are both air-related and nonair-related, and the
absence of any discernible threshold levels in the health effects
evidence. Further, the Administrator recognizes that there are a number
of important uncertainties and limitations inherent in the available
health effects evidence and related information, including
uncertainties in the evidence of associations between total blood Pb
and neurocognitive effects in children, especially at the lowest blood
Pb levels evaluated in such studies, as well as uncertainties in key
parameters used in this evidence-based framework, including C-R
functions and air-to-blood ratios. In addition, the Administrator
recognizes that there are currently no commonly accepted guidelines or
criteria within the public health community that would provide a clear
basis for reaching a judgment as to the appropriate degree of public
health protection that should be afforded to neurocognitive effects in
sensitive populations, such as IQ loss in children.
The air-related IQ loss evidence-based framework considered by the
Administrator focuses on quantitative relationships between air-related
Pb and neurocognitive effects (e.g., IQ loss) in children, building on
recommendations from CASAC to consider the body of evidence in a more
quantitative manner. More specifically, this framework is premised on a
public health goal of selecting a standard level that would prevent
air-related IQ loss (and related effects) of a magnitude judged by the
Administrator to be of concern in populations of children exposed to
the level of the standard, taking into consideration uncertainties
inherent in such estimates. In addition to this public health policy
judgment regarding IQ loss, two other parameters are relevant to this
framework--a C-R function for population IQ response associated with
blood Pb level and an air-to-blood ratio. Based on the discussion of
these parameters in section II.E.3.a above, the Administrator concludes
that, in considering alternative standard levels below the level of the
current standard, it is appropriate to take into account the same two
sets of C-R functions, recognizing uncertainties in the related
evidence, as was done in considering the adequacy of the current
standard (as discussed above in section II.D). He notes that the first
set of C-R functions reflects the evidence indicative of steeper slopes
in relationships between blood Pb and IQ in children, and that the
second set of C-R functions reflects relationships with shallower
slopes between blood Pb and IQ in children. In addition, the
Administrator concludes that it is appropriate to consider various air-
to-blood ratios, again recognizing the uncertainties in the relevant
evidence. He notes that an air-to-blood ratio of 1:5 is within the
reasonable range of values that EPA considers to be generally supported
by the available evidence, which includes ratios of 1:3 up to 1:7.
With regard to making a public health policy judgment as to the
appropriate level of protection against air-related IQ loss and related
effects, the Administrator first notes that ideally air-related (as
well as other) exposures to environmental Pb would be reduced to the
point that no IQ impact in children would occur. The Administrator
recognizes, however, that in the case of setting a NAAQS, he is
required to make a judgment as to what degree of protection is
requisite to protect public health with an adequate margin of safety.
The NAAQS must be sufficient but not more stringent than necessary to
achieve that result, and does not require a zero-risk standard.
Considering the advice of CASAC and public comments on this issue,
notably including the comments of the American Academy of Pediatrics,
the Administrator proposes to conclude that an air-related population
mean IQ loss within the range of 1 to 2 points could be significant
from a public health perspective, and that a standard level should be
selected to provide protection from air-related population mean IQ loss
in excess of this range.
The Administrator considered the application of this air-related IQ
loss framework with this target degree of protection in mind, drawing
from the information presented in Table 7 above in section II.E.3.a.ii
that addresses a broad range of standard levels. In so doing, the
Administrator first focused on the estimates associated with the first
set of C-R functions in conjunction with the range of air-to-blood
ratios considered by EPA in this framework. Specifically, using an air-
to-blood ratio of 1:5, the Administrator notes that a standard level of
0.10 [mu]g/m\3\ would limit the estimated degree of impact on
population mean IQ loss from air-related Pb to no more than 1.5 points,
the mid-point of the proposed range of protection. Using the full range
of air-to-blood ratios considered in this framework (1:3 to 1:7), he
notes that a standard set at this level (0.10 [mu]g/m\3\) would limit
the estimated degree of air-
[[Page 29243]]
related impact on population mean IQ loss to a range from less than 1
point to around 2 points. Again based on the first set of C-R
functions, the Administrator notes that a standard level of 0.20 [mu]g/
m\3\ would also limit the estimated degree of air-related impact on
population mean IQ loss to within the proposed range of protection
based on using an air-to-blood ratio of 1:3.
In considering the use of the second set of C-R functions in
conjunction with the range of air-to-blood ratios considered in this
framework (1:3 to 1:7), the Administrator notes for example that a
standard set within the range of 0.10 to 0.30 [mu]g/m\3\ would limit
the estimated degree of air-related impact on population mean IQ loss
to a range from less than one-half point to just under 2 points. More
specifically, based on using an air-to-blood ratio of 1:5 (the
approximately central estimate) in conjunction with the second set of
C-R functions, the Administrator notes that a standard level of 0.30
[mu]g/m\3\ would limit the estimated degree of impact on population
mean IQ loss from air-related Pb to just under 1.5 points, the mid-
point of the proposed range of protection.
Taking these considerations into account, and based on the full
range of information presented in Table 7 above on estimates of air-
related IQ loss in children over a broad range of alternative standard
levels, the Administrator concludes that it is appropriate to propose a
range of standard levels, and that a range of levels from 0.10 to 0.30
[mu]g/m\3\ is consistent with his target for protection from air-
related IQ loss in children. In recognition of the uncertainties in
these key parameters, the Administrator believes that the selection of
a standard level from within this range is conditional on judgments as
to the most appropriate parameter values to use in the context of this
evidence-based framework. For example, he notes that placing more
weight on the use of a C-R function with a relatively steeper slope
would tend to support a standard level in the lower part of the
proposed range, while placing more weight on a C-R function with a
shallower slope would tend to support a level in the upper part of the
proposed range. Similarly, placing more weight on a higher air-to-blood
ratio would tend to support a standard level in the lower part of the
proposed range, whereas placing more weight on a lower ratio would tend
to support a level in the upper part of the range. In soliciting
comment on a standard level within this proposed range, the
Administrator specifically solicits comment on the appropriate values
to use for these key parameters in the context of this evidence-based
framework, reflecting that his proposal to revise the level of the
primary Pb standard, defined in terms of the current Pb-TSP indicator,
to within the range of 0.10 to 0.30 [mu]g/m\3\ is conditional on
judgments as to the appropriate values of key parameters to use in this
context.
The Administrator has also considered the results of the exposure
and risk assessments conducted for this review to provide some further
perspective on the potential magnitude of air-related IQ loss. The
Administrator finds that these quantitative assessments provide a
useful perspective on the risk from air-related Pb. However, in light
of the important uncertainties and limitations associated with these
assessments, as discussed above in sections II.C and II.E.3.b, for
purposes of evaluating potential new standards, the Administrator
places less weight on the risk estimates than on the evidence-based
assessments. Nonetheless, the Administrator finds that the risk
estimates are roughly consistent with and generally supportive of the
evidence-based air-related IQ loss estimates described above, as
discussed above in section II.E.3.b. This lends support to the proposed
range based on this evidence-based framework.
In the Administrator's view, the above considerations, taken
together, provide no evidence- or risk-based bright line that indicates
a single appropriate level. Instead, there is a collection of
scientific evidence and judgments and other information, including
information about the uncertainties inherent in many relevant factors,
which needs to be considered together in making this public health
policy judgment and in selecting a standard level from a range of
reasonable values. Based on consideration of the entire body of
evidence and information available at this time, as well as the
recommendations of CASAC and public comments, the Administrator is
proposing that a standard level within the range of 0.10 to 0.30 [mu]g/
m\3\ would be requisite to protect public health, including the health
of sensitive groups, with an adequate margin of safety. He also
recognizes that selection of a level from within this range is
conditional on judgments as to what C-R function and what air-to-blood
ratio are most appropriate to use within the context of the air-related
IQ loss framework. The Administrator notes that this proposed range
encompasses the specific level of 0.20 [mu]g/m\3\, the upper end of the
range recommended by CASAC and by many public commenters. The
Administrator provisionally concludes that a standard level selected
from within this range would reduce the risk of a variety of health
effects associated with exposure to Pb, including effects indicated in
the epidemiological studies at low blood Pb levels, particularly
including neurological effects in children, and cardiovascular and
renal effects in adults.
Because there is no bright line clearly directing the choice of
level within this reasonable range, the choice of what is appropriate,
considering the strengths and limitations of the evidence, and the
appropriate inferences to be drawn from the evidence and the exposure
and risk assessments, is a public health policy judgment. To further
inform this judgment, the Administrator solicits comment on the air-
related IQ loss evidence-based framework considered by the Agency and
on appropriate parameter values to be considered in the application of
this framework. More specifically, we solicit comment on the
appropriate C-R function and air-to-blood ratio to be used in the
context of the air-related IQ loss framework. The Administrator also
solicits comment on the degree of impact of air-related Pb on IQ loss
and other related neurocognitive effects in children considered to be
significant from a public health perspective, and on the use of this
framework as a basis for selecting a standard level.
For the reasons discussed above, the Administrator proposes to
revise the level of the primary Pb standard, defined in terms of the
current Pb-TSP indicator, to within the range of 0.10 to 0.30 [mu]g/
m\3\, conditional on judgments as to the appropriate C-R functions and
air-to-blood ratio to use in the context of the air-related IQ loss
framework.
The Administrator notes that this framework indicates that for
standard levels above 0.30 [mu]g/m\3\ up to 0.50 [mu]g/m\3\, the
estimated degree of impact on population mean IQ loss from air-related
Pb would range from approximately 2 points to 5 points or more with the
use of the first set of C-R functions and the full range of air-to-
blood ratios considered, and would extend from somewhere within the
proposed range of 1 to 2 points IQ loss to above that range when using
the second set of C-R functions and the full range of air-to-blood
ratios considered. The Administrator proposes to conclude in light of
his consideration of the evidence in the framework discussed above that
the magnitude of air-related Pb effects at the higher blood Pb levels
that would be allowed by standards above 0.30 up to 0.50 [mu]g/m\3\
would be greater than what is requisite to protect
[[Page 29244]]
public health with an adequate margin of safety.
In addition, the Administrator notes that for standard levels below
0.10 [mu]g/m\3\, the estimated degree of impact on population mean IQ
loss from air-related Pb would generally be somewhat to well below the
proposed range of 1 to 2 points air-related population mean IQ loss
regardless of which set of C-R functions or which air-to-blood ratio
within the range of ratios considered are used. The Administrator
proposes to conclude that the degree of public health protection that
standards below 0.10 [mu]g/m\3\ would likely afford would be greater
than what is requisite to protect public health with an adequate margin
of safety.
Having reached this proposed decision based on the interpretation
of the evidence, the evidence-based frameworks, the exposure/risk
assessment, and the public health policy judgments described above, the
Administrator recognizes that other interpretations, frameworks,
assessments, and judgments are possible. There are also potential
alternative views as to the range of values for relevant parameters
(e.g., C-R function, air-to-blood ratio) in the evidence-based
framework that might be considered supportable and the relative weight
that might appropriately be placed on any specific value for these
parameters within such ranges. In addition, the Administrator
recognizes that there may be other views as to the appropriate degree
of public health protection that should be afforded in terms of air-
related population mean IQ loss in children that would provide support
for alternative standard levels different from the proposed range.
Further, there may be other views as to the appropriate weight and
interpretation to give to the exposure/risk assessment conducted for
this review. Consistent with the goal of soliciting comment on a wide
array of issues, the Administrator solicits comment on these and other
issues.
In particular, the Administrator solicits comment on alternative
levels of a primary Pb-TSP standard of above 0.30 [mu]g/m\3\ up to 0.50
[mu]g/m\3\. In considering the air-related IQ loss framework and the
case when the second set of C-R functions is used in conjunction with
the lowest air-to-blood ratio considered in this framework (i.e., 1:3),
a standard level as high as 0.50 [mu]g/m\3\ would still limit the
estimated degree of impact on population mean IQ loss from air-related
Pb to no more than 1.5 points, the mid-point of the proposed range of
protection. Comment is solicited on levels within this range and the
associated rationale for selecting such a level in terms of the
appropriate weight to place on relevant parameter values that may
extend to values outside the ranges of values considered by EPA, or in
terms of alternative evidence- or risk-based frameworks that might
support standard levels within this range.
In addition, the Administrator solicits comment on alternative
levels below 0.10 [mu]g/m\3\. In considering the evidence-based
framework discussed above, a standard level within this range would
likely provide a degree of protection in terms of air-related
population mean IQ loss that is greater than the proposed range based
on the use of any of the relevant parameter values within the ranges
considered by EPA. Comment is solicited on levels within this range and
the associated rationale for selecting such a level in terms of the
appropriate weight to place on relevant parameter values that may
extend to values outside of the ranges considered by EPA, or
alternative public health policy judgments as to the degree of
protection that is warranted, or the appropriate weight to place on the
results of the risk assessment.
More broadly, as discussed above, the Administrator recognizes that
Pb can be considered a non-threshold pollutant.\152\ In recognizing
that no threshold has been identified below which we are scientifically
confident that there is no risk of harm, EPA's views are consistent
with the views of the CDC, the Federal agency that tracks children's
blood Pb levels nationally and provides guidance on levels at which
medical and environmental case management activities should be
implemented (CDC, 2005a; ACCLPP, 2007). In 2005, CDC revised its
statement on Preventing Lead Poisoning in Young Children, specifically
recognizing the evidence of adverse health effects in children and the
data demonstrating that no ``safe'' threshold for blood Pb had been
identified (CDC, 2005a). EPA's views are also consistent with other
organizations, including, for example, the American Academy of
Pediatrics that recognized in commenting on the ANPR that ``[t]here is
no known ``safe'' level of blood lead in children'' (AAP, 2008). In
addition, the California Environmental Protection Agency, in a recent
risk assessment report, recognizes that ``no safe level has been
definitively established'' for effects of Pb in children (CalEPA, 2007,
p. 1). Given the current state of scientific evidence, which does not
resolve the question of whether or not there is a threshold, we
recognize that there is no level below which we can say with scientific
confidence that there is no risk of harm from exposure to ambient air
related lead.
---------------------------------------------------------------------------
\152\ Similarly, in the most recent reviews of the NAAQS for
ozone and PM, EPA recognized that the available epidemiological
evidence neither supports nor refutes the existence of thresholds at
the population level, while noting uncertainties and limitations in
studies that make discerning thresholds in populations difficult
(e.g., 73 FR 16444, March 27, 2008; 71 FR 61158, October 17, 2006).
---------------------------------------------------------------------------
The Administrator also recognizes, as discussed in section I.A
above, that the CAA does not require that NAAQS be established at a
zero-risk level, but rather at a level that reduces risk sufficiently
so as to protect public health with an adequate margin of safety. In
setting primary standards that are ``requisite'' to provide the this
degree of public health protection, the Supreme Court has affirmed that
EPA's task is to establish standards that are neither more nor less
stringent than necessary for this purpose. The question then becomes
how the Agency should reconcile these scientific and legal
understandings in reviewing the Pb NAAQS.
As discussed above, EPA is proposing a range of levels for the
primary Pb NAAQS, with the range extending down to 0.10 [mu]g/m\3\.
This range reflects the Administrator's proposed conclusion that lower
levels would be more than necessary to protect public health with an
adequate margin of safety. This proposed conclusion is based in large
part on EPA's evaluation of the evidence, recognizing important
uncertainties in the scientific evidence and related assessments, and
reflects the proposed public heath policy judgment of the Administrator
on these issues. As discussed above, these uncertainties stem in part
from the complexities of determining the health impact of air-related
Pb given the multi-media exposure pathways for exposure to lead and the
persistence of Pb in the environment. The major areas of uncertainty
include the appropriate air-to-blood ratio; the apportionment of Pb
between air-related and nonair Pb; the increasing uncertainty at lower
blood Pb levels as to the existence, nature, and degree of health
effects; and the uncertainty over the public health significance of
smaller and smaller impacts on IQ or other similar neurocognitive
metrics from exposure to air-related Pb. In recognition of such
uncertainties, EPA is also soliciting comment on a lower range of
standard levels below 0.10 [mu]g/m\3\.
In so doing, EPA fully recognizes that a standard set at the lowest
proposed level of 0.10 [mu]g/m\3\, or any non-zero level, would not be
a risk-free standard.
[[Page 29245]]
As in numerous prior NAAQS reviews, we recognize that the CAA does not
require that EPA set a risk-free standard. Instead, EPA is to recognize
and take risk into account, and set a standard that is requisite to
protect public health with an adequate margin of safety based on the
currently available information. This calls for a public health policy
judgment informed by many factors, most notably the nature and severity
of the health effects at issue, the size of the population(s) at risk,
and the kind and degree of uncertainties involved. After considering
all of these factors in this review, the Administrator's proposed
judgment is that a standard set below 0.10 [mu]g/m\3\ would not satisfy
this statutory directive.
The Administrator recognizes that the current state of the
scientific evidence clearly indicates that health effects from Pb occur
at much lower blood Pb levels than we understood in the past, and that
the appropriate level for ambient air Pb is much lower than we thought
in the past. Further the Administrator expects that, as time goes on,
future scientific studies will continue to enhance our understanding of
Pb, and anticipates that such studies might lead to a situation where
there is very little, if any, remaining uncertainty about human health
impacts from even extremely low levels of Pb in the ambient air. As
noted above, this has the potential to raise fundamental questions as
to how the Agency can continue to reconcile such evidence with the
statutory pro