UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
April 26, 2006
EPA-CASAC-06-005
Honorable Stephen L. Johnson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Subject: Clean Air Scientific Advisory Committee's (CASAC) Peer Review of the
Agency ' s Air Quality Criteria for Lead (First External Review Draft), Volumes
7ara///(EPA/600/R-05/144aA-bA, December 2005)
Dear Administrator Johnson:
EPA's Clean Air Scientific Advisory Committee (CASAC), supplemented by subject-
matter-expert Panelists — collectively referred to as the CASAC Lead Review Panel (Lead
Panel) — met in a public meeting held in Durham, NC, on February 28 and March 1, 2006, to
conduct an initial peer review of the Agency ' s Air Quality Criteria for Lead (First External
Review Draft), Volumes I and II (EPA/600/R-05/144aA-bA, December 2005), also known
simply as the 1st draft Lead Air Quality Criteria Document (AQCD). The current CASAC roster
is found in Appendix A of this report, and the Lead Panel roster is attached as Appendix B. The
charge questions provided to the Lead Panel by EPA staff are contained in Appendix C to this
report, and Lead Panelists' individual review comments are provided in Appendix D.
The members of the Lead Panel were generally pleased with the quality of this 1st draft
Lead AQCD, but regret the lack of an integrative synthesis chapter in this initial draft. The Lead
Panel approved of the organization that was used in most chapters, i.e., starting with a brief
review of what was in the earlier previous AQCD, followed by the new information obtained
since the publication of the earlier document, and ending with a short summary. However, the
Lead Panel suggested that the beginning of the initial chapter should clearly state the question to
be addressed, i.e., Is there evidence that the current lead standards need to be made more (or less)
stringent? In order to address this crucial question, the Lead AQCD needs to place greater
emphasis on the adverse effects that occur at low levels of lead exposure. The Lead Panel was
concerned that the document, as written, focuses too much on material relevant to occupational,
high-level exposures. In addition, the Panel expressed concerns with respect to the superficial
and incomplete discussion of neurobehavioral effects of lead exposure and the lack of update to
the neurobehavioral literature published since EPA's issuance of the previous Lead AQCD.
-------
1. Background
The CASAC, comprising seven members appointed by the EPA Administrator, was
established under section 109(d)(2) of the Clean Air Act (CAA or "Act") (42 U.S.C. § 7409) as
an independent scientific advisory committee, in part to provide advice, information and
recommendations on the scientific and technical aspects of issues related to air quality criteria
and national ambient air quality standards (NAAQS) under sections 108 and 109 of the Act.
Section 109(d)(l) of the CAA requires that EPA carry out a periodic review and revision, where
appropriate, of the air quality criteria and the NAAQS for "criteria" air pollutants such as lead.
The CASAC, which is administratively located under EPA's Science Advisory Board (SAB)
Staff Office, is a Federal advisory committee chartered under the Federal Advisory Committee
Act (FACA), as amended, 5 U.S.C., App. The CASAC Lead Review Panel consists of the seven
members of the chartered (statutory) Clean Air Scientific Advisory Committee, supplemented by
thirteen technical experts.
EPA is in the process of updating, and revising where appropriate, the AQCD for lead.
Section 109(d)(l) of the Clean Air Act (CAA) requires that EPA carry out a periodic review and
revision, as appropriate, of the air quality criteria and the national ambient air quality standards
(NAAQS) for the six "criteria" air pollutants including lead. On December 1, 2005, EPA's
National Center for Environmental Assessment National, Research Triangle Park (NCEA-RTP),
within the Agency's Office of Research and Development (ORD), made available for public
review and comment a revised draft document, Air Quality Criteria for Lead (First External
Review Draft), Volumes I andII (EPA/600/R-05/144aA-bA). This first draft Lead AQCD
represents a revision to the previous EPA document, Air Quality Criteria for Lead., EPA-600/8-
83/028aF-dF (published in June 1986) and an associated supplement (EPA-600/8-89/049F)
published in 1990. Under CAA sections 108 and 109, the purpose of the revised AQCD is to
provide an assessment of the latest scientific information on the effects of ambient lead on the
public health and welfare, for use in EPA's current review of the NAAQS for lead. Detailed
summary information on the revised draft AQCD for lead is contained in a recent EPA Federal
Register notice (70 FR 72300, December 2, 2005).
2. CASAC Lead Review Panel's Peer Review of the 1st Draft Lead AQCD
The initial peer review of EPA's first external review draft AQCD for lead took place in a
public meeting held in Durham, NC, on February 28 and March 1, 2006. Specific comments
aimed at improving the individual chapters of the 1st draft Lead AQCD are given below. These
are aimed at pointing out omissions, places in the document where additional or more-balanced
interpretations are needed, and sections where the organization of the AQCD could be improved.
Responses to the charge questions are given either directly or indirectly in the comments on each
chapter or, in some cases, in the individual comments of Lead Panel members attached to this
letter (Appendix D).
In general, Chapter 2, "Chemistry, Sources, and Transport of Lead," is well-written
and adequately summarizes pertinent information regarding chemistry, natural and
anthropogenic sources and transport of lead in the environment. However, the information
relative to production, active sources, emission rates, particle size, total lead emissions and
ambient air lead levels is outdated or missing. For example, lead emission data from coal and
-------
fuel oil combustion and some metallurgical processes rely, for the most part, on older references
(e.g., Pacyna, 1986). Accurate and informed emission inventory data are critical to provide a
perspective in establishing and implementing protective health and environmental standards for
atmospheric lead.
EPA has limited this review to ".. .where information is available in the peer-reviewed
literature." However, data and better information for production, emissions, industry transition
and economic indicators may be found in the trade literature and government agency records.
The Lead Panel felt that, if adequate peer-review literature data do not exist, the use of publicly-
available reports and reliable compilation of data is justified. In that event, data quality and
reliability should be assessed and discussed as the material is incorporated in the document. If
usable emission and source characterization data are unavailable or unsuitable for use in the
standard setting process, the critical need to have these updated must be emphasized.
Other concerns with Chapter 2 are that it: (1) fails to put the various emission categories
in context; (2) is somewhat fragmented; and (3) is not well-integrated with the remainder of the
document. Putting the information into a broader historic, national and global context and
acknowledging local problem areas in the U.S. would be beneficial. Several sources and source
categories are listed, but it is not clear which are the most important ones. Some ranking should
be provided. In addition, a number of additional examples that tie chemical and physical
mechanisms in environmental and biological processes to material presented in later Chapters
would be helpful. Since particle size information, for example, is scattered through different
sections; a summary section or table would be useful.
Chapter 3 on "Routes of Human Exposure to Lead" is, in general, a good discussion
of this field. There are, however, several modifications that would enhance the chapter, such as
an overview and introduction. It would be helpful if there was a description of the scope of the
systematic approach that was used to identify the critical papers on lead exposure published since
1990. There should be a stronger focus on the relative contributions of various sources of lead
exposure.
Chapter 3 includes substantial information about dust lead as opposed to airborne lead.
This is important because these two exposure sources are substantially interrelated. Moreover,
dust is the most proximal exposure for contemporary children. The information on these two
related exposure sources should be organized into two separate sections of Chapter 3 for clarity.
The chapter should also include a discussion of how soil and road dust are affected by human
activity. In addition, as was pointed-out during the meeting, this chapter omits information on
studies that relate drinking water lead to blood lead, differences between first-flush, partially-
flushed, and fully-flushed samples, etc. Since EPA is also in the process of revising its lead
standard for drinking water, this information should be updated and included in this chapter.
The chapter should include insights that explain the contribution and trends in lead
exposure. For example, it may not be obvious to all readers that the various sources of lead
intake are cumulative, and that blood lead (in children) and bone lead (in adolescents and adults)
are cumulative biomarkers of exposure. Residential exposures (i.e.., lead-based paint and lead-
-------
contaminated house dust and water) have become increasingly important sources of lead
ingestion following the phase-out of leaded gasoline and reductions in dietary intake.
This review should describe the relative contribution of various sources of lead exposure
that vary by age using epidemiologic studies. For example, children's blood lead levels rise
rapidly between six and 12 months of age, peak between 18 months to 36 months and then
gradually decline. Lead-contaminated floor dust is a source of lead intake throughout early
childhood, but lead-contaminated dust on windowsills is not a major source of intake until the
second year of life, when children stand upright. Soil ingestion, as reported by parents, peaks
during the second year of life and diminishes thereafter. It would be useful if the chapter
summarized how our understanding of lead exposure has changed since the 1990 supplement.
For example, there have been several randomized trials published since 1990 that provide insight
into the relative contribution of lead intake from various sources that the chapter authors
overlooked (e.g., Haynes, 2001; Jordan, 2004; Brown, 2005; Aschengrau, 1994).
Finally, the chapter leaves the impression that exterior sources of lead are more important
sources of lead in house dust than interior sources, such as lead-contaminated paint. This may be
true for mining, milling or smelting communities, but it is not true for many older urban
communities. There are also a number of relevant studies that document sources of lead in
human dust that are not currently included in Chapter 3. (See Lead Panel members' individual
review comments found in Appendix D.) A discussion of the efficacy of lead paint abatement
should also be included in the chapter.
The authors have provided a good first draft of Chapter 4, "Models of Human
Exposure to Lead and Observed Environmental Concentrations " They have captured most
of the basic information needed to understand the strengths and weaknesses of the various
kinetic-based dosimetry models for lead in humans incorporating the oral, dermal, and inhalation
routes of exposure. The Lead Panel recognizes the need for models to relate blood lead levels to
environmental lead concentrations. Some Panel members also recognize the potential utility of
slope-factor (i.e., epidemiologic) models for this purpose. There is some disagreement among
Lead Panel members as to the most scientifically-valid approach for estimating blood lead levels
from environmental lead media concentrations. Most Lead Panel members think the biokinetic
and physiologically-based models are the most valid to use for estimating blood lead levels,
while some feel that the slope-factor models are more appropriate. The various opinions of Lead
Panel members on this issue are captured in members' individual review comments, which are
attached as Appendix D. In any case, the CASAC wishes to emphasize that Agency staff needs
to explore, carefully and in detail, the comparative usefulness of the slope-ratio models and the
biokinetic- and physiologically-based models for the purpose of addressing:
• Short and long-term exposures;
• Low exposure concentrations;
• Relevant exposure routes and media types;
• Both site-specific and national average exposures; and
• All age ranges and sensitive populations.
-------
An assessment of these issues should be reflected in the 2nd draft Lead AQCD.
In addition to the individual comments of Lead Panel members that are attached to this
letter, the Lead Panel offers the following comments and recommendations to improve the
chapter. The chapter is currently missing a "bottom line" as to which model or models would be
the most appropriate for use in the assessment of potential risks in humans from exposure to
lead. The Lead Panel does not consider the EPA "All-Ages" Lead Model (AALM) to be ready
for "prime time" use in routine applications given that it is still in development. The material
related to this model should be minimized or deleted altogether.
Numerous studies have been conducted that have related concentrations of lead in various
media in and around people's homes with the resident's blood lead. Regression or slope-factor
models that relate the blood lead to media lead have resulted from these efforts. There is also an
international pooling project of such studies. These studies and slope-factor models should be
summarized and reviewed, along with an assessment of their strengths and limitations
comparable to the assessment that was done for the physiologically-based models.
The Leggett and O'Flaherty models are closer to being ready for use but still require
more work before they can be regarded as the "workhorse" models for use in risk assessments
for lead-impacted groups of individuals. That being said, there are instances where these models
can be useful. For example, the O'Flaherty model is probably the most robust of the models in
terms of estimating blood lead levels (chronic and transient) associated with "absorbed" doses of
lead. To this degree, the use of slope factors that relate absorbed doses of lead to blood lead
levels is perfectly valid and easily-conducted using the O'Flaherty model after stripping off its
exposure modules.
The effective interface between modeling and epidemiological data as both relate to Pb
NAAQS should include:
• the role of air Pb-related dust lead and dust lead loadings for children in light of new
sub-10 ng/dl thresholds for adverse health effects;
• the hazards to children of dust lead per the EPA floor dust rule; and
• the multi-media impacts of existing lead levels in these media.
In addition, the Integrated Exposure Uptake Biokinetic (IEUBK) model should be
validated by using it to predict the distribution of blood lead concentrations in the National
Health and Nutrition Examination Survey (NHANES) data, using reasonable assumptions about
exposure, and adjusted as necessary. Modeling of child Pb exposure via the IEUBK model
should include lowering the not-to-exceed 10 |ig/dl level that sets IEUBK distributions for
percentile protections to a lower value consistent with current findings about Pb toxicity.
Section 4.7 on Slope Factor Models needs to be expanded to provide results available from
regression models derived from epidemiological studies. In addition, Section 4.8 (Model
Comparisons) should discuss when the use of models derived from epidemiological studies
might be preferable to use of mechanistic models for estimation of blood lead levels.
-------
Chapter 4's appraisal of the IEUBK model's blood lead level predictions should
acknowledge limitations identified in the literature (in particular, Bowers and Mattuck, 2001).
The chapter should also caution against inappropriate superimposition of the lognormal
distribution on the IEUBK model output to estimate the risk of having a blood lead level above a
specified threshold. The IEUBK model should be modified in its exposure module to handle
dust lead loadings from air Pb and loading-related intakes in children by ingestion. The chapter
currently focuses on blood lead in children, but the epidemiology data provide results for adverse
effects in adults as well. Thus, the Lead Panel recommends Chapter 4 provide more information
for predicting blood lead levels in adults. The IEUBK model, which is the model most currently
used, only addresses children up to seven years of age.
Finally, the chapter currently does not do an adequate job of addressing lead deposition
and clearance by the inhalation route of exposure nor does it recognize the importance of the size
of lead particles in determining where and how much is deposited in various regions of the
respiratory tract. In addition, the authors need to address how these aspects vary for children
compared to adults. More specifics are needed on the bioavailability of lead once it has been
ingested, inhaled or absorbed through the skin. Information on model parameter values and
variables should be included in an annex to this chapter.
Chapter 5, "Toxicological Effects of Lead in Laboratory Animals, Humans and In
Vitro Systems," could be better-organized to provide balanced treatment of the topics and
reduce redundancies. There should be succinct conclusions at the end of each section. Specific
suggestions for reorganizations are in the individual comments attached to this letter. The
experimental animal behavior literature has shown comparable results that occur at blood lead
levels corresponding to those at which such effects are seen in humans. Much of that literature
has been reported since the last iteration of the lead air quality criteria document.
Unfortunately, however, the current draft Lead AQCD does not adequately or critically
cover this literature, focusing instead primarily on neurochemical and electrophysiological
effects of Pb. In addition, the blood Pb levels at which effects are observed is inconsistently
reported. The experimental literature on Pb-associated behavioral toxicity has reported effects at
levels of 10-11 |ig/dl and has also resulted in significant new information related to further
defining the basis of reported changes in IQ. For example, there are several studies examining
changes in various aspects of attention, particularly sustained attention and impulsivity. These
are not covered critically, and in fact, the interpretation that sustained attention is a major effect
of Pb fails to consider contradictory findings, or the deficiencies in the one study that is cited,
where the magnitude of the reported changes in sustained attention are minimal. Therefore, this
section of the document requires significant revision. Also to be re-considered is whether the
extensive coverage of the other aspects of nervous system effects requires reporting in such
depth and detail. Moreover, the discussion of cardiovascular effects of lead jumps immediately
into a discussion of studies suggesting potential pathways for the lead-blood pressure
association, and should briefly summarize the earlier literature that established that association.
Finally, the Lead Panel had extensive concerns about Section 5.3 with respect to the
superficial and incomplete discussion of neurobehavioral effects of lead exposure. Contradictory
findings are presented and are neither explained nor evaluated in any depth. In addition, this
-------
section does not update the neurobehavioral literature published since EPA issued its previous
Lead AQCD. For example, potential lead-induced deficits in attention are not critically
reviewed. These studies are particularly important as they relate to the behavioral mechanisms
underlying cognitive deficits and no doubt relate to the changes observed in IQ. These concerns
are reported in-depth in the individual comments of Dr. Cory-Slechta, which are found in
Appendix D, beginning on Page D-6.
Although Chapter 6, "Epidemiologic Studies of Human Health Effects Associated
with Lead Exposures," is well-organized in first presenting the findings from the 1986 and
1990 update of the previous Lead AQCD and then updating the published literature to 2005, it
suffers from several aspects of the time pressure expressed by Agency staff at the meeting with
respect to meeting court-ordered deadlines. The chapter was clearly put together by a number of
writers and needs to undergo significant editing. In many places much of the material is repeated
and many of the same studies are used in each of the sections. This leads to redundancy of
presentation of material that all needs to be there but should be said only once.
More important for those using the AQCD who only read this chapter, there needs to be
some place where the multimedia exposure sources are presented. Clearly there are multiple
biomarker methods that are discussed to define exposure. This presents a significant advantage
over other criteria pollutants; however, in many places the text reads more like a medical
textbook and, particularly for the naive reader, the fact that the toxin being discussed is for a
multimedia environmental and occupational pollutant gets lost. Some brief summary of
information contained in Chapter 2 needs to be included in this chapter.
There are several places where reference is made to the Annex Tables, particularly in the
biomarker discussion and in the neurobehavioral sections. In large part the Lead Panel agreed
with this approach. However, there are a number of places where a summary graph, analysis, or
table in the text would be useful. This is particularly true in the neurobehavioral sections where
important effects at low level of exposure are discussed. The later sections on Renal and
Cardiovascular effects seem to have included sufficient tabular and graphic examples.
Furthermore, there are a number of specific issues that are discussed in detail in the
individual comments attached. Several broad specifics are mentioned as follows:
• The issue of measurement error in outcomes and the implication of such errors need
further discussion.
• The conclusion that a single blood lead is a relatively-poor index of body lead burden
is too broad.
• The discussion that long-term lead body burden represents the "gold standard" for
exposure is not the case. Acute exposures that affect blood lead levels may or may
not change body burden but may be important predictors of adverse effects.
• The statement that there is no consistent evidence of effects in adults if competing
risks are taken into account is unsupportable.
-------
• While the section on neurobehavioral effects mentions the consistency with animal
toxicology, this is not done with the blood pressure/cardiovascular effects where
similar mechanistic data exist.
• The section on exposure misclassification needs to focus on the epidemiological
implications.
With regard to the Agency charge questions related to this chapter, for the most part the
reviewers believed that most of the questions could be answered in the affirmative. The
exception was with Charge Question QF2a, which had to do with model selection. The pooled
analyses used a log-linear analysis to quantify the lead-associated IQ decrements. It was not
explicit in the write up of this work that the non-linear values observed were not due to the
influence of the model. Several suggestions in the individual comments were made to help
clarify this section.
Finally, several suggested data sets were offered to all Agency staff to test a number of
the models considered. (These were provided to EPA in subsequent e-mails). These need to be
incorporated in the next draft of the Lead AQCD.
Chapter 7, the critical "Integrative Synthesis" chapter, was not completed in time for
the Lead Panel's review of the 1st draft Lead AQCD. NCEA-RTP is developing this chapter for
the 2nd draft of the Lead document, which the Lead Panel is scheduled to review in June 2006.
Chapter 8, "Environmental Effects of Lead" summarizes a large fraction of the
accumulated body of knowledge concerning the effects of "atmospherically-deposited lead" on
the soils, sediments, waters, and biota of terrestrial and aquatic ecosystems. The chapter is well-
organized, with clearly written summaries of terrestrial effects (8.1.1) and aquatic effects (8.2.1),
along with more detailed information intended to be included as annexes.
Nevertheless, some significant additional intellectual work is needed in preparing the
Second External Review Draft and especially its Executive Summary and the Integrative
Synthesis chapter. Both of these additional sections should summarize scientific knowledge
regarding effects of lead on both public health and the environment. The information in Chapter
8 needs to be presented in a way that is more directly relevant to the issue of whether the EPA
Administrator should retain, increase, or decrease the present primary and secondary NAAQS for
lead. Since secondary standards are often (neglectfully) set equal to primary standards, a key
question is whether there are environmental effects that occur at lead concentrations lower than,
or for indicators, forms, or averaging times different from, those that affect human health.
Very substantial decreases in air concentrations and atmospheric deposition of lead into
the environment were achieved in recent decades. Thus, most current exposures of living
organisms in natural and managed ecosystems are caused primarily by redistribution of
environmentally persistent airborne lead compounds deposited in soils, sediments, and surface
waters prior to the phase-out of leaded gasoline in the 1970s and 1980s. Chapter 8 specifically
needs to more clearly indicate how any continuing environmental effects of lead might respond
to changes in current and future atmospheric lead emissions, concentrations or deposition. The
chapter might better help EPA prepare for such changes if it included a more complete and/or
-------
balanced analysis of the status of new advances in the science relevant to environmental
management of lead. For example, consideration of monitoring needs and the implications of
dietary exposure and trophic transfer are needed, as is more balance in considering equilibrium
partitioning in sediments and uses of the biotic ligand model. It would also be useful to consider
how environmental effects of historically deposited lead or future increases in deposition (if
current laws are relaxed) might be modified by land-use changes, or soil amendment treatments,
or interactions with other pollutants, including other metals or acidifying pollutants, or with
changes in climate and climate processes.
Members of CASAC were especially pleased to see the relatively-thorough discussion at
the end of Chapter 8 regarding the alternative concepts of critical loads, critical limits, target
loads, and target times that have been developed in European and Canadian scientific literature
to guide the processes of decision making regarding both environmental and public health effects
of airborne chemicals.
In conclusion, the Clean Air Scientific Advisory Committee and the Lead Panel
encourage EPA in its continued efforts to protect the public health and our environment from
adverse effects of ambient lead. The Committee looks forward to reviewing the 2nd draft Lead
AQCD — and particularly Chapter 7, the Integrative Synthesis — and the first draft of the
Agency's Lead Staff Paper. As always, we wish the Agency staff well in this important
endeavor.
Sincerely,
/Signed/
Dr. Rogene Henderson, Chair
Clean Air Scientific Advisory Committee
Appendix A - Roster of the Clean Air Scientific Advisory Committee
Appendix B - Roster of the CASAC Lead Review Panel
Appendix C - Agency Charge to the CASAC Lead Review Panel
Appendix D - Review Comments from Individual CASAC Lead Review Panel Members
-------
Appendix A - Roster of the Clean Air Scientific Advisory Committee
U.S. Environmental Protection Agency
Science Advisory Board (SAB) Staff Office
Clean Air Scientific Advisory Committee (CASAC)
CHAIR
Dr. Rogene Henderson, Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM
MEMBERS
Dr. Ellis Cowling, University Distinguished Professor-at-Large, North Carolina State
University, Colleges of Natural Resources and Agriculture and Life Sciences, North Carolina
State University, Raleigh, NC
Dr. James D. Crapo, Professor, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO
Dr. Frederick J. Miller, Consultant, Cary, NC
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard
Medical School, Boston, MA
Dr. Barbara Zielinska, Research Professor, Division of Atmospheric Science, Desert Research
Institute, Reno, NV
SCIENCE ADVISORY BOARD STAFF
Mr. Fred Butterfield, CASAC Designated Federal Officer, 1200 Pennsylvania Avenue, N.W.,
Washington, DC, 20460, Phone: 202-343-9994, Fax: 202-233-0643 (butterfield.fred@epa.gov)
(Physical/Courier/FedEx Address: Fred A. Butterfield, III, EPA Science Advisory Board Staff
Office (Mail Code 1400F), Woodies Building, 1025 F Street, N.W., Room 3604, Washington,
DC 20004, Telephone: 202-343-9994)
A-l
-------
Appendix B - Roster of the CASAC Lead Review Panel
U.S. Environmental Protection Agency
Science Advisory Board (SAB) Staff Office
Clean Air Scientific Advisory Committee (CASAC)
CASAC Lead Review Panel
CHAIR
Dr. Rogene Henderson*, Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM
MEMBERS
Dr. Joshua Cohen, Faculty, Center for the Evaluation of Value and Risk, Institute for Clinical
Research and Health Policy Studies, Tufts New England Medical Center, Boston, MA
Dr. Deborah Cory-Slechta, Director, University of Medicine and Dentistry of New Jersey and
Rutgers State University, Piscataway, NJ
Dr. Ellis Cowling*, University Distinguished Professor-at-Large, North Carolina State
University, Colleges of Natural Resources and Agriculture and Life Sciences, North Carolina
State University, Raleigh, NC
Dr. James D. Crapo [M.D.]*, Professor, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO
Dr. Bruce Fowler, Assistant Director for Science, Division of Toxicology and Environmental
Medicine, Office of the Director, Agency for Toxic Substances and Disease Registry, U.S.
Centers for Disease Control and Prevention (ATSDR/CDC), Chamblee, GA
Dr. Andrew Friedland, Professor and Chair, Environmental Studies Program, Dartmouth
College, Hanover, NH
Dr. Robert Goyer [M.D.], Emeritus Professor of Pathology, Faculty of Medicine, University of
Western Ontario (Canada), Chapel Hill, NC
Mr. Sean Hays, President, Summit Toxicology, Allenspark, CO
Dr. Bruce Lanphear [M.D.], Sloan Professor of Children's Environmental Health, and the
Director of the Cincinnati Children's Environmental Health Center at Cincinnati Children's
Hospital Medical Center and the University of Cincinnati, Cincinnati, OH
Dr. Samuel Luoma, Senior Research Hydrologist, U.S. Geological Survey (USGS), Menlo
Park, CA
B-l
-------
Dr. Frederick J. Miller*, Consultant, Cary, NC
Dr. Paul Mushak, Principal, PB Associates, and Visiting Professor, Albert Einstein College of
Medicine (New York, NY), Durham, NC
Dr. Michael Newman, Professor of Marine Science, School of Marine Sciences, Virginia
Institute of Marine Science, College of William & Mary, Gloucester Point, VA
Mr. Richard L. Poirot*, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Michael Rabinowitz, Geochemist, Marine Biological Laboratory, Woods Hole, MA
Dr. Joel Schwartz, Professor, Environmental Health, Harvard University School of Public
Health, Boston, MA
Dr. Frank Speizer [M.D.]*, Edward Kass Professor of Medicine, Channing Laboratory,
Harvard Medical School, Boston, MA
Dr. Ian von Lindern, Senior Scientist, TerraGraphics Environmental Engineering, Inc.,
Moscow, ID
Dr. Barbara Zielinska*, Research Professor, Division of Atmospheric Science, Desert Research
Institute, Reno, NV
SCIENCE ADVISORY BOARD STAFF
Mr. Fred Butterfield, CASAC Designated Federal Officer, 1200 Pennsylvania Avenue, N.W.,
Washington, DC, 20460, Phone: 202-343-9994, Fax: 202-233-0643 (butterfield.fred@epa.gov)
* Members of the statutory Clean Air Scientific Advisory Committee (CASAC) appointed by the EPA
Administrator
B-2
-------
Appendix C - Agency Charge to the CASAC Lead Review Panel
OVERVIEW OF SALIENT ASPECTS OF THE DECEMBER 2005 1st DRAFT LEAD
AQCD AND ASSOCIATED CHARGE QUESTIONS FOR THE FEBRUARY 28 -
MARCH 1, 2006 CASAC LEAD REVIEW PANEL PUBLIC MEETING
A. Format and Structure of the Draft Lead AQCD.
In developing the December 2005 1st Draft Lead AQCD, NCEA followed past CASAC
advice to streamline the format of the document, in order to facilitate timely CASAC and public
review by focusing more clearly on those issues most relevant to the policy assessment to be
provided in the Lead Staff Paper. As described in Chapter 1 of the 1st draft Lead AQCD, chief
emphasis is placed on interpretative evaluation and integration of evidence in the main body of
the document, with more detailed descriptions of individual studies being presented in a series of
accompanying annexes. Key information from lead-related literature previously assessed in
prior lead NAAQS reviews is only succinctly summarized (usually without citation) at the
opening of each section or subsection, to provide a very brief overview of previous work. For
more detailed discussion of such information, readers are referred to EPA's 1986 Lead AQCD,
an associated 1986 Addendum, and its follow-on 1990 Supplement. This format is intended to
make each chapter of the main Lead AQCD a manageable length, to focus on interpretation and
synthesis of relevant new research, and to lessen or avoid redundancy with previous Lead AQCD
materials.
As for overall structure and content, after an introductory chapter (Chapter 1), the 1st
Draft Lead AQCD presents chapters addressing three main topic areas:
• Characterization of properties of lead and its environmental dispersal including
discussion of: (a) the chemistry, sources, and transport (Chapter 2); and (b) observed
environmental concentrations and routes of human exposure (Chapter 3);
• Pb-related health effects, including discussion of modeling of human exposure impacts
on lead body burdens (Chapter 4), toxicological effects in animals, humans, and in vitro
test systems (Chapter 5), and epidemiology studies (Chapter 6). (Please note that the
integrative synthesis of Pb-related health effects will be included as Chapter 7 in the
Second External Review Draft of the Lead AQCD, to be made available later in 2006 for
public comment and CASAC review); and
• Pb-related welfare effects, including discussion of environmental effects of Pb on
vegetation and ecosystems (Chapter 8).
C-l
-------
Charge Questions Al. To what extent is the document format (i.e.., main chapters of the
1st draft AQCD focused on evaluative/interpretive aspects, with descriptive materials and
tables presented in annexes) useful and desirable? Can the structure be further improved?
If so, how?
B. Lead Chemistry, Sources and Transport (Chapter 2).
Chapter 2 summarizes available information on the chemistry, natural and anthropogenic
sources, and transport of Pb in the environment. The discussion of lead's chemistry is limited to
properties of importance in the environment and in biological systems. Industrial uses of lead
are summarized in tables. Sources and transport mechanisms are described in greater detail.
Important mechanisms for transport of Pb in the environment that are discussed include:
advection, deposition, resuspension, runoff, leaching, aquatic cycling, plant uptake, and ingestion
by livestock and wildlife. Advection in the atmosphere is the mechanism of greatest importance
in this discussion. The major reservoirs identified are soils and sediments.
Charge Questions Bl. Overall, does Chapter 2 provide adequate coverage of important
chemical properties of lead and concise summarization of pertinent information on
sources of Pb and Pb emissions, especially in relation to the United States? In particular,
how well does Chapter 2 identify the most pertinent available datasets that contain
information on emission rates for point and area sources? Also, does the discussion of
available data adequately address issues such as the spatial distribution of point and area
sources and emissions estimate uncertainties?
Furthermore, does the discussion satisfactorily address emissions by key industrial
sectors? Does Chapter 2 adequately address other important issues relating to the
dispersal and/or accumulation of Pb in the environment, e.g., resuspension of roadside
dust or the potential for Pb to accumulate in some media, like soils, due to its relatively
low mobility? (The latter fact means that fairly low air Pb concentrations have the
potential to produce elevated soil concentrations over time due to wet and dry
deposition.) In addition, does the chapter adequately discuss key chemical and transport-
related factors that should be considered in evaluating long-term buildup of Pb in the
environment? Finally, are the discussions of the leaching of Pb from soil and sediment
into surface and groundwater sufficiently complete for this chapter?
C. Environmental Exposure Pathways and Concentrations (Chapter 3).
Chapter 3 summarizes scientific information on routes by which humans are exposed to
lead and the concentrations observed in pertinent environmental media, including air (i.e.,
indoors, outdoors and occupational settings) and soil and dust (near-point sources, roads, and in
urban settings). The available information on lead found in drinking water, food, and other
sources (e.g., paint, dietary supplements, pottery glazes, window blinds and hair dye) is also
discussed. The techniques used for measuring environmental Pb concentrations are described so
as to provide background for the reader on detection issues and potential sources of uncertainty.
Available evidence indicates that Pb concentrations are elevated in all environmental media in
urban areas. Highest concentrations are found near stationary sources and roadways. The most
C-2
-------
important routes of exposure in the U.S. are by ingestion of food and waterborne lead or, in some
areas, via contact with soils and/or house dust contaminated with Pb from deteriorated older
leaded paint or from secondary deposition of airborne Pb from smelter emissions.
Charge Questions Cl. Does Chapter 3 provide adequate coverage of pertinent available
information (especially as it pertains to the United States) on Pb exposure routes, as well
as environmental Pb concentrations, including those in air, drinking water, food, soils,
and dust? Also, does the chapter delineate adequately interconnections between airborne
Pb and its potential contributions (via secondary deposition) to Pb in other media (e.g.,
indoor dust)?
Moreover, given the potential importance of historical deposition of Pb from mobile
sources, does the chapter adequately identify key sources of information characterizing
the magnitude and distribution of lead soil concentrations near roadways in urban,
suburban and rural areas? Also, given the importance of characterizing "background" Pb
concentrations in conducting health/ecological impact analyses (where background refers
to both natural and generalized anthropogenic contributions as distinct form specific
point sources), does the chapter adequately denote key sources of information
characterizing existing "background" Pb levels in urban, suburban and rural/pristine
areas?
D. Modeling of Lead Exposure Impacts on Internal Lead Burdens (Chapter 4).
The multimedia nature of Pb exposure must be considered in making decision on standards
to lessen risks for adverse health effects projected to be associated with Pb exposures of
susceptible populations. Scientific rationales underlying most EPA lead-related regulatory or
remedial action decisions typically include estimation of the impact of exposures to Pb in air,
water, food, soil/dust or other media on internal Pb body burden indices, for example, blood or
bone Pb levels. Chapter 4 discusses historical evolution of the modeling of external Pb exposure
impacts on internal Pb body burdens in various tissues, especially those widely used to index
increased risk of Pb-induced health effects (e.g., concentrations in blood and bone). This
includes modeling activities related to development of EPA's 1978 Lead NAAQS and the
generation of the EPA Integrated Exposure, Uptake, Biokinetic Model for Lead (i.e., Lead
IEUBK Model) in the late 1980s. The IEUBK Model has provided a tool for estimating
distributions of blood Pb levels among pediatric populations less than six (< 6) years old likely to
result from exposures to varying levels of lead from one or another media. As such, it has been
widely-used to support development of standards or guidance for control of Pb in air or drinking
water or remediation of Pb-based paint and Pb-contaminated soils and house dust. During recent
years, EPA has also initiated efforts to further refine and expand the Lead IEUBK Model and its
software to create an All-Ages Lead Model (AALM), which not only estimates the impact of Pb
exposures from various media on blood lead levels in young infants and children < 6 yrs. old (as
per its progenitor, the IEUBK Model), but also aims to project Pb exposure impacts on blood and
bone Pb of older children and adults through age 90 years (as well as the unborn fetus exposed
via transplacental transfer of Pb). Thus, the AALM aims to broaden the array of potentially-
susceptible population groups that can be more readily evaluated with regard to the extent to
which various Pb exposure scenarios may pose risks of undue elevations of internal Pb body
burdens and associated health impacts.
C-2
-------
Charge Questions Dl. How well does Chapter 4 concisely characterize key information
on: (a) the evolution and key features of important available approaches to the modeling
of external Pb exposures and their impacts on internal Pb body burdens; and (b) the status
of model evaluation efforts, e.g., PBPK model code verification and comparisons of
model-predicted versus observed impacts on blood or bone Pb distributions of particular
lead exposure scenarios for affected population groups? Also, does Chapter 4
sufficiently characterize the ability of different models to handle key factors related to
lead exposure modeling, including: temporal variation in external exposure profiles; low-
level lead exposure; multi-pathway lead exposure; and the contribution of
historical/artifact lead exposure in influencing blood lead levels?
Furthermore, given that the October 2005 SAB review of the AALM suggested that
further model validation and verification was needed before the AALM should be used in
support of regulatory development, does Chapter 4 clearly identify which alternative
models (e.g.., IEUBK, O'Flaherty) should be used for adult and/or child modeling instead
of the fledgling AALM? In addition, does Chapter 4 adequately identify the strengths
and weaknesses of the recommended models in modeling adult and child populations?
Finally, overall, how can Chapter 4 be improved without notable extension of length?
E. Toxicologic Evaluation of Lead Health Effects (Chapter 5).
An extensive lead toxicology literature is available, derived from controlled laboratory
experiments carried out in various laboratory animals, including primates. Chapter 5 mainly
focuses on newer scientific literature that has accumulated in the past 15 years or so since the last
prior Pb criteria review. This includes discussion of interesting new findings elucidating novel
information regarding lead effects on cardiovascular system and immune system functioning, as
well as impacts on bone and teeth, in addition to new insights into effects on more traditionally-
recognized lead target organs and tissues, e.g., the nervous system, the renal and hepatic systems,
and blood components.
Charge Questions El. Have any important new animal toxicology studies been
overlooked in Chapter 5 discussions on short- and long-term effects of Pb? Also, what
guidance can be provided by the CASAC Lead Review Panel with regard to the
following sub-questions or -issues?
Ela. Discussions in the neurotoxicology section focus mainly on lead effects on
glutamatergic synapses, synaptic plasticity, protein kinase C, and sensory systems
and learning. Are there any other areas pertinent to Pb neurotoxicology now
missing or inadequately covered in this section?
Elb. To what extent does the existing scientific literature provide evidence for
developmental Pb toxicity having a permanent impact on bone and teeth structure
and for these tissues serving as Pb storage pools forming long-term internal
sources of lead exposure for other body tissues?
C-4
-------
Elc. Are the animal toxicology studies with chelation/intervention agents
relevant to analogous studies in humans, and is the discussion of such studies of
sufficient relevance for current purposes to include coverage of them here?
Eld. Do the newer insights gained on Pb-induced micro molecular alterations on
erythrocyte biology, Pb-binding and transport kinetics, and altered nucleotide
pools suggest molecular mechanisms of action? Are they suggestive of
mechanisms underlying specific health endpoints?
Ele. Does the oxidative stress theory appear plausible for Pb toxicity and perhaps
represent a common mode of action operating across organs and species?
Elf. Concentrations of Pb compounds used in animal toxicology studies often
appear high and not necessarily very representative of ambient exposure
scenarios. What advice can the Panel provide to identify a cut-off value for
utilizing the biochemical and molecular toxicologic observations obtained under
these exposure conditions in extrapolating animal toxicology study findings to
humans in later development of an integrative synthesis chapter?
F. Epidemiologic Studies of Lead Exposures and Health Effects (Chapter 6).
Chapter 6 mainly assesses evidence derived from epidemiologic studies on associations
between both short-term and long-term biomarkers of Pb exposure and various health endpoints.
Such endpoints include: neurotoxic effects of lead in children and adults; renal effects;
cardiovascular effects; reproductive and developmental effects; genotoxic and carcinogenic
effects; effects on the immune system; and effects on various other organ systems. Important
new findings from numerous studies have been published since the 1986 Lead
AQCD/Addendum and the 1990 Supplement — including, perhaps most notably, evidence for
increased risk of neurotoxicity in children at low blood Pb levels below 10 //g/dL. Numerous
issues are discussed in Chapter 6 with regard to assessing: (a) the credibility of newly-reported
findings being attributable to Pb acting alone or in combination with other potential confounders
(e.g., socioeconomic status and home environment, inter-individual variability in susceptibility to
Pb toxicity); and (b) the health significance of changes observed on an individual or population
basis. EPA is seeking advice from the CASAC Lead Panel with regard to the following
questions or sub-issues related to Chapter 6.
Charges Questions Fl. Different biological markers of Pb exposure and body burden
are discussed in Chapter 6. The discussion concludes that higher blood Pb concentrations
can be interpreted as indicating higher exposures (or lead uptakes), but do not necessarily
predict appreciably higher body burdens. Bone lead is considered an indicator of
cumulative Pb exposure, with Pb in the skeleton being regarded as a potential continuing
internal source of Pb exposure for other tissues. Are the discussions on the various
biomarkers adequate to elucidate their role in assessing human health effects from Pb
exposure? Also, as the results from prospective cohort studies of Pb exposure have
become available, our understanding of the optimal exposure metric to use in modeling
specific health endpoints has evolved (e.g., initially peak blood Pb levels were favored
for child IQ, but that position now appears to be shifting toward concurrent or lifetime-
C-5
-------
averaged blood Pb levels). Does Chapter 6 adequately address this issue of which
exposure metrics are now believed to be most strongly associated with specific health
endpoints and, therefore, should be the focus of exposure and risk assessments targeting
those endpoints?
Charge Questions F2a. Newly-available human epidemiologic studies provide evidence
for slowed physical and neurobehavioral development being associated with blood Pb
levels ranging well below 10 //g/dL, and possibly to as low as 2 //g/dL. There is a
focused discussion of one large pooled study that examined the association between
blood Pb levels and IQ deficits in children from Boston, MA; Cincinnati, OH; Cleveland,
OH; Rochester, NY; Mexico City, Mexico; Port Pirie, Australia; and Kosovo,
Yugoslavia. The individual studies, which cover a wide range of exposure and outcome
values, generally found negative associations between blood Pb levels and IQ. The
pooled analysis shows a significant negative Pb effect on IQ measured at school age,
after adjusting for common confounders. Due to the log-linear relationship, the slope of
the Pb effect on IQ was greatest at the lower blood Pb level range, i.e., below 10 //g/dL.
Does this chapter adequately address questions regarding significant neurotoxic effects
observed at low blood Pb levels (<10 //g/dL)? Also, is the issue of the influence of
model selection on the estimated health effects adequately discussed?
Charge Questions F2b. In addition to other neurotoxic effects of Pb (e.g., disturbances
in behavior, mood, and social conduct; neuromotor function; and brain anatomical
development and activity), other important Pb effects involving the renal system,
cardiovascular system, reproductive and developmental system, immune system, and
various other organ systems are discussed in Chapter 6. The genotoxic and carcinogenic
potential of lead is also discussed. Does this chapter provide an adequate overview of
key Pb health effects of concern? Are the key summary statements and conclusions
regarding the effects of Pb on various organ systems sufficiently substantiated by the
assessed epidemiologic evidence?
Charge Question F2c. Recent studies have observed significant effects on various
health outcomes at relatively low lead levels. Examples include effects of lead on IQ,
blood pressure, and early biomarkers of preclinical renal damage. However, there is
concern as to what level of change for various health endpoints may be considered
adverse or clinically significant on an individual or population basis. What are the views
of the Panel regarding this issue?
Charge Questions F2d. Drawing causal inferences between increased Pb exposure and
adverse health effects in epidemiologic studies is complicated by the presence of many
potential confounders that may both affect Pb exposure and be associated with the health
outcome of interest. Examples of potential confounders for Pb health effects include
socioeconomic status, maternal IQ, maternal smoking, alcohol use, birth weight, and
many others depending on the health outcome. In this chapter, is the discussion of the
various potential confounders of Pb health effects adequate? Given the concern
regarding the influence of such confounders on the effect estimates, are the stated key
conclusions regarding Pb effects on various health outcomes appropriate?
C-6
-------
Charge Questions F3. Discussions of epidemiologic studies mainly focus on studies of
potential Pb effects among infants, school-aged children, the general population, and
occupationally-exposed populations. Some studies also examined potentially susceptible
individuals such as those with chronic medical diseases and specific genetic
polymorphisms. Does Chapter 6 adequately cover key populations that should be
considered for present purposes? Are the discussions of differences in individual
vulnerability and susceptibility adequate?
G. Integrated Synthesis (Chapter 7).
Due to time constraints, NCEA staff did not attempt to craft an Integrative Synthesis
(Chapter 7). At this time, CAS AC Panel suggestions as to the format and content of this chapter
would be welcomed. Such input would allow NCEA staff to focus on those points of greatest
importance for inclusion in the next (2nd) Draft Lead AQCD.
H. Characterization of Pb-Related Welfare Effects (Chapter 8).
1. Terrestrial Ecosystem Effects. Sections 8.1.1 through 8.1.6 present new information on
relevant measurement methods and the distribution of Pb in ecosystems and its effects on
terrestrial species and ecosystems. Section 8.1.1 is intended to serve as the main body of the
terrestrial effects portion of Chapter 8, while the other sections will ultimately serve as annexes
to Chapter 8, similar to the format used for the Ozone AQCD. Thus, the initial perceived
redundancy between Section 8.1.1 and the other sections in the chapter will be resolved later in
preparing the Second Draft Lead AQCD.
Charge Questions HI: Do the subject sections adequately cover the most current (since
1996) information on the measurement methods, distribution, and effects of Pb on
terrestrial ecosystems? Is there important material that was missed that should be
covered in the next draft of the chapter?
2. Aquatic Ecosystem Effects. Sections 8.2.1 through 8.2.6 present new information on the
measurement methods, distribution, and effects of Pb on aquatic species and ecosystems.
Section 8.2.1 is intended to serve as the main body of the aquatics effects portion of Chapter 8,
while the other sections will serve as annexes to Chapter 8, similar to the format used for the
Ozone AQCD. Thus, the initial redundancy between Section 8.2.1 and the other sections in the
chapter will be resolved in the Second Draft Lead AQCD.
Charge Questions H2: Do the subject sections adequately cover the most current (since
1996) available information on the measurement methods, distribution, and effects of Pb
on aquatic ecosystems? Is there any important additional material that should be covered
in the next draft of the chapter?
3. Critical Loads for Pb in Aquatic and Terrestrial Ecosystems. Sections 8.3 presents the
latest information on the application of a "critical loads" approach for protecting aquatic and
terrestrial ecosystems from the detrimental effects of atmospherically-delivered Pb.
C-7
-------
Charge Questions H3: Does the subject section adequately cover the most current
(since 1996) information on the potential use of critical loads? Is there important
additional material that should be covered in the next draft of the chapter?
-------
Appendix D - Review Comments from
Individual CASAC Lead Review Panel Members
This appendix contains the preliminary and/or final written review comments of
the individual members of the Clean Air Scientific Advisory Committee (CASAC) Lead
Review Panel who submitted such comments electronically. The comments are included
here to provide both a full perspective and a range of individual views expressed by
Panel members during the review process. These comments do not represent the views
of the CASAC Lead Review Panel, the CASAC, the EPA Science Advisory Board, or
the EPA itself. The views of the CASAC Lead Review Panel and the CASAC as a
whole are contained in the text of the report to which this appendix is attached. Panelists
providing review comments are listed on the next page, and their individual comments
follow.
D-l
-------
Panelist Page#
Dr. Joshua Cohen D-3
Dr. Deborah Cory-Slechta D-6
Dr. Ellis Cowling D-12
Dr. Bruce Fowler D-19
Dr. Andrew Friedland D-23
Dr. Robert Goyer D-26
Mr. Sean Hays D-34
Dr. Bruce Lanphear D-36
Dr. Samuel Luoma D-57
Dr. Frederick J. Miller D-60
Dr. Paul Mushak D-62
Dr. Michael Newman D-90
Mr. Rich Poirot D-93
Dr. Michael Rabinowitz D-96
Dr. Joel Schwartz D-100
Dr. Frank Speizer D-108
Dr. Ian von Lindern D-110
Dr. Barbara Zielinska D-l 18
D-2
-------
Dr. Joshua Cohen
Comments - Chapter 4
Joshua Cohen
February 18,2006
Section 4.1
The introduction in Section 4.1 is generally good, although there are several passages that are
slightly inaccurate and should be revised. Substantively, the biggest problem with this section is that it
does not clearly explain the advantages and disadvantages of the two categories of modeling -
mechanistic modeling and regression modeling - for quantifying the relationship between environmental
lead levels and lead body burden (in particular, blood lead levels). EPA must develop and present a more
compelling argument the advantages of mechanistic models. One advantage is that mechanistic models
can be adjusted so that they can be used in contexts in different contexts. For example, they can be
adapted to account for differences in lead bioavailability. They can also integrate multiple sources of
exposure (e.g., lead in drinking water AND lead in soil). Those arguments are not made clearly in the
existing text.
• Page 4-1, lines 1+ - The incorrectly asserts that a distinguishing characteristic between
regression and mechanistic models is that regression models "can have relatively few
parameters", whereas mechanistic models include more or all of the parameters needed to
specify a relationship. Regression models can have many or few parameters, and the
same is true of mechanistic models. I think that the main distinction is that regression
models include only those quantities that can be measured and associated with
measurements of the outcome of interest. For example, there are datasets that include
residential soil lead measurements and corresponding child blood lead levels. As a result,
it is possible to regress blood lead levels against soil lead levels to develop a statistical
model. On the other hand, there is are no datasets that include both soil ingestion rates
and blood lead levels. Hence, no regression model can be built relating blood lead levels
and soil ingestion rates. The strength of regression models is that they are empirically
verified (at least within the range of observation). Mechanistic models can incorporate
quantities that have not been measured along with the outcome of interest, making it
possible to characterize the impact of changes in those quantities on the outcome of
interest. On the other hand, they are not directly verified empirically.
• Page 4-2, line 2: The term "Exposure-biokinetic" is incorrect in this context. "Exposure-
intake" would be a better term because you are referring to models that represent
"relationships between levels of lead in environmental media and human lead intakes."
• Page 4-2, line 7 - Likewise, the term "biokinetic model" in this context could be replaced
with "intake-biokinetic model" because you are describing models that characterize the
relationship "between lead intakes and levels of lead in body tissue". EPA can then go
on to say that combining an exposure-intake model and an intake-biokinetic model results
in an exposure-biokinetic model that quantifies the relationship between the amount of
lead in environmental media and resulting lead concentrations in tissue.
• Page 4-2, lines 21-24 - The claim that mechanistic models "integrate complex
information on lead exposure and biokinetics into a form that provides predictions, rather
than just an organized grouping of observations" does not make sense. First, the
statement does not make clear what the comparator is, although it appears to be
D-3
-------
regression models. Second, regression models make predictions, so the statement
appears to be incorrect.
• Page 4-3, line 6 - Consistency - Replace "exposure-biokinetics models" with "exposure-
biokinetic models".
Section 4.2
No comments
Section 4.3
This section does a good job at describing the components of the IEUBK model (Section 4.3.1).
The discussion of the model's calibration and evaluation (Section 4.3.2) is inadequate. Page 4-16, line
30-32 states that the IEUBK model has been evaluated by Hogan et al. (1998). It then goes on to describe
the reasonable agreement that Hogan et al. reported between observed and lEUBK-predicted blood lead
levels. EPA does not mention that another evaluation (Bowers and Mattuck, 2001) that found that "the
IEUBK Model reproduces blood lead levels in children well for some communities, but poorly for others"
(p. 1706). EPA does mention this paper (p. 4-17, line 17) but only in the context of stating that empirical
comparisons have shown that numerous factors influence agreement between the model and observed
blood lead values. EPA does not mention that, at least according to Bowers and Mattuck, these factors
can be so idiosyncratic that it is impossible to account for them unless empirical blood lead measurements
are available for the community in question (p. 1708), something that would obviate use of the model.
Even if the IEUBK model predicts GM blood lead levels without bias, I am concerned about the
way in which the model is used to predict the probability that a child's blood lead level will exceed a
specified level of concern (in particular, 10 ug/dL). Because the model's predictions are imperfectly
correlated with actual blood lead levels, its low predictions will tend to be underestimates, while its high
predictions will tend to be overestimates. It follows that the model will overpredict the probability that
blood lead levels will exceed a specified level of concern at higher levels of environmental lead exposure
(and likewise underpredict this probability at lower levels of environmental lead exposure). Even if the
residual GSD (as estimated by Griffin et. al. (1999b) is reasonably correct), and hence the aggregate
predicted risk of exceeding 10 ug/dL is close to observed values (i.e., summed over all individuals in the
population), the model will tend to overestimate this risk at higher levels of environmental lead exposure
(and underestimate it at lower levels of environmental lead exposure). These discrepancies can have
ramifications for the use of such models to identify appropriate regulatory limits for exposure to lead.
Section 4.4
No comments.
Section 4.5
No comments.
Section 4.6
Page 4-31, line 13 - Does EPA mean "arithmetic average" or geometric mean here?
Please specify the intended modifier.
Page 4-32, lines 26-27 - EPA states that "To the extent that model validation evaluations
have indicated reasonably good matches between IEUBK or Leggett model outputs and
empirical observations, the same can be reasonably expected for the AALM." I do not
see how this claim follows. It is my understanding that the AALM can be thought of as
D-4
-------
using the IEUBK model exposure module together with the Leggett model biokinetics.
Section 4.3 of this chapter describes how the IEUBK model's predictions are consistent
with observed blood lead levels. On p 4-38, EPA notes that the relationship between
predicted blood lead levels and lead uptake is more than twice as steep for the Leggett
model (0.88 ug/dL per ug/day uptake) as it is for the IEUBK model (0.36 ug/dL per
ug/day uptake). That difference (see also Figure 4-13) implies that the blood lead levels
predicted by the AALM will be substantially higher than the blood lead levels predicted
by the IEUBK model. Why does EPA believe the AALM will produce valid blood lead
level estimates given these differences between the Leggett and the heavily tested and
validated IEUBK model?
Section 4.7
No comments.
Section 4.8
No comments.
Section 4.9
Page 4-42, lines 5-7. The suggestion that remaining differences between the major
models are "minor discrepancies" does not seem appropriate. As noted above, the impact
of lead uptake on blood lead levels predicted by the Leggett model (and hence the
AALM) exceeds the corresponding IEUBK model prediction by a factor of two.
Page 4-42, lines 20+. EPA states that "While this magnitude of difference [a factor of
approximately 2] may be substantial in the context of regulatory use of the models (e.g.,
for establishing cleanup goals at hazardous waste sites), it represents a remarkable
convergence of various approaches taken to reduce the complex biokinetics of lead to
tractable, and relatively simple, mathematical expressions." It may in fact be true that
this degree of agreement between the m models is remarkable from a scientific
perspective. However, EPA's purpose in studying and developing these models has been
to aid regulatory efforts. The importance of these differences should therefore not be
downplayed.
Page 4-42, lines 24-28. Given that the Leggett and IEUBK models appear to differ
substantially, it is difficult to see how the AALM model's predictions will simultaneously
converge with those of both of these models.
D-5
-------
Dr. Deborah Cory-Slechta
Air Quality Criteria Document
Chapter 5 Comments
Deborah Cory-Slechta
General Comments
1. Statements about the non-linear effects of Pb should probably be summarized, particularly
with similar effects now being described for IQ and low blood Pbs, e.g.,
p. 5-11, lines 24-26.
pp. 5-20 line 30 through p. 5-21.
p. 5-28, lines 1-9, and figure. Particularly important given the non-linearity emerging
with IQ
p. 5-29, lines 15-17
p. 5-131 lines 18-20 relating to hypertension
The delineation of components in Chapter 5 is peculiar. The experimental literature is followed
by a rather strange configurations of topics, most of which seem to belong more appropriately in
Chapter 6. Particularly confusing is the section on dose-response (p 5-66 onwards) said to be
there to bridge the gap between the findings in 5.3.1 and those to be discussed in section 6. The
presentation of information in Figures 5 and 6 and what is included in each of these two chapters
needs to be reconsidered.
Specific Comments
5-2, lines 20-23 - Logic doesn't follow
5-4, line 4 - Furthermore... was not observed. Seems a non sequitur since accumulation was not
previously referred to.
5-5, line 2 - interfacial?
Line 7-100 ug/100 ml should be changed to |ig/dL as it is other places in the document;
this should be corrected throughout since it is inconsistent. The most standard use, |ig/dL should
be adopted.
5-19, lines 2-3. Why is this approach said to be of little value? Seems a judgment out of context,
or in hindsight, and dismissive in a broad context.
5-19, line 6 - the most reliable evidence? In who's judgment? Why so?
5-19, line 11 - Says this was the significant advance in the field. Again, by who's judgment?
Also contradictory, since there has just been criticism of the low relevance of in vitro
approaches. Not clear how such conclusions were reached; they should be deleted and are not
necessary for the document.
5-20, lines 1-2, drawn the most attention? By what criteria?
D-6
-------
5-22, lines 23-24, it is critical to point out the dangers of measuring a few parameters of a
neurochemical system, finding that they all change, and assuming that they are related. It is
likely that there were changes in other neurotransmitter systems as well that could also have been
correlated with behavioral changes. Unless experiments are explicitly carried out to examine the
nature of the relationship, these interpretations can be misleading.
5-23, lines 1-10. This interpretation seems inconsistent with the findings. Specifically, an
increase in NMDA receptors would seem to suggest that the system sees inadequate agonist and
therefore an up-regulation occurs, while the decrease in MK-801 binding would suggest that the
binding sight sees an excess of antagonist and down-regulates in response to this. Indeed, the
study of Lasley and Gilbert (1996), Gilbert et al. (1996) found decreased stimulated release of
glutamate in hippocampus. There is of course a great danger in trying to engender unitary effects
of Pb; it seems clear that the outcome can change greatly depending upon the timing, level and
chronicity of exposure, and therefore one size fits all may not be realistic, nor should it be
suggested that it should be the case.
5-25, lines 26-27'. The interpretation is not supported by any data; what studies directly show that
nicotinic cholinergic changes play a role in Pb-induced cognitive deficits?
p. 5-26, line 13, why is this one of the 'most significant' advances? Its one more line of evidence,
supporting what has already been shown in a functional capacity.
5-31, line 11, pathology, not function
5-33 through 5-34 really is abbreviated relative to what is described in excruciating detail at the
biochemical level; this seems particularly odd given that the levels of concern are based on
cognitive and behavioral deficits
5-34 through 5-35 same story with rodents; contradictory findings are presented within rodent
studies and contradictory to non-human primate; these are not evaluated in any depth. This
section is missing numerous studies related to the whole area of attention; there are contradictory
findings on sustained attention; the work by Morgan et al. 2001 shows a trivial effect that
required 20 animals per group. The work by Brockel et al fails to support it and in fact
demonstrates the importance of delay of reward. None of this work is included. These studies are
particularly important as they relate to the behavioral mechanisms underlying cognitive deficits
and no doubt relate to the changes observed in IQ.
5-35 through 5-36 Not clear why the work of interactions of Pb and cocaine is described as
producing scientifically important results? This is not at all surprising given the well known
effects of Pb on dopaminergic systems, particularly the mesocorticolimbic system.
5-36 There are numerous other studies using drug discrimination to evaluate dopaminergic
function in Pb-treated rats, none of which are noted or described.
D-7
-------
5-39 lines 12-14 Again, there is no real presentation of the attention work that has been done;
deficiencies in sustained attention are not well demonstrated behaviorally and stronger studies
documenting problems with delay of reward by Brockel et al. are not presented. In general, the
summary on this page does not really capture the understanding from the experimental literature
on behavioral changes.
5-41 to 5-42 mentions the half life of Pb elimination from brain, only citing unpublished
alterations. There are several other published studies, including one that has examined the issue
on a regional basis in brain that are more appropriate to cite here; they provide half-lives that are
longer than that indicated in an unpublished (non-peer reviewed) study.
5-43, summary, lines 1-6 seems a gross overgeneralization of the literature and by no means
anything demonstrated, especially since all of the neurotransmitters exist within the
mesocorticolimbic circuitry that is critical to executive functions.
5-43, lines 18-20 makes little sense. Why wouldn't there be susceptibility factors? For example,
gender, we know there are differences between males and females (e.g., Cory-Slechta et al.,
2004) and these have received no attention in the literature (even though we also know that
males and females exhibit different PbBs). Why wouldn't we expect polymorphisms to interact
with Pb. And certainly period of exposure is a susceptibility factor.
5-43 and prior - there is no mention of the possibility of fetal programming with Pb. Recent
findings by Cory-Slechta et al. (2004) and by Zawia et al. (ref) definitely speak to the possibility
of permanent effects on important systems and proteins need to be identified, particularly as they
invoke an etiological role for Pb in many other disorders and diseases. Also, never really
describes what effect levels are in animals; says 15 and above, but there are studies documenting
effects below 15 and below 10; it is important to note these given the emerging literature on
children and IQ.
5-44 to 5-45 discussion of biomarkers. This seems quite distorted, what are typically known as
susceptibility factors are somehow here renamed as biomarkers of susceptibility. This makes
little sense and isn't consistent with what are typically deemed biomarkers. Also, the issue of
biomarkers as selective or specific to Pb are never mentioned
5-46 lines 7-16; unclear why these changes in peripheral 5-HIAA and HVA are listed as
biomarkers. Not clear what these mean of if they are at all specific to Pb exposure.
5-46, line 18 clinically 'oriented'? Just call these effects. This is a tautology when saying
biomarkers of effect.
5-51 lines 1-23 are really tangential here and reads more like a textbook than a criteria document;
this should be shortened or eliminated.
5-51 lines 24-31 are repetitive of what has already been said.
5-52 lines 1-10, again read like a textbook and add nothing.
D-8
-------
5-53, line 9-10 unclear what is meant here, does it mean that the presumption was that there
should have been a relationship and there wasn't?
5-53 lines 20-21 same as lines 9-10; shouldn't the null hypothesis be no relationship?
5-54, lines 20-22 also give support to an absence of an objective basis of comparison of studies
here; they seem to presume that there has to be a relationship, thus only studies supporting it are
defined as "compelling".
5-58 lines 5-26 describes a study that doesn't even include PbBs and imparts Pb effects to
differential IQs in suburban vs. urban Detroit. This is a highly inappropriate inclusion in this
criteria document and should be deleted.
5-62 lines 4-7 again, a finding that seems to be considered negative since it doesn't show the
positive relationship the author apparently wants to support. It seems critical for a criteria
document in particular to be based on the null hypothesis.
5-64, line 22, halotypes should be haplotypes
5-66; this section seems out of place. It is not critically presented. For example, how was the
calculation of .48uM as the in vitro equivalent of 10 |ig/dL blood Pb determined? States that
almost all of the Pb in a 'neurochemistry experimental system' (whatever that is) can participate
in a reaction.. .this completely ignores precipitation and binding of Pb in these assays.
5-67 lines 1-10; this argument is unclear and not presented in a critical way. The data for a half
life of 2 years for Pb in brain cites a study of modeling and is not consistent with what has been
described in experimental studies. The fact that plasma Pb is not the dose to brain seems to make
little sense. The plasma compartment would remain a source of Pb if blood Pb remains elevated.
What human data can be cited to support the 2 year brain half life?
5-67 lines 21-26 This is total conjecture, unsupported by data and should be deleted; the entire
section on dose-response is hypothetical and should not be included in the criteria document.
5-68 lines 1-11 Again, confusing and unsupported by any substantial data. The section should be
deleted.
5-68 lines 14 through 20; not clear what the point is here. Very confusing sentences.
5-68 through 5-70 Not clear why this section is included here instead of in the chapter on
epidemiology where these effects are described. Moreover, the conclusions that are presented
here are flawed (5-70, lines 9-24). These approaches, while they certainly on a group basis, help
to more specifically delineate the basis of the IQ deficits reported, will not be able to provide a
specific link to lead, i.e., to demonstrate that a lower performance capability can be attributed to
Pb.
D-9
-------
5-73, lines 18-20 again fail to be based on a null hypothesis.
5-75, lines 9-11, again, failure to assume the null hypothesis; any lack of a relationship is always
presumed to be due to some other source of variability.
5-75 and 5-76 on SES; this is really demographic data and should be included in sections of the
document describing demographics; the same can be said for the preceding section on children
and SES. The same can be said for sections on nutritional status since these are really co-
variates in the populations and not describing health outcomes.
5-77 lines 15-17; again, failure to assume the null hypothesis; moreover, the putative explanation
for the absence of an effect of calcium here is inconsistent within the sentence: high lead
burdens, but blood Pbs of 8.5 |ig/dl.
5-86 lines 1-2; need to elaborate the evidence for the adaptation. What is it?
5-86 line 16-17 missing the word "in" after gross changes.
5-88 lines 1-10; blood Pb values of 35-40 |ig/dl can hardly be considered lower levels of
exposure
5-102 lines 13 through 20; the Cory-Slechta paper shows main effects in both males and females
and no interactions, so developmental only Pb exposure produces fetal glucocorticoid
programming.
5-105 through 5-108 conclusions seems like a restatement of all that preceded it rather than
conclusions; perhaps bullet points here would be preferable.
5-111, line 8, change 'administrating' to 'administering'
5-163 lines 13-21 cites the study of Sanchez-Fructuoso et al. (2002) and states that "The authors
emphasized that there was no redistribution to brain. Cory-Slechta et al. (1987) had originally
reported that with CaNa2EDTA chelation in rats Pb is preferentially mobilized from bone and
then redistributed to other organs including brain. The Sanchez-Fructuoso et al. (2002a,b)
findings stand in contrast..." In fact, that is not the case; the redistribution of Pb to brain reported
by Cory-Slechta et al. (1987) occurred in response to a single injection of CaNa2EDTA, and
indeed with further injections of CaNa2EDTA, levels of Pb in brain ultimately declined. The
protocol used by Sanchez-Fructuoso et al. (2002) never examined effects in response to a single
CaNa2EDTA injection. Levels of Pb in brain were only examined after 3 courses of 5
CaNa2EDTA treatments. Thus, this is not a failure to replicate as the text suggests.
Pages 5-178 to 5-183 There is a study by Meja et al. (Neurotoxicology and Teratology, 1997,
vol. 6:489-497) examining combined effects of Pb and As that reports that Pb levels in brain are
increased by this co-exposure.
5-214 line 2, change 'neurological' to 'nervous.'
D-10
-------
5-215 to 5-218 This introduction to the immune system should be deleted; no such section is
presented for any other target organ.
D-ll
-------
Dr. Ellis Cowling
Dr. Ellis Cowling
North Carolina State University
March 10, 2006
Review of the Air Quality Criteria for Lead
(First External Review Draft)
General Comments on the Organization, Format, and Content of Air Quality Criteria
Documents and their Relationship to Staff Papers
This past year has provided an unusual opportunity for CASAC, NCEA, and OAQPS to work
together in efforts to further optimize the design and organization of Air Quality Criteria
Documents and Staff Papers. During this one year, in rather rapid succession, CASAC has
reviewed both planning documents, and review drafts of both criteria documents and staff papers
for three of the five pollutants presently recognized criteria pollutants. In each case, CASAC has
been presented with very large documents that require very careful attention from the standpoint
of many different scientific disciplines in order to summarize the present state of scientific
understanding about:
1) the chemistry and physics of the pollutant itself,
2) the sources of air emissions of the pollutant or its precursors,
3) the transport, transformation, and atmospheric deposition processes by which the
pollutant is delivered to sensitive receptors,
4) the nature and magnitude of the effects of the pollutant on both human health and on
human welfare, and
5) the establishment of science- based national ambient air quality standards that in the
judgment of the Administrator of EPA will be useful and effective in decreasing
exposures and therefore decreasing the magnitude and prevalence of adverse effects on
both human health and human welfare with an "adequate margin of safety" at lest in the
case of effects on public health, and finally
6) the continuously evolving historical development of both scientific understanding about
all five of these preceding aspects of the pollutant, its health and environmental effects,
and the art and practice of its management over time.
The laws of our country require that this difficult and challenging intellectual work should be
accomplished periodically (ideally every five years) by scientists, engineers, policy analysts, and
decision makers who are charged by our society to do their respective parts — leading to
scientifically sound, policy effective, and socially acceptable decisions in a contentious
democratic society that often is resistant to change and frequently uses the courts of our country
to set demanding deadlines for the development of Criteria Documents, Staff Papers, and the
promulgation and implementation of Regulations and Rules for air quality management.
During the past year CASAC Members and Panelists have reviewed and offered our
carefully considered individual and collective advice and counsel about the adequacy of the
D-12
-------
criteria documents, staff papers, and the proposed rules and regulations for ozone and other
photochemical oxidants, fine and coarse particulate matter, and now lead.
In all three of these cases, CAS AC has done its best to review the documents prepared by
NCEA and OAQPS and to offer our individual and collective counsel and advice about the
scientific content, organization, and the scientific objectivity and tone of impartiality of these
very large criteria documents.
Beginning in the case of the Criteria Document and Staff Paper for Ozone and Related
Photochemical Oxidants, a somewhat different organizational structure was used by NCEA.
The new organizational format called for relatively brief Main Chapters that consist of two
parts:
1) a concise summary of "Key findings/conclusions" from earlier assessment documents,
and
2) carefully prepared descriptions of advances in scientific understanding that have been
developed since the time of the last review and published in more recent scientific
literature.
The new structure also calls for development of very detailed Annexes for each Main
Chapter in which many important advances in scientific understanding are presented in much
more thorough fashion than in the corresponding Main Chapter.
The final features of the new structure and organization of Air Quality Criteria Documents
were development of both an Integrative Summary Chapter and an Executive Summary for the
whole Criteria Document. The purpose of these two additional parts of the Criteria Document
was to draw together the major findings and conclusions of scientific understanding developed
within each of the Main Chapters and corresponding Annexes and to to present in an integrative
way the Key Findings and Conclusions (from both earlier assessment reports and the most recent
description of scientific advances) and thus provide a maximally useful foundation for the Staff
Paper.
In the words of OAQPS, the purpose of the Staff Paper is to:
"provide a critical assessment of the latest available scientific information upon which the
National Ambient Air Quality Standards are to be based. Drawing upon the AQCD, staff in
EPA's Office of Air Quality planning and Standards (OAQPS) within the Office of Air and
Radiation prepares a Staff Paper that evaluates policy implications of the key studies and
scientific information contained in the AQCD and presents the conclusions and
recommendations of the staff for standard setting options for the EPA Administrator to
consider. The Staff Paper is intended to 'bridge the gap' between the scientific assessments
contained in the AQCD and the judgments required of the Administrator in determining
whether it is appropriate to retain or to revise the primary and secondary NAAQS."
Many members of CASAC were very pleased with the good sense of the revised structure
and format of Criteria Documents. We are convinced that these innovations in the overarching
method of organization of Criteria Documents will better serve the interests of the wide variety
of audiences that are interested to learn more about scientific understanding of each of the
criteria pollutants and their effects on both human health and welfare. Thus, many of us believe
that these innovations in structure should be retained and used not only in preparing the Second
D-13
-------
External Review Draft of the Criteria Document for Lead but also in preparing other Criteria
Documents for other Criteria Pollutants.
In doing so, it is of course important that the different target audiences for the Executive
Summary, the Main Chapters of the Criteria Document itself, and the various Annexes be very
well defined and well understood by all of the staff, consultants, and editors that prepare these
three different treatments of the same body of scientific knowledge.
It is even more imperative that the scientific content, objectivity, and tone of impartiality of
the Executive Summary and the Integrative Summary Chapter of the Criteria Document [and the
Staff Paper as well!] be consistent not only with the scientific content, objectivity, and tone of
impartiality of the Main Chapters of the Criteria Document itself, but also with the scientific
content and objectivity of the more detailed Annexes. Differences in content of these different
parts of the same Criteria Document [and the related parts of the Staff Paper] should be based
primarily on their relevancy to their respective purposes and target audiences. Discrepancies in
scientific content, objectivity, and tone of impartiality in these distinct parts of the Criteria
Document and Staff Paper will inevitably lead to decreased confidence in the validity and
reliability of the different parts of both types of documents. Thus such discrepancies must be
carefully avoided. This will require a larger degree of common understanding among authors,
consultants, editors, and managers of the Criteria Document and Staff Paper development
processes than many members and Panelists within CASAC believe has been achieved to date.
A useful mechanism for ensuring that there is an effective and concise summary of "Key
Findings and Conclusions" in each Main Chapter is to require that an Executive Summary be
prepared for each Main Chapter and that these statement of Key Findings and Conclusions from
individual Main Chapters be used in constructing both the Executive Summary for the whole
Criteria Document and in developing the organizational framework for the Integrative Summary
Chapter.
One suggestion for avoiding discrepancies in communication among these several different
parts of Criteria Documents is to require that the same carefully-crafted summary statements of
scientific findings and conclusions are not only included (but also printed in bold-face type)
within all parts of complex scientific assessment documents. This editorial device is used in
many high-quality National Research Council assessment reports that also deal with very
complex policy relevant scientific issues.
In written comments on the Criteria Document for Ozone and Other Photochemical Oxidants
dated December 2, 2005 I recommended [and affirm here once again] that all authors,
consultants, editors, and managers engaged in the preparation of Criteria Documents and EPA
Staff Papers take full advantage of- and use the attached published "Guidelines for the
Formulation of Statements of Scientific Findings to be Used for Policy Purposes."
These guidelines, written in the form of checklist questions, were developed by the members
of the Oversight Review Board (ORB) of the National Acid Precipitation Assessment Program to
assist scientists, engineers, and policy analysts dealing with other environmental research and
assessment programs in formulating statements of scientific findings to be used in policy
decision processes. The distinguished members of the ORB who prepared these guidelines
included: Milton Russell, former Assistant Administrator for EPA, Chauncey Starr, former
Director of Research for the Electric Power Research Institute (EPRI), Tom Mai one, former
D-14
-------
Foreign Secretary for the National Academy of Sciences, John Tukey, Distinguished Professor of
Statistics at Princeton University, and Kenneth Starr, Nobel Prize Winner in Economics.
D-15
-------
GUIDELINES FOR FORMULATION OF STATEMENTS OF SCIENTIFIC FINDINGS
TO BE USED FOR POLICY PURPOSES
The following guidelines in the form of checklist questions were developed by the NAPAP Oversight
Review Board to assist scientists in formulating presentations of research results to be used in policy decision
processes.
1) IS THE STATEMENT SOUND? Have the central issues been clearly identified? Does each statement contain
the distilled essence of present scientific and technical understanding of the phenomenon or process to which it
applies? Is the statement consistent with all relevant evidence that is available in the published literature. Is the
statement contradicted by any important evidence in the published literature? Have apparent contradictions or
interpretations of available evidence been considered in formulating the statement of principal findings?
2) IS THE STATEMENT DIRECTIONAL AND, WHERE APPROPRIATE, QUANTITATIVE? Does the
statement correctly quantify both the direction and magnitude of trends and relationships in the phenomenon or
process to which the statement is relevant? When possible, is a range of uncertainty given for each quantitative
result? Have various sources of uncertainty been identified and quantified, for example, does the statement
include or acknowledge errors in actual measurements, standard errors of estimate, possible biases in the
availability of data, extrapolation of results beyond the mathematical, geographical, or temporal relevancy of
available information, etc. In short, are there numbers in the statement? Are the numbers correct? Are the
numbers re levant to the general meaning of the statement?
3) IS THE DEGREE OF CERTAINTY OR UNCERTAINTY OF THE STATEMENT INDICATED
CLEARLY? Have appropriate statistical tests been applied to the data used in drawing the conclusion set forth
in the statement? If the statement is based on a mathematical or novel conceptual model, has the model or
concept been validated? Does the statement describe the model or concept on which it is based and the degree
of validity of that model or concept?
4) IS THE STATEMENT CORRECT WITHOUT QUALIFICATION? Are there limitations of time, space, or
other special circumstances in which the statement is true? If the statement is true only in some circumstances,
are these limitations described adequately and briefly?
5) IS THE STATEMENT CLEAR AND UNAMBIGUOUS? Are the words and phrases used in the statement
understandable by the decision makers of our society? Is the statement free of specialized jargon? Will too
many people misunderstand its meaning?
6) IS THE STATEMENT AS CONCISE AS IT CAN BE MADE WITHOUT RISK OF
MISUNDERSTANDING? Are there any excess words, phrases, or ideas in the statement which are not
necessary to communicate the meaning of the statement? Are there so many caveats in the statement that the
statement itself is trivial, confusing, or ambiguous?
7) IS THE STATEMENT FREE OF SCIENTIFIC OR OTHER BIASES OR IMPLICATIONS OF
SOCIETAL VALUE JUDGMENTS? Is the statement free of influence by specific schools of scientific
thought? Is the statement also free of words, phrases, or concepts that have political, economic, ideological,
religious, moral, or other personal-, agency-, or organization-specific values, overtones, or implications? Does
the choice of how the statement is expressed rather than its specific words suggest underlying biases or value
judgments? Is the tone impartial and free of special pleading? If societal value judgments have been discussed,
have these judgments been identified as such and described both clearly and objectively?
8) HAVE SOCIETAL IMPLICATIONS BEEN DESCRIBED OBJECTIVELY? Consideration of alternative
courses of action and their consequences inherently involves judgments of their feasibility and the importance of
effects. For this reason, it is important to ask if a reasonable range of alternative policies or courses of action
have been evaluated? Have societal implications of alternative courses of action been stated in the following
general form?:
"If this [particular option] were adopted then that [particular outcome] would be expected."
9) HAVE THE PROFESSIONAL BIASES OF AUTHORS AND REVIEWERS BEEN DESCRIBED
OPENLY? Acknowledgment of potential sources of bias is important so that readers can judge for themselves
the credibility of reports and assessments.
D-16
-------
More Specific Comments:
Target Audiences
During the CAS AC meeting on the First External Review Draft of the Criteria Document on
Lead on February 28-March 1, several Members and Panelists mentioned that the original 1977
Criteria Document on lead, the 1986 updated revision and its accompanying Addendum, and the
1990 Supplement to the Criteria Document on lead had been valuable sources of scientific
background information (and even for inspiration and definition of career goals) for scientists
and engineers in years past. Thus, in addition to the most immediate value of Criteria
Documents and Staff Papers as background for decisions by the Administrator of EPA, there are
a number of other target audiences which have in the past (and no doubt also in the future) will
profit from the rigorous scientific reviewing and evaluation that is accomplished by these
documents.
Thus, graduate students and post-doctoral fellows in universities, in other federal and state
offices, leaders in industry, leaders in public interest groups and trade organizations, teachers of
courses in universities, and members of the public at large should be recognized as target
audiences. This wide array of target audiences should be recognized and borne in mind by all
the authors, consultants, editors, and managers involved in the design, organization, preparation,
evaluation, and response to reviewer comments concerning the information contained in both
Criteria Documents and Staff Papers.
Multi-Media Nature of Lead
More than any other of the five Criteria Pollutants which CAS AC has been charged to review
in recent years, lead crosses more if not all of the "media of concern" to USEPA. These multi-
media aspects include: 1) air emissions and deposition of lead from transportation vehicles,
metal smelters, battery production and recycling facilities, 2) lead content of drinking water, 3)
lead containing paints, 4) lead containing pesticides involved in food production, 5) soil
contamination with lead, 6) lead in municipal and solid waste management, 7) lead
contamination of superfund sites, etc. The multi-media nature of this pollutant is touched upon
in several different parts of this the First External Review Draft. It may be worthwhile to draw
these multi-media aspects of lead together in a single part of the Criteria Document, probably in
Chapter 1.
Content and Placement of the Integrative Summary Chapter
Lack of an Integrative Summary Chapter and a comprehensive Executive Summary for the
whole Criteria Document is a major shortcoming of the First External Review Draft. Although
the original intent was to prepare an Integrative Summary Chapter (Chapter 7) that would deal
only with effects of lead on public health, (and thus to include information only from Chapters 1-
6), it clearly would be much more advantageous and appropriate for the Integrative Summary
Chapter also to include Environmental Effects of Lead (Chapter 8) as well as effects of lead on
public health.
This is especially desirable because it is historically deposited lead that is the principal object
of current concern rather than current "ambient" air concentrations. Very substantial decreases
in air concentrations and atmospheric deposition of lead into the environment have been
achieved in recent decades. Thus, most current exposures of both people and living organisms in
natural and managed ecosystems are caused primarily by redistribution of environmentally
D-17
-------
persistent airborne lead compounds deposited in soils, sediments, and surface waters during
earlier decades of the present century.
This more broad perspective can be achieved by reversing the chapter numbers for chapters 7
and 8 so that the Integrative Summary Chapter comes at the end of the Criteria Document and
provides an integrative summary of both health and environmental effects of lead. A similarly
broad perspective will be desirable in the design and content of the Executive Summary of the
whole Criteria Document for lead.
Response to Specific Issues in Chapter 8
It was good to learn from Lester Grant's transmittal memo dated February 15, 2006 that the
intent of Section 8.1.1 and 8.2.1 is to serve as the main body of the terrestrial effects and aquatic
effects portions of Chapter 8, respectively, while the other sections (8.1.2 through 8.1.6 and 8.2.1
through 8.2.6) will ultimately serve as annexes to chapter 8, similar to the format used for the
Criteria document on Ozone and Related Photochemical Oxidants. In this way the redundancy
between Section 8.1.1 and Sections 8.1.2 through 8.1.6 regarding terrestrial effects, and between
Section 8.2.1 and Sections 8.2.2 through 8.2.6 regarding aquatic effects will be resolved in the
Second External Review Draft for Lead.
Charge Question H3 - Discussion of the Concepts of Critical Loads
Many of us were especially pleased to see the relatively thorough discussion at the end of
Chapter 8 regarding the alternative concepts of critical loads, critical limits, target loads, and
target times that have been developed in European and Canadian scientific literature to guide the
processes of decision making regarding both environmental and public health effects of airborne
chemicals. Although these alternative concepts and processes of analysis of multiple
pollutant/multiple effects have not been carefully considered for use in the United States, we
believe, together with the authors of the National Research Council/National Academy of
Sciences 2004 report on "Air Quality Management in the United States," that these alternatives
should be considered very carefully as air quality management tools for use in this country as
well.
D-18
-------
Dr. Bruce Fowler
Bruce Fowler 3/10/06
Post - Meeting comments on EPA DRAFT CASAC and response to Chapter 5 Charge Questions
1. First - I would like to reaffirm that my pre- meeting comments regarding the need for
strong editorial assistance in Chapter 5 with regard to organization and the need to put the
relevant references on the brain and kidney lead binding proteins into their respective
sections near the front of the chapter. These references are generally well captured and
discussed in Section 5.11 but missing or covered in passing in the brain and kidney
sections in those earlier sections. This is really an editorial matter.
2. If the recent work by Drs. Michael Waalkes / Robert Goyer and colleagues at NIEHS on
MT and alpha Synuclein is published or in press by the time this EPA document is ready
for publication, that should also be cited. The recent data presented at the SOT meeting
seem to indicate that MT may be interacting with this protein as well.
3. I note that there were some references in the document to non-mammalian systems so I
offer the following 2 references regarding lead-binding proteins in catfish and altered
susceptibility to lead inhibition of liver ALAD in this species for completeness.
Conner EA, Fowler BA. Preliminary purification and partial characterization studies of a
low-molecular weight cytosolic lead-binding protein in liver of the channel catfish
(Ictalurus punctatus). Aquatic Toxicology 28: 29-36, 1994.
Conner EA, Fowler BA. Biological and immunological properties of Fish Hepatic 5-
aminolevulinic acid dehydratase (Porphobilinogen synthetase). Aquatic Toxicology
28:37-52, 1994.
4. Interactions between lead and other toxic elements such as arsenic and cadmium. As
discussed in the meeting, there are published data on additive interactions among lead
cadmium and arsenic from Mahaffey and co-workers during the 1977-1981 time period,
and I can forward these if needed. The more recent studies (in vivo and in vitro) from my
lab group are preparation and will hopefully be submitted soon. I will forward them as
they are accepted or in press. It is my opinion that interaction among these elements is
important since they frequently occur together in Superfund sites and in aerosols from
smelting or coal- fired power plants. The literature is limited so it should not be an
onerous task to have an up to date summary in a short time period.
5. I agree with the other suggestions offered by Dr. Goyer.
D-19
-------
Bruce Fowler Responses to Charge Question El
El a. I would suggest adding some references to lead binding proteins in brain since they are
found in both animals and humans and appear to play an important intracellular role in mediating
low dose lead bioavailability to other sensitive molecular processes in brain. They may also help
to explain inter -individual differences in sensitivity to lead neurotoxicity. Some suggested
references are as follows:
1. Oskarsson A, Squibb KS, Fowler BA. Intracellular binding of lead in the kidney: Partial isolation
and characterization of post-mitochondrial supernatant lead-binding components. Biochem Biophys
Res Commun 104:290-298, 1982.
2. Goering PL, Mistry P, Fowler BA. A high affinity lead-binding protein in brain attenuates lead
inhibition of 5-aminolevulinic acid dehydratase: Comparison with a renal lead-binding protein. J
Pharmacol Exper Therap 237:220-225, 1986.
3. DuVal GE, Fowler BA. Preliminary purification and characterization studies of a low molecular
weight high affinity cytosolic lead-binding protein in rat brain. Biochem Biophys Res Comm
159:177-184, 1989.
4. Quintanilla-Vega B, Smith DR, Kahng MW, Hernandez JM, Albores A, Fowler BA. Lead-binding
proteins in brain tissue of environmentally-lead exposed humans. Chem Biol Interact 98:193-209,
1995.
Elb. There is an extensive literature on bone and other calcified tissues as storage sites for lead
but the basic scientific literature regarding permanent impact of lead on bone development is
more limited. The work of JE Puzas and colleagues at the University of Rochester is perhaps the
most recent relevant in this regard. Other suggested references regarding lead storage in bone
and molecular effects are given below.
Suggested possible references:
1. Sauk J, Smith T, Silbergeld EK, Fowler BA, Somerman MJ. Lead inhibits secretion of
ostenectin/SPARC without significantly altering collagen or Hsp47 production in osteoblast-like
ROS 17/2/8 cells. Toxicol and Appl Pharmacol 116:240-247, 1992.
2. Todd AC, McNeill FE, Fowler BA. In vivo X-Ray fluorescence of lead in bone. Environmental
Research 59: 326-335, 1992.
3. Silbergeld EK, Sauk J, Somerman M, Todd A, McNeil F, Fowler B, Fontaine A, van Buren J. Lead
in bone: storage site, exposure source, and target organ. Neurotox 14(2-3):225-36, Summer-Fall
1993.
4. McNeill FE, Todd AC, Fowler BA, Laughlin NK. The in vivo measurement of bone lead stores by
109Cd K X-ray fluorescence in a non-human primate (Macaca mulatta). Basic Life Sci. 60:315-8,
1993.
D-20
-------
5. McNeill FE, Laughlin NK, Todd AC, Sonawane BR, Van DeWal KM, Fowler BA.
Geriatric bone lead metabolism in a female non-human primate population. Environ
Research 72:131-139, 1997.
Elc. The cited animal studies regarding chelation / intervention are relevant to analogous studies
in humans since similar biochemical/molecular principles are operating. The question of
relevance may also be considered from the perspective of risk management for humans exposed
to lead and if this is a concern for EPA, then it is appropriate to cover them in this document.
Eld. The newer findings regarding the interactions of lead with specific molecules, lead binding,
transport kinetics and nucleotide pools are certainly useful information but more work is needed
to in order to make the linkage to mechanisms underlying specific health endpoints. These
parameters represent a number of possible molecular events that are in operation following
exposure to lead. There are other factors such as changes in gene expression patterns and other
compensatory mechanisms which may be of particular importance at low dose exposure levels.
Some possible references for consideration relating to lead interactions with specific target
molecules are listed below. These are offered as suggestions only.
Suggested possible references:
1. Goering PL, Fowler BA. Regulation of lead inhibition of 5-aminolevulinic dehydratase by a high
affinity renal lead-binding protein . J Pharmacol Exp Therap 231:66-71, 1984.
2. Victery WW, Miller CR, Fowler BA. Lead accumulation by rat renal brush border membrane
vesicles. J Pharmacol Exp Therap 231:589-596, 1984.
3. Mistry P, Lucier GW, Fowler BA. High affinity lead-binding proteins from rat kidney cytosol:
Mediate cell-free nuclear translocation of lead. J Pharmacol Exp Therap 232:462-469, 1985.
4. Goering PL, Fowler BA. Mechanisms of renal lead-binding protein protection against lead-
inhibition of 5-aminolevulinic acid dehydratase. J Pharmacol Exp Therap 234:365-371, 1985.
5. Oskarsson A, Fowler BA. Effects of lead inclusion on subcellular distribution of lead in rat kidney:
The relationship to mitochondrial function. Exper Molec Pathol 43:409-417, 1985.
6. Oskarsson A, Fowler BA. Effects of lead on the heme biosynthetic pathway in rat kidney. Exper
Molec Pathol 43:397-408, 1985.
7. Mistry P, Mastri C, Fowler BA. Influence of metal ions on renal cytosolic lead-binding proteins and
nuclear uptake of lead in the kidney. Biochem Pharmacol 35:711-713, 1986.
8. Goering PL, Mistry P, Fowler BA. A high affinity lead-binding protein in brain attenuates lead
inhibition of 5-aminolevulinic acid dehydratase: Comparison with a renal lead-binding protein. J
Pharmacol Exper Therap 237:220-225, 1986.
9. Goering PL, Fowler BA. Mechanism of kidney metallothionein reversal of lead inhibition
of 5-aminolevulinic acid dehydratase. Arch Biochem Biophys 253:48-55, 1987.
D-21
-------
10. Goering PL, Fowler BA. Metal constitution of metallothionein influences inhibition of 5-
aminolevulinic acid dehydratase (porphobilinogen synthase) by lead. Biochem J 245:339-345,
1987.
11. Fowler BA, Kahng MW, Smith DR, Conner EA, Laughlin NK. Implications of lead- binding
proteins for risk assessment of lead exposure. J Exposure Analysis and Environ Epidemiol 3: 441-
448, 1993.
12. Smith DR, Kahng MW, Quintanilla-Vega B, Fowler BA. High affinity renal lead-binding proteins
in environmentally exposed humans. Chem Biol Interact 115:39-52, 1998.
E.le. The oxidative stress theory does represent a plausible general mechanism of action that is
likely to occur across organs and species. The degree to which it may have an impact will depend
upon a number of other factors such as anti-oxidative stress mechanisms (e.g., GSH,
metallothionein), cellular repair mechanisms such as the stress proteins, duration of exposure and
nutritional status and concomitant exposure to other oxidative stress inducing agents such as
arsenic and cadmium. Does EPA wish to take up the issue of interactions with other toxic metal
s/metalloids commonly found with lead in this document as well as compensatory mechanisms
against oxidative stress? Please see comments and reference below on oxidative stress from
combined exposures.
Elf. The issue of animal - human and apparent dose differences between species is a long-
standing concern for the risk assessment of many chemicals. Perhaps a better way to look at the
problem is to consider dosages at the molecular or target cell levels of biological organization.
Dosage at the target cell level may be more relevant to risk than trying to use administered or
intact organism exposure levels for such purposes. As scientific understanding of what doses of
lead, in this case, cause biological disruption of critical target or cellular pathways increases
confidence in such an approach would also increase. The potential confounding scientific
variables then become changes in lead kinetics as a function of dose and time as well as the
influence of compensatory molecular mechanisms which appear to be operational at low dose
exposure levels. The issue of populations at special risk as a function of age, gender, nutritional
status and genetic predisposition would also complicate the selection of a specific cut -off value.
A probabilistic approach may prove to be more satisfactory in the long run for estimating a cut-
off range.
One suggested reference regarding exposure to lead, cadmium and arsenic at LOEL doses
levels that may be useful to this discussion is given below. A series of full papers with both in
vitro and in vivo studies is in preparation by my former students.
1. Fowler BA, Whittaker MH, Lipsky M, Wang G, Chen XQ. Oxidative stress induced by lead,
cadmium and arsenic mixtures: 30-day, 90-day, and 180-day drinking water studies in rats: an
overview. Biometals 17(5): 567-8, 2004.
D-22
-------
Dr. Andrew Friedland
Andy Friedland, Dartmouth College
26 February 2006
Preliminary Comments on First External Review Draft Lead Air Quality Criteria Document
(Dated December 2005) with a specific focus on Chapter 8: Environmental Effects of Lead
Charge Question Al:
The document format is certainly useful. The subject of historical trends in atmospheric
emissions, and the history of deposition of Pb at specific locations over time occur repeatedly
throughout the document. Within Chapter 8, there are a number of locations where atmospheric
deposition is discussed and the history of the adding of alkyl-Pb and then elimination of
additives is described. Chapter 2 also contains some discussion of sources of Pb and subsequent
atmospheric transport. I have not yet read other chapters closely, but it appears that history of Pb
deposition occurs elsewhere in the document. It would be useful to discuss as a group the
clearest, most consistent and most efficient way to present historical trends in atmospheric
emissions and deposition throughout the document.
A discussion of the intended purpose of the Integrative Synthesis (Chapter 7) and the choice for
its location would be useful. From the 15 February 2006 Memorandum from Lester Grant to
Fred Butterfield, I presume this is an Integrative Synthesis of health related topics only; if so, this
should be stated clearly in the chapter title in the Table of Contents. As it is listed now, the
chapter appears to be a synthesis of the entire subject and if that is the case, it is unclear why it
occurs before Chapter 8.
Charge Question HI:
Yes the subject section adequately covers the most current and most important information on
the measurement methods, distribution and effects of Pb on terrestrial ecosystems. All major
bodies of work have been included. There can be better organization of the material, and some
inconsistencies can be removed. These minor weaknesses of the document may be the result of a
multi-author team, or perhaps they reflect an organizational structure not immediately clear to
me. I look forward to discussing this subject at the meeting.
The authors make it abundantly clear that the reduced use of Pb additives in gasoline has
decreased substantially the atmospheric deposition of Pb in the US since the mid-1970s.
However, information related to the relative role of other sources of Pb is inconsistent in Chapter
8. Page 8-1 line 11 lists waste incineration before the combustion of fossil fuel and metal
smelting and production, yet it is not clear if this is listed in order of importance, today or
historically, or if the list is random. The same ordering appears on page 8-35 (lines 30-33) and
lists the same references (in the opposite order). In this instance, fuel combustion is not
included. Certainly, even without gasoline additive Pb emissions, the natural occurrence of Pb in
coal and petroleum products other than natural gas should be mentioned. Page 8-47, lines 21-24
D-23
-------
include the metal production industry and the combustion of fossil fuels but do not mention
waste incineration.
Charge Question H2:
Yes the subject section adequately covers the most current and most important information on
the measurement methods, distribution and effects of Pb on aquatic ecosystems.
The following comment is relevant for both HI and H2: Multiple contaminants, interactions
with other pollutants, chemical mixtures including synergistic effects of Pb plus other metals are
discussed in multiple locations and in different ways throughout the terrestrial and aquatic
ecosystem sections. There are also discussions on this topic elsewhere in the document. This is
another area where consistency and uniformity, when appropriate, at least throughout Chapter 8,
would be beneficial. The Metal Assessment Panel of the Science Advisory Board of the EPA
addressed the issue in its two reports of 2003 and 2005. It might be useful for the thought
process from that group on how to address "mixtures" in terrestrial and aquatic ecosystems and
in human systems to be communicated to the authors of this document.
The specific case study illustrated in the aquatic section 8.2.6.2 is useful and effective. It would
be valuable to discuss if the case study should be expanded and whether or not a parallel
treatment of a case study in the terrestrial section would be beneficial.
Charge Question H3:
I believe that the subject section does contain the most current information on the potential use of
critical loads. This is a difficult question to answer, and a difficult topic to write about because
there are many fewer publications on critical load analysis for metals than there are critical load
analysis publications for sulfur, nitrogen and hydrogen ion. Furthermore, the most important
document referred to in this section is a paper I was not aware of previously, DeVries et al.
(2004), which is not a peer-reviewed document and appears to be available only from a website.
Perhaps the panel could discuss the paucity of information on this subject and the apparent lack
of any critical loads analysis literature for Pb in the United States, and how to best respond to
this relative lack of information. I believe that critical load analysis in general—not even
specifically critical load analysis for metals—is relied upon much less in the USA than in Europe
and perhaps Canada. Perhaps this too could be a topic for discussion among panelists.
Specific items needing clarification or elaboration:
Page 8-1 lines 19-20. The statement "Pb leached into mineral soil appears to be 20%-50% of
total anthropogenic Pb deposition" needs a reference and needs elaboration. At a minimum, it
should refer to a specific location or region.
D-24
-------
Page 8-18, Figure 8-1.2.1 Relationship of bioaccessibility versus speciation. This figure needs
units and needs a better description. Is this an illustration or should the x and y axes confer a
scale and directionality?
Page 8-80, Figure 8-1.5.1 Avian toxicity data.... This figure needs units or an indication of
directionality on the x axis.
Page 114, lines 4-6. Other studies in this section are described and the percentage reduction that
occurred over time is presented. In the discussion of Evans et al. (2005), the percentage
reductions are not presented. Why not?
D-25
-------
Dr. Robert Goyer
February 24,2006
From: Robert A. Goyer
To Rogene Henderson, CASAC Chair; Fred Butterfield, CASAC Federal Officer
Comments in response to charge questions E 1
I am not aware of new animal studies not included in the CASAC draft but have some comments
regarding sub-questions as follows.
Ela. Neurotoxicology
I am not able to provide comments on mechanisms of neurotoxicology, section 5.3.1 but do have
comments in regard to organization and continuity for other parts of Section 5.3.
The section (not numbered) beginning on p. 5-66 titled Dose-response paradigms serves to
connect neurological toxicities to the next chapter, clinical effects, in section 6.3.1 suggest this
Dose Response discussion is a good bridge to the clinical chapter and might directly follow
section 5.3.1. The discussion in Dose-Response Paradigms also addresses the Charge Question
Elf
The remainder of section 5.3 might be reorganized to provide better emphasis on the
toxicological basis for vulnerable populations and susceptibility. Reasons and rationale for
reorganizing the remainder of section 5.3 are contained in the following comments.
The third bullet, p. 5-43 Integration of research findings questions the rationale for studying
susceptibility factors in animals
There are compelling reasons for understanding susceptibility factors. It is true that susceptibility
of humans to lead toxicity is difficult to study in experimental models but much has been learned
about effects of nutrition, and age on susceptibility from animal studies that has lead to further
studies in humans. Factors affecting susceptibility may explain differences in health (toxic)
effects that might be observed in different people with comparable levels of exposure and/or
biomarkers of exposure and perhaps differences in responses among different age groups. Also,
it may be possible to create genetic models in animals that imitate human polymorphisms. To say
that there is a "lack of compelling rationale for their investigation" (in laboratory animals) is
likely to discourage further research.
Susceptibility factors e.g., nutrition, polymorphisms etc. are well stated in p- 5-51, para. 2 lines
13-23 but are not highlighted as such.
D-26
-------
Section 5.3.2 concerns effects of lead in at different ages and the influence of various
susceptibility factors. The organization of this Section (5.3.2 p. 5-43 line 22 etc) is difficult to
follow largely because of dividing the discussion into major groups based on age e.g.
5.3.2.1 (p.5-44) Children and adolescents, 5.3.2.2 (p 5-70), Adults with childhood lead poisoning
and 5.3.2.3 (p 5-73), Adults with ambient exposures. This organization may be meant to parallel
epidemiological studies addressed in Chapter 6. However, many of the subsections in this part of
the chapter concern susceptibility factors, e.g. effects of age, nutrition , polymorphisms. This
results in repetitions/redundancies in some topics.
The following is a suggested format for reorganization of sections 5.3.2 etc.
Discussion of biomarkers of exposure/effect is fine in introduction.
Susceptibility Factors might be highlighted as a major subsection including Age, SES, Nutrition
and polymorphisms (currently two nutrition and polymorphism sections) are under children and
adults. This new section might end with a series of conclusions regarding susceptibility factors. It
should also reference related discussion of role of other metals on Pb distribution beginning in
section 5-7, p. 5-178.
Biochemical biomarker discussion in adult section p5-74, 75 really identifies homocystine as a
susceptibility factor and might be considered as such or might even be included as a
polymorphism.
P 5-75 para. 2, line 18 begins a section titled, "Vulnerability and Susceptibility" which includes
discussion of SES and nutrition, (redundancy). This section might be an introduction to a major
section on Susceptibility.
P 5-78, section titled Neurotoxicology of Lead is really a discussion of bone lead as a biomarker
and might be included with earlier discussion of other biomarkers.
Other sections in 5-3 on neurotoxicology including Neuro Epidemiological studies, p 5-68,
Clinical aspects of Adult lead poisoning, p.5-70 appropriately precede the next chapter on
epidemiology but are concerned with relationships between experimental studies and clinical
indications of lead health effects. These sections currently are not followed by a set of
conclusions.
Question Elb. Bone and teeth as internal pools. No comment.
Question Elc. Relevance of study of chelation of lead in animals to humans
Section 5.10.1.4 provides a detailed account of the effects of chelating agents on oxidative stress
in the liver. A broader discussion of effects of chelating agents on toxicity in other organ
systems, particularly the CNS, would provide information on changes in tissue/cellular content
of lead and effects on mechanisms (neural transmitters). Such information provides background
for the selection and potential role of chelating agents in management of lead exposure in
D-27
-------
humans. The present draft contains very little about experimental studies of effects of chelating
agents, only as oxidants (Section 5.2 re: as erythrocyte antioxidants, also section 5.10.1.4 Effects
of Chelation on ROS p 5-p266).
Question Eld. Do insights gained on Pb-induced alterations in erythrocyte biology do
provide information regarding molecular mechanisms of action? Yes.
Section 5-1 reviews effects of lead on erythrocyte biology and function, heme metabolism and
erythrocyte enzymes. The section is clearly written and provides a detailed summary of current
information of lead in the red blood cell. Many of the heme intermediates and enzymes are used
as biomarkers for assessment of both lead exposure and potential health effects so mechanisms
of action of lead on these molecular systems is important in understanding the significance of
changes in red blood cell metabolism. As to whether they are suggestive of mechanisms
underlying specific health points I believe the answer is yes. For example, interactions of Pb and
Ca on membrane transport are likely to be similar to effects on transport in kidney tubule cells,
hepatic cells and possibly cells in the CNS.
Question Ele. Is the oxidative stress theory plausible for Pb toxicity?
The answer is yes
Oxidative stress is likely as a common mode of action operating across organs and species. It is
cited and discussed as a mode of action in various organ systems in this chapter. Oxidative stress
has been invoke and a common mode of action for toxicity of other metals as well as Pb so that it
is non specific and other more specific modes of action must be present in different organ
systems to explain differences in effects between organs and between different metals.. It is
likely a common mode of action but is non-specific and not the only mode of action. Role of
oxidative stress is discussed in section 5.2.6 including effects of antioxidants on reducing B-Pb
levels.
Question Elf. How to use animal data to identify cut-off values for lead effects in humans?
It is difficult directly extrapolate quantitative data from animal studies to humans but results in
animals can suggest that there may not be a cut-off value for a particular effect in humans, e.g Pb
induced increases in NMDA receptor density, p5-23 line 7-9, and may indicate beneficial or
adverse effects of modifying factors, e.g. nutritional supplements, dose-response, PB/Ca).
The discussion of Integration of Research Findings (neuro)atthe end of section 5.3.1 and
bullet 2 on page 5-43 debates low level effects of lead on the CNS in animal models. From my
own experiences many years ago it was clear that it required a much higher blood lead level in
the rat to attain a particular brain lead level than in a human so there are clearly species
differences. However, animal models might address the question of reversibility. Also animal
models might address the question of linearity of effect at low levels of exposure
Other examples that show how animal studies assist in identifying low-level sub-clinical effects
are cited in section 5.4, e.g. (p 5-93 para 1). Placental effects of Pb exposure in squirrel monkeys
D-28
-------
were determined without overt toxicity to mothers, page 5-94 para 5.4.3.4. Pb induced
endocrine mediated alterations of female reproductive system of rats and non-human primates
suggest that non-human primates are particularly relevant to extrapolations to humans and
provide dose-response information for effects on female sex hormones and menstrual cycle.
Other comments about specific sections in Chapter 5
Section 5.2 Effects on Heme Synthesis
Section is well written, succinct with good summary. Last three paragraphs of section 5.2
concerns chelation as protective antioxidants for erythrocytes. Para 2 page 5-16 second
sentence makes the general statement that metal chelators form insoluble complexes. Shouldn't
that be soluble complexes?
Summary for this section 5.2.7 is very good.
Section 5.3 Neurological Effects
See comments in response to Charge question El.
Section 5.4 Reproductive Effects
Section 5.4.7.3 (p5-104) Developmental effects on the Retina might be crossed referenced with
Section p 5-31 Retinal function in Rodents (not numbered).
Conclusions to this section p 5104-108 are really a lengthy repletion with cited references of
earlier text. Might be integrated into the chapter followed by a more succinct set of conclusions.
Section 5.5 Effects on CV system.
No comments, Good summary and conclusions p5130-131.
Section 5.6 Genotoxic and Carcinogenic Effects
Section 5.6 P5-135 lines 1-3 the statement that the "data (from the Waalkes et paper on MT-null
mice) convincingly indicate that metallothionein binds Pb as part of an inclusion body and
prevents tumors" is not correct. Pb induced inclusion bodies are not formed in mt-null mice. The
study is correctly interpreted page 5-293, line 28-29, - that mt (gene) or a closely related gene is
involved in the formation of Pb-binding proteins in the kidney.
Section 5.6.6 Conclusions. I agree that overall conclusions have not changed much since the
1986Pb AQCD.
D-29
-------
Section 5.7 Lead and the Kidney
This section is well done. I suggest adding in summary Section 5.7.5 that earlier experimental
studies have shown that acute effects on tubular cells are generally reversible. With continued
exposure acute renal effects may progress to a chronic irreversible nephropathy. Also none of
the biomarkers for renal effects are specific for lead effects on the kidney.
I have no comments on Section 5.8 Effects on Bone and Teeth and Section 5.9, Effects on
Immune system
Section 5.10 Other organ systems
Series of heterogeneous topics.
Summary 5.10. l(re: liver effects) is v. good.
Summary of Section on Gasrointestinal absorption 5.10.2.7 is good and summarizes factors that
influenced rate/percent of absorption. However, I did not find any discussion of this in this
document particularly concerns about effects of chemical speciation of lead absorption, e.g Pb
acetate versus Pb sulfate and exposure to dust, lead ores, mine tailings etc. Is all of this
incorporated in the modeling Chapter.?
Section 5.11 Lead-binding proteins is a complex topic which is still evolving.
Separating the discussion of proteins in the intranuclear inclusion body from cytoplasmic lead
binding proteins implies that the proteins involved may be different. That may not be the case.
The paper describing the formation of inclusion bodies by Mclaughlin et al. 1980 (cited in
another context line 29, p284, describes the ultrastructural appearance of fibrils in the
cytoplasma in response to lead exposure with subsequent formation of intranuclear inclusion
bodies. These studies suggest that intranuclear inclusion body protein may be derived from the
cytoplasm. I suggest that sections 5.11.1 and 5.11.2 be merged with inclusion of the McLauglin
et al.1980 observation .
Also discussion of the studies by Harry, et al. 1996 (Tox Appl Pharmacol 139:84-93) regarding
fibrillar acidic protein in the developing rat brain should also be included. This fibrillary protein
has some similarities to inclusion body protein described by Goyer et al. 1970a and Moore and
Goyer, 1974.
Lead binding protein in erythrocytes seems to be an independent phenomenon but inconclusive
at this point. The proteins describe by Fullmer et al. (1985) in the intestine are more likely Ca
transport proteins as suggested.
The summary or a newly written conclusion section might provide a synthesis of the different
studies and approaches to metal binding proteins and how they are similar/different.
The summary does not include a comment about inclusion body protein.
D-30
-------
Discussion of Mt in bullets in the second and fourth bullet might be combined.
March 10, 2006
From: Robert A. Goyer
To Rogene Henderson, CASAC Chair; Fred Butterfield, CASAC Federal Officer
Post-meeting review comments on 1st draft Pb AQCD
I was not in attendance when Chapter 5 was discussed but would like to submit the following
brief comments based on my earlier review and the review meeting discussions I did attend.
Chapter 5 needs to be better organized to provide balanced treatment of the topics in each. Also
redundancies should be omitted and there should be succinct conclusions at the end of each
section. The 1st draft is into 11 sections concerning 10 organ systems and one section, and a last
one, on lead- binding proteins. I suggest consolidating discussions regarding various
susceptibility factors into a new single section including the following topics.
Polymorphisms/genetics
Nutrition
Age, all ages? (from conception to the elderly)
SES
Biochemical biomarker discussion in adult section p5-74, 75 really identifies homocystine as a
susceptibility factor and might be considered as such or might even be included as a
polymorphism.
There might also be some discussion concerning role of other metals on Pb distribution
beginning in section 5-7, p. 5-178.
The section (not numbered) beginning on p. 5-66 titled Dose-response paradigms serves to
connect neurological toxicities to the next chapter, clinical effects, in section 6.3.1 suggest this
Dose Response discussion is a good bridge to the clinical chapter. This might form part of a
succinct summary of the whole chapter.
The following comments were included in my pre review meeting submission but concern
corrections and omissions I wish to emphasize
D-31
-------
Section 5.2 Effects on Heme Synthesis
Section is well written, succinct with good summary. Last three paragraphs of section 5.2
concerns chelation as protective antioxidants for erythrocytes. Para 2 page 5-16 second
sentence makes the general statement that metal chelators form insoluble complexes. Shouldn't
that be soluble complexes?
Summary for this section 5.2.7 is very good.
Section 5.6 Genotoxic and Carcinogenic Effects
Section 5.6 P5-135 lines 1-3 the statement that the "data (from the Waalkes et al. paper on MT-
null mice) convincingly indicate that metallothionein binds Pb as part of an inclusion body and
prevents tumors" is not correct. Pb induced inclusion bodies are not formed in mt-null mice. The
study is correctly interpreted page 5-293, line 28-29, - that mt (gene) or a closely related gene is
involved in the formation of Pb-binding proteins in the kidney.
Section 5.6.6 Conclusions. I agree that overall conclusions have not changed much since the
1986Pb AQCD.
Section 5.7 Lead and the Kidneyn is well done but I suggest adding in summary Section 5.7.5
that earlier experimental studies have shown that acute effects on tubular cells are generally
reversible. With continued exposure acute renal effects may progress to a chronic irreversible
nephropathy. Also none of the biomarkers for renal effects are specific for lead effects on the
kidney.
Summary of Section on Gastrointestinal absorption 5.10.2.7 is good and summarizes factors that
influenced rate/percent of absorption. However, I did not find any discussion of this in this
document particularly concerns about effects of chemical speciation of lead absorption, e.g Pb
acetate versus Pb sulfate and exposure to dust, lead ores, mine tailings etc.
Section 5.11 Lead-binding proteins is a complex topic which is still evolving.
Separating the discussion of proteins in the intranuclear inclusion body from cytoplasmic lead
binding proteins implies that the proteins involved may be different. That may not be the case.
The paper describing the formation of inclusion bodies by Mclaughlin et al. 1980 (cited in
another context line 29, p284, describes the ultrastructural appearance offibrils in the cytoplasma
in response to lead exposure with subsequent formation of intranuclear inclusion bodies. These
studies suggest that intranuclear inclusion body protein may be derived from the cytoplasm. I
suggest that sections 5.11.1 and 5.11.2 be merged with inclusion of the McLauglin et al. 1980
observation.
Also discussion of the studies by Harry, et al. 1996 (Tox Appl Pharmacol 139:84-93) regarding
fibrillar acidic protein in the developing rat brain should also be included. This fibrillary protein
has some similarities to inclusion body protein described by Goyer et al. 1970a and Moore and
Goyer, 1974.
D-32
-------
Lead binding protein in erythrocytes seems to be an independent phenomenon but inconclusive
at this point. The proteins describe by Fullmer et al. (1985) in the intestine are more likely Ca
transport proteins as suggested.
The summary or a newly written conclusion section might provide a synthesis of the different
studies and approaches to metal binding proteins and how they are similar/different.
The summary does not include a comment about inclusion body protein.
Discussion of Mt in bullets in the second and fourth bullet might be combined.
D-33
-------
Mr. Sean Hays
Comments on Chapter 4 - Lead NAAQS
Submitted by: Sean Hays (Summit Toxicology)
Chapter 4 includes descriptions of most of the major lead kinetic models available for relating
environmental and dietary lead exposures with blood lead in humans. Chapter 4 includes a
description of the All Ages Lead Model (AALM), which is probably inappropriate given its'
incomplete status. It should be included once it has been formally accepted by the All Ages
Lead Model review panel of the Science Advisory Board.
The glaring omission from Chapter 4, and is contained no where else in the rest of the lead
NAAQS document, is a description of the pharmacokinetics of lead in humans. While the
models described in Chapter 4 synthesize what is known about the pharmacokinetics of lead, a
knowledge of the underlying pharmacokinetic science is required for reviewers to understand
and appreciate, 1) how historical exposures to lead can impact current blood lead levels, 2) how
pharmacokinetics of lead in particular populations yield some sensitivities, and 3) how transient
changes in exposure may or may not be an important contributor to blood lead levels and thus
impacts on public health. Therefore, Chapter 4 should include a review of the pharmacokinetics
of lead in humans, with particular emphasis placed on sensitive populations, including;
• In utero exposures
• Children, including neonates,
• Potential for lead to be excreted in breast milk
• Post-menopausal women and individuals with osteoporosis
It is not enough to simply review the models of lead pharmacokinetics without a description of
the underlying pharmacokinetic literature that forms the basis of these models. Given that not all
models are applicable for all the sensitive populations, a knowledge of the pharmacokinetics in
these populations will be critical for reviewers to understand how the pharmacokinetics of lead in
these populations may impact the risk assessment.
Without knowing how a model will be used in developing an air quality standard, it is impossible
yet to provide insights on which model(s) would be most appropriate for this exercise. Having
an actual requirement and use for a model imposes certain constraints and requirements. Once
these are known, more insights can be provided on which model(s) are valid and/or best suited
for the application, in this case for a risk assessment of lead.
D-34
-------
The following are specific comments on Chapter 4.
Page Line Comment
4-10 29 The term "probabilistic model" is inappropriate as a description of the IEUBK model
4-13 1 "unlimited" should be replaced with the term "linear"
4-13 6 replace "a function of with "with increasing total lead intake and age"
4-13 7 Need to define what is meant by absorption fractions are medium specific.
4-13 17 The reference to Table 4-1 seems to indicate different content than actually exists.
Reference to "probability of elevated" should be deleted. The model is actually
designed to predict blood lead concentrations, not probability of elevated blood lead
4-17 25 levels.
4-18 7 It is not accurate to describe the GSD function as a "probability model"
The PBPK model has been used to address post-menopause and osteoporosis
(O'Flaherty, 2000). This should be included here and in various other places later in
4-26 9 the chapter.
The reference to 70% and 30% are backwards. 70% of lead elimination is attributed
4-26 23 to urine in the PBPK model.
4-29 14 Another approach is to use a GSD approach like that used in the IEUBK model.
D-35
-------
Dr. Bruce Lanphear
Comments on "Routes of Human Exposure to Lead and Observed Environmental
Concentrations" (Chapter 3)
General Comments:
The Chapter on "Routes of Human Exposure to Lead" is a good beginning. There are, however,
several modifications that would enhance the Chapter. An overview and introduction of the
Chapter would be useful. As written, the chapter meanders through various sources of exposure
without a logical format or outline. It would also be helpful if the authors provided a description
of the scope of the review and the systematic approach that was used to identify the various
papers on lead exposure published since 1990. There were major gaps in their literature review
and it wasn't clear why there wasn't a greater focus on data to quantify the relative contributions
of various sources of lead exposure. If the contribution of various sources of lead exposure is to
be included in the Synthesis Chapter, it would be worth noting this in the introduction.
This overview should include insights that would help the reader understand the contribution and
trends in lead exposure. For example, it may not be obvious to all readers that the various
sources of lead intake are cumulative, and that blood lead (in children) and bone lead (in
adolescents and adults) are cumulative biomarkers of exposure.
Charge Question Cl. Does Chapter 3 provide adequate coverage of pertinent available
information (especially as it pertains to the United States) on lead exposure routes, as well
as environmental lead concentrations, including those in air, drinking water, food, soils,
and dust? Does the chapter adequately delineate interconnections between airborne lead
and its potential contributions (via secondary deposition) to lead in other media (e.g. indoor
dust)?
As written, the Chapter does not delineate the interconnections between airborne lead and other
media. Nor does it adequately cover the available information on routes of lead exposure. Most
of the following specific comments attempt to fill some of those gaps. The description of the
contribution of airborne lead was inadequate.
Given that dust is the most proximal exposure for contemporary children, it deserves
considerably more attention. There is now considerable data on the probability of a child having
a blood lead level > 10 |j,g/dL if exposed to various levels of lead-contaminated house dust. In
2001, the US EPA promulgated residential lead standards of 40 ng/ft2 for floors and 250 ng/ft2
for window sills. Data from epidemiologic studies show that 5% of children have a blood lead
level > 10 |j,g/dL at a median floor dust lead level of 5 ng/ft2 (Lanphear, 1996; Lanphear, 1998;
Malcoe, 2002; Lanphear, 2005). At a floor standard of 50 ng/ft2, 20% of children were
estimated to have a blood lead level > 10 ng/dL (Lanphear, 1998). Children who were exposed
to floor dust lead levels > 25 |J,g/ft2 were at 8-times greater risk of having blood lead levels > 10
Hg/dL compared with those exposed to levels below 2.5 ng/ft2 (Lanphear 2005).
D-36
-------
Malcoe LH, Lynch RA, Keger MC, Skaggs VJ. Lead sources, behaviors, and socioeconomic
factors in relation to blood lead of native american and white children: a community-based
assessment of a former mining area. Environ Health Perspect. 2002 Apr; 110 Suppl 2:221-31.
Lanphear BP, Hornung R, Ho M. Screening housing to prevent lead toxicity in children. Pub Health Rep
2005;120:305-310.
Lanphear BP, Hornung R, Ho M, Howard CR, Eberly S, Knauf K. Environmental lead exposure
during early childhood. J Pediatr. 2002 Jan;140(l):40-7.
Lanphear BP, Matte TD, Rogers J, Clickner R, Dietz B, Bornschein RL, Succop P, Mahaffey
KR, Dixon S, Galke W, Rabinowitz M, Farfel M, Rohde C, Schwartz J, Ashley P and Jacobs
DE. The contribution of lead-contaminated house dust and residential soil to children's blood
lead levels: A pooled analysis of 12 epidemiologic studies. Environmental Research 1998;79:51-
68.
As to the contribution of airborne lead to interior house dust, there is an article by Caravanos and
others that should be incorporated even though it was only published in 1996. Caravanos
collected weekly sample collection of interior and exterior settled dust in New York City to
monitor accumulation of atmospheric deposition of lead (Caravanos, 2006). The median values
of leaded dust for the interior plate (adjacent to the open window), unsheltered exterior plate, and
the sheltered exterior plate were 4.8, 14.2, and 32.3 |ig/feet2/week, respectively. The data suggest
that there is a continuous source of deposited leaded dust in interior and exterior locations within
New York City. Additional data from a control plate (interior plate with the window closed)
showed that the source of the interior lead deposition was primarily from exterior
(environmental) sources.
Caravanos J, Weiss AL, Jaeger RJ. An exterior and interior leaded dust deposition survey in
New York City: Results of a 2-year study. Environ Res. 2006;100:159-164.
Introduction (or embedded in the section on various sources of lead exposure):
It would be useful if the authors would provide a description of the relative contribution of
various sources of lead exposure that vary by age. Children's blood lead levels rise rapidly
between 6 and 12 months of age, peak between 18 months to 36 months and then gradually
decline (Clark, 1991). The peak in children's blood lead levels is due to the confluence of
normal mouthing behaviors and increasing mobility. Lead-contaminated floor dust is a source of
lead intake throughout early childhood, but lead-contaminated dust on windowsills is not a major
source of intake until the second year of life, when children stand upright. Soil ingestion, as
reported by parents, peaks during the second year of life and diminishes thereafter (Lanphear,
2002).
Clark S, Bornschein R, Succop P, Roda S, Peace B. Urban lead exposures of children in Cincinnati,
Ohio. Chemical Speciation Bioavailability 1991;3:163-171.
Lanphear BP, Hornung R, Ho M, Howard CR, Eberly S, Knauf K. Environmental lead exposure
during early childhood. J Pediatr. 2002 Jan;140(l):40-7.
D-37
-------
It would also be useful if the authors could summarize how our understanding of lead exposure
has changed since the 1990 supplement. For example, there have been several randomized trials
published since 1990 that provide insight into the relative contribution of lead intake from
various sources that were entirely overlooked by the authors. In a meta-analysis of 4 dust control
trials and 1 low-cost housing intervention, there was no significant reduction in children's mean
blood lead concentration (Haynes, 2001). There was, however, a > 50% reduction in children
having blood lead concentrations > 15 |j,g/dL and > 20 |j,g/dL in the experimental groups
compared with the control groups, indicating some benefit of dust control for children with
higher blood lead levels (Haynes, 2001). Since the publication of this systematic review, two
additional studies were published (Jordan, 2004; Brown, 2005). One community based study
showed a 34% (non-significant) reduction in the proportion of children with a blood lead level >
10|j,g/dL (Jordan, 2004). As described later, there have been studies of soil abatement that
provide insight into the contribution of lead from soil (e.g., Aschengrau, 1994).
Haynes E, Lanphear BP, Tohn E, Farr N, Rhoads GG. The effect of dust controls on children's
blood lead concentrations: A systematic review. Env Health Perspect 2001; 110:103-107.
Jordan CM, et al. A randomized trial of education to prevent lead burden in children at high risk
for lead exposure: efficacy as measured by blood lead monitoring. Environ Health Perspect.
2003;! 11:1947-51.
Aschengrau A, Beiser A, Bellinger D, Copenhafer D, Weitzman M. The impact of soil lead abatement
on urban children's blood lead levels: phase II results from the Boston Lead-in-Soil Demonstration
Project. Environ Res 1994;67:125-148.
Brown MJ, McLaine P, Dixon S, Simon P. A randomized, community-based trial of home
visiting to reduce blood lead levels in children. Pediatrics. 2006 Jan;117(l):147-53.
Page 3-1, line 12-21: This paragraph leaves the impression that exterior sources of lead are more
important source of lead in house dust than interior sources, such as lead-contaminated paint.
This may be true for mining, milling or smelting communities, but it is not true for many older
urban communities. (I am certainly not trying to revive or perpetuate the old "environment
versus housing" wars; my interpretation is that they are probably contributing equally and both
certainly need to be reduced to impact human exposure.) The statement also needs to be
clarified because there is evidence that paint is a particularly important source for children with
elevated blood lead levels (Sachs 1970; McElvaine, 1990; Shannon 1995; Bates, 1995; Lanphear
1996).
There are several relevant studies shed some light on this specific question about sources of lead
in house dust. Hunt used a classification scheme to categorize the house dust particles as auto
exhaust, road dust, garden soil or paint in England (Hunt, 1993). The primary contributing source
in the 64-1000-microm size range of the house dusts was paint. In the 0-64-microm size fraction,
paint, road dust and garden soil all made significant contributions.
D-38
-------
In a U.S. study, Sterling and others showed that all three sources - mining waster, paint and soil
- were determinants of house-dust. Not surprisingly, they concluded that soil and mining wastes
accounted for over 50% of lead in house dust whereas only 16% (23% using a weighted formula
to account for the lead concentration per particle) was from paint (Sterling, 1998). Paint was
responsible for 16% to 23% (29% was of unidentified origin) of lead in house dust in a mining
community (Sterling, 1998).
Adagte et al. used lead stable isotope ratio analysis examine the relationship between sources of
lead in 22 dust wipe samples collected from 10 homes in Jersey City, NJ (Adgate, 1998). They
found high correlations between isotope ratios of wipe samples for street dusts and exterior soils,
indicating that these two sources were indistinguishable. They were treated as a single exterior
source in a source apportionment using isotope ratio matching. The upper-bound estimate of the
contribution of interior lead-based paints to 10 floor- and eight sill-wipe samples was 56% and
50%, respectively (Adgate, 1998).
In a case study of two households in Oakland, California, Yaffe et al. found that paint and
surface soil samples collected in and around both households of children with elevated blood
lead levels (Yaffe, 1983). The isotopic ratios of lead in the blood of these children were close to
the average lead ratios of paints from exterior walls and of surface soils in adjacent areas where
the children played. In both case studies, the data suggest that the lead in the soil was derived
mainly from weathering of lead-based exterior paints and that the lead-contaminated soil was a
proximate source of lead in the blood of the children.
In a study from New Zealand, Bates and coworkers found that children with elevated lead levels
were more likely to live in a house greater than 50 years old where paint removal had taken place
in the last 2 years (RR = 14.4, 95% CI: 2-107) (Bates, 1995). Eating dirt, particularly for children
who usually played outside within 2 meters of the house, was also a risk factor for elevated blood
lead levels. Soil lead levels increased with the age of the house and were correlated with blood
lead levels (r = 0.32).
In a study to examine risk factors to explain the racial differences in children's blood lead
concentrations, the primary differences in exposure that explained racial disparities in children's
blood lead concentration were dust lead loading and paint lead variables (Lanphear, 1996).
Predictors of blood lead concentration for Black children, who had a geometric mean blood lead
of 8.8 ng/dL, were dust lead loading, paint lead concentration and condition, water lead, soil lead
concentration and putting soil or dirt in their mouths. In contrast, predictors for White children,
who had a geometric mean blood lead of 4.4 ng/dL, was limited to soil lead concentration, time
spent outdoors, and putting soil or dirt in their mouths (Lanphear, 1996). The authors concluded
that a major reason for the racial disparity (and the significantly higher blood lead levels) was
that Black children were exposed to both interior and exterior sources of lead.
Page 3-2, line 22-31: The authors should consider discussing other relevant research using
isotopic ratios to enhance the discussion of the source of lead in housing (see Gwiazda RH, et al.
EHP2000;108:1091-1097).
D-39
-------
It is also worth noting that airborne lead and exterior lead may be derived from lead-based paint
due to demolition or remodeling of other dwellings (Farfel, 2003).
Sterling DA, Johnson PL, Murgueytio AM, Evans RG. Source contribution of lead in house
dust from a lead mining waste superfund site. J Expo Anal Environ Epidemiol. 1998 Jul-
Sep;8(3):359-73.
Hunt A, Johnson PL, Thornton I, Watt JM. Apportioning the sources of lead in house dusts in
the London borough of Richmond, England. Sci Total Environ. 1993 Sep 30;138(1-3):183-206.
Adgate JL, Rhoads GG, Lioy PJ.The use of isotope ratios to apportion sources of lead in Jersey
City, NJ, house dust wipe samples. Sci Total Environ. 1998 Oct 8;221(2-3): 171-80.
Yaffe Y, Flessel CP, Wesolowski JJ, del Rosario A, Guirguis GN, Matias V, Gramlich JW, Kelly
WR, Degarmo TE, Coleman GC. Identification of lead sources in California children using the
stable isotope ratio technique. Archives of Environmental Health 1983;237-45.
Sachs HK, Blanksma LA, Murray EF, O'Connell MJ. Ambulatory treatment of lead poisoning:
report of 1,155 cases. Pediatrics 1970;46:389-396.
McElvaine MD, DeUngria EG, Matte TD, Copley CG, Binder S. Prevalence of radiograph!c evidence of
paint chip ingestion among children with moderate to severe lead poisoning, St Louis, Missouri, 1989
through 1990. Pediatrics 1992;89:740-742.
Shannon MW, Graef JW. Lead intoxication in infancy. Pediatrics 1992;89:87-90.
Bates M, Malcolm M, Wyatt R, Garrett, N, Galloway Y, Speir T, Read, D. Lead in children
from older housing areas in the Wellington region. New Zealand Medical Journal 1995; 108:400-
404.
Lanphear BP, Weitzman M, Eberly S. Racial differences in environmental exposures to lead.
American Journal of Public Health 1996; 86:1460-1463.
Farfel MR, Orlova AO, Lees PS, Rohde C, Ashley PJ, Chisolm JJ Jr. A study of urban housing
demolitions as sources of lead in ambient dust: demolition practices and exterior dust fall.
Environ Health Perspect. 2003; 111:1228-34.
Page 3-1, line 27: There are several other relevant articles that should be cited to indicate that
various methods of lead hazards control can result in an increase in children's blood lead levels,
especially for young children who exhibit frequent mouthing behaviors. (Amitai, 1991, Rey-
Alvarez 1987, Swindell 1994, Aschengrau 1997, Clark 2004).
Two studies of paint abatement are particularly noteworthy because they show that undue lead
exposure can occur despite the use of dust clearance tests after lead abatement (Aschengrau,
1997, Clark 2004). In a prospective, controlled study of children with baseline blood lead below
22 ng/dL, Aschengrau reported a 6.5 |j,g/dL increase in blood lead for children who had paint
abatement (Aschengrau, 1997). Clark and coworkers found, despite using HUD post-abatement
D-40
-------
standards, that 6-month old children were 11-fold more likely to have an increase in blood lead
concentration > 5 |j,g/dL. Collectively, these studies raise questions about the adequacy of
existing clearance standards to protect children from lead hazards following abatement or other
lead hazard controls, especially for younger children (Clark, 2004).
Amitai Y, Brown MJ, Graef JW, Cosgrove E. Residential deleading: effects on the blood lead
levels of lead-poisoned children. Pediatrics 1991;88:893-8987.
Rey-Alvarez S, Menke-Hargrave T. Deleading dilemma: Pitfall in the management of childhood
lead poisoning. Pediatrics 1987;79:214-217.
Swindell SL, Charney E, Brown MJ, Delaney J. Home abatement and blood lead changes in
children with class III lead poisoning. Clin Pediatr 1994;33:536-541.
Aschengrau A, Beiser A, Bellinger D, Copenhafer D, Weitzman M. Residential lead-based-paint
hazard remediation and soil lead abatement: their impact among children with mildly elevated
blood lead levels. Am J Public Health 1997;87:1698-1702.
Clark S, Grote J, Wilson J, Succop P, Chen M, Galke W, McLaine P. Occurrence and
determinants of increases in blood lead levels in children shortly after lead hazard control
activities. Environ Res. 2004:(2): 196-205.
Page 3-2, line 7: The authors usually refer to dust lead concentration, but some of the studies
express dust in loading (|j,g/ft2). This raises the question about whether concentration or loading
is a better measure of exposure. While there are some advantages to dust lead concentration
(mostly limited to modeling), dust lead loading is a significantly better predictor of children's
blood lead levels (Lanphear, 1995; Lanphear 1998). Indeed, the US EPA residential dust lead
standards rely on dust lead loading, not concentration. This distinction is key to understand
childhood lead exposure.
Lanphear BP, Emond M, Jacobs DE, Weitzman M, Winter NL, Tanner M, Yakir B, Eberly S. A
side-by-side comparison of dust collection methods for sampling lead-contaminated house-dust.
Environmental Research 1995;68:114-123.
Lanphear BP, Matte TD, Rogers J, Clickner R, Dietz B, Bornschein RL, Succop P, Mahaffey
KR, Dixon S, Galke W, Rabinowitz M, Farfel M, Rohde C, Schwartz J, Ashley P and Jacobs
DE. The contribution of lead-contaminated house dust and residential soil to children's blood
lead levels: A pooled analysis of 12 epidemiologic studies. Environmental Research 1998;79:51-
68.
Page 3-4, Table 3-1: The description and tabulation of dust lead levels is vague. I would
imagine, for example, that the reason Cincinnati had dust lead loading values ranging from 20
Hg/ft2 to 293 ng/ft2 is because they were collected from various surfaces. But the reader is left to
guess why the levels vary. This raises the problem of treating all dust lead levels the same
regardless of whether they were collected from different surfaces using different sampling
methods (e.g., floor versus window sill or window trough).
D-41
-------
Page 3-1, lines 6-29:
The authors indicate that soil is an important source of lead exposure, but their review did not
include more recent studies that attempted to estimate the contribution of lead to children's lead
intake. Duggan and Inskip estimated that children's blood lead concentrations increased 5 |j,g/dL
for every 1000 ppm increase in soil lead concentration, but this estimate didn't account for other
sources of lead intake (Duggan, 1985). In a randomized, controlled trial of soil abatement,
Aschengrau and coworkers reported a 1.12 to 1.35 |j,g/dL decrease in blood lead for every 1000
ppm reduction in soil lead concentration (Aschengrau, 1994). In a pooled analysis of 12 studies,
there was an estimated 3.8 |j,g/dL increase in blood lead concentration for every 1000 ppm
increase in soil lead concentration (Lanphear, 1998). Finally, in a Superfund site, soil abatement
was shown to lead to a 3.5 |j,g/dL decrease in blood lead levels of 6 to 36 month old children; the
decrease was less for 36 to 72 month old children (Lanphear 2003). The variation in the reported
relationship of lead-contaminated soil is due to a number of factors, including the age of children
studied, adjustment for the contribution of lead intake from other sources, and mouthing
behaviors. Indeed, once adjusted for age (i.e., redistribution of bone lead stores - see Gwiazda
R, Campbell C, Smith D. A noninvasive isotopic approach to estimate the bone lead contribution
to blood in children: implications for assessing the efficacy of lead abatement. Environ Health
Perspect. 2005; 113:104-10) the estimated contribution of soil lead to children's blood lead
concentration from these latter three studies is quite similar.
Duggan MJ, Inskip MJ. Childhood exposure to lead in surface dust and soil: a community health
problem. Public Health Rev 1985; 13:1-54.
Aschengrau A, Beiser A, Bellinger D, Copenhafer D, Weitzman M. The impact of soil lead abatement
on urban children's blood lead levels: phase II results from the Boston Lead-in-Soil Demonstration
Project. Environ Res 1994;67:125-148.
Lanphear BP, Matte TD, Rogers J, Clickner R, Dietz B, Bornschein RL, Succop P, Mahaffey
KR, Dixon S, Galke W, Rabinowitz M, Farfel M, Rohde C, Schwartz J, Ashley P and Jacobs
DE. The contribution of lead-contaminated house dust and residential soil to children's blood
lead levels: A pooled analysis of 12 epidemiologic studies. Environmental Research 1998;79:51-
68.
Lanphear BP, Succop P, Roda S, Henningsen G. The Effect of Soil Abatement on Blood Lead Levels in
Children living near a Former Smelting and Milling Operation. Public Health Reports 2003;! 18:83-90.
Page 3-9, lines 1-6: It would be helpful if there was additional and more specific information
about the four areas (Liberty-Acadia, MO, Herculaneum MO, East Helena, Lame Deer, MT) that
exceeded the EPA air lead standard, including trends over the past 5 to 10 years because these
sites are likely to be of particular interest and relevance to the review. Although the authors
write that only two sites exceeded EPA standards in 2004, my understanding is that
Herculaneum (MO) once again exceeded standards again in 2005. Was this one of the sites? It
would useful to plot the airborne levels of lead for each of the sites up to the present.
D-42
-------
Page 3-11: The authors describe occupational exposure, but they do not consider or describe
para-occupational (take-home) exposures for children (see Roscoe, 1999). This includes
construction, lead mining and smelting, renovation or construction workers, and manufacturing.
There can also be community exposure from paint removal of bridges.
Roscoe, RJ, et al. Blood lead levels among children of lead-exposed workers: A meta-analysis.
Am J Ind Med 1999;36:475-481.
Page 3-13, line 1-10: If the authors want to conclude that "The dominant source of lead to soil is
atmospheric deposition from local sources and long-range transport." they either need to put
some conditions on this statement to indicate a specific setting or more thoroughly review the
literature to make a convincing argument.
Page 3-14, lines 16: All of the studies the authors cite to argue that "the major source of lead in
the urban environment is soil" did not measure paint lead. Thus, it is not a surprise that age of
housing or paint wasn't found to be a major contributor to lead in soil. (To be fair, most of the
urban studies did not examine the contribution from airborne lead. Still, the authors should be
careful not to over-interpret the conclusions of studies that did not evaluate all sources of lead
exposure.) In one study, for example, which studied housing units in an urban setting, paint lead
concentration and deteriorated lead-based paint predicted soil lead concentration (Lanphear
1998). Jacobs reported that soil lead levels were related to deteriorated exterior lead-based paint.
They found that for units with and without deteriorated exterior lead-based paint, the percent of
units with bare soil lead levels > 1,200 ppm decreased from 24% to only 4% (Jacobs, 2002).
Thus, as written, the statement that "the major source of lead in the urban environment is soil" is
overly broad.
Jacobs DE, et al. The Prevalence of Lead-Based Paint Hazards in U.S. Housing. Environ Health
Perspect 110:A599-A606 (2002).
Lanphear BP, Roghmann KJ. Pathways of lead exposure in urban children. Environmental
Research 1997;74:67-73.
Page 3-33, lines 8-18: I found it odd that lead-based paint was mentioned almost as an
afterthought when it is arguably the major source of lead exposure for contemporary children,
especially those whose blood lead concentrations exceed 10 |j,g/dL. For a comprehensive review
of lead exposure, it would be worth describing that paint that was used through the 1950s, and
continuing to some extent through the 1970s, often contained high concentrations of lead. It
would also be worth providing insight into the relative contribution from lead-based paint for
children with varying degrees of lead exposure. For example, over 80% of children with blood
lead levels greater than 50 |j,g/dL were reported to ingest paint chips or broken plaster (Sachs,
1970). Children with blood lead above 55 |j,g/dL were 10-times more likely to have paint chips
observable on abdominal radiographs than children who had blood lead levels below this value
(McElvaine, 1992). The majority of preschool children with blood lead over 25 |j,g/dL were
reported to put paint chips in their mouth (Shannon, 1997). Because all of these children were
exposed to background levels of leaded gasoline, these studies really indicate that it was children
with exposure to both leaded gasoline and leaded paint that ultimately developed lead poisoning
D-43
-------
or elevated blood lead concentrations. Still, this perspective provides insight into past sources
and provides an opportunity for the authors to describe trends in childhood lead exposure.
Sachs HK, Blanksma LA, Murray EF, O'Connell MJ. Ambulatory treatment of lead poisoning:
report of 1,155 cases. Pediatrics 1970;46:389-396.
McElvaine MD, DeUngria EG, Matte TD, Copley CG, Binder S. Prevalence of radiograph!c evidence of
paint chip ingestion among children with moderate to severe lead poisoning, St Louis, Missouri, 1989
through 1990. Pediatrics 1992;89:740-742.
Shannon MW, Graef JW. Lead intoxication in infancy. Pediatrics 1992;89:87-90.
Page 3-35, lines: 1-22: There was no mention of soil sampling or dust sampling in the section on
measurement methods. Yet, it is well known that dust lead loading varies considerably by the
surface sampled and the sampling methods used (Lanphear, 1995). This section should include
discussion of how sampling different areas of the floor or using different sampling methods also
affects dust lead loading values (Lanphear, 1995). Dust lead loading collected from troughs is
oftentimes 1000-fold greater than floors samples. The levels of lead in house dust collected from
the midpoint of a room tends to be lower than those found under a window or in the perimeter of
a room (Sayre, 1974). Finally, soil lead concentration can vary by location (perimeter of
foundation versus yard samples), by the sieve size used, and the depth of collection.
Lanphear BP, Emond M, Jacobs DE, Weitzman M, Winter NL, Tanner M, Yakir B, Eberly S. A
side-by-side comparison of dust collection methods for sampling lead-contaminated house-dust.
Environmental Research 1995;68:114-123.
Sayre JW, Katzel MD. Household surface lead dust: its accumulation in vacant homes. Environ
Health Perspect. 1979;29:179-82.
Page 3-35, lines 26-29: In the summary (and on page 3-14, lines 14-16), the authors refer to the
fact that "people in cities, especially in poor and minority-dominated neighborhoods, are the
most at risk for lead exposure". But, they do not adequately review the literature to clarify the
specific differences in the types of exposure by urban status. Similarly, I was surprised that there
was no mention of the striking racial disparity in blood lead levels (Pirkle, 1998) due, in large
part, to differences in environmental exposures (Lanphear, 2002). This deserves to be included
in the review.
Pirkle JL, Kaufmann RB, Brody DJ, Hickman T, Gunter EW, Paschal DC. Exposure of the U.S.
population to lead, 1991-1994. Environ Health Perspect 1998;11:745-50.
Lanphear BP, Hornung R, Ho M, Howard CR, Eberly S, Knauf K. Environmental lead exposure
during early childhood. J Pediatr. 2002 Jan;140(l):40-7.
In their review of the human exposure to lead-contaminated water, it would be valuable to
review evidence on its contribution to blood lead concentrations in pregnant women and children
(Watt, 1996), and the trends in its relative contribution to lead intake (Levin 1989). In a
D-44
-------
prospective study of 248 children followed from 6 to 24 months, children who were exposed to
water lead > 5 ppb had blood lead concentrations 1.0 |j,g/dL (20%) higher than children with
water lead levels < 5 ppb (Lanphear 2002). Thus, water is not a trivial source of lead intake for
young children in many communities. Moreover, as predicted, lead in water is becoming an
increasingly important source of lead intake as other sources of lead intake have diminished
(Levin, 1989).
Watt GCM, Britton A, Gilmour WH, et al. Is lead in tap water still a public health problem? An
observational study in Glasgow. BMJ 1996;313:979-981.
Levin R, Schock MR, Marcus A. Exposure to Lead in US Drinking Water. Trace Substances Environ
Health 1989;319-344.
Lanphear BP, Hornung R, Ho M, Howard CR, Eberly S, Knauf K. Environmental lead exposure
during early childhood. J Pediatr 2002;140(l):40-7.
Page 3-27, lines 26-29: In the section on exposure via food ingestion, it would be important to
describe various factors that modify lead absorption. In the experimental setting, for example,
fasting has been shown to modify lead absorption in adults. There was a 10-fold increase in lead
absorption among fasting volunteers who ingested lead compared with those who had recently
eaten (Rabinowitz 1980; Maddaloni 1998).
Maddaloni M, Lolacono N, Manton W, Blum C, Drexler J, Graziano J. Bioavailability of soilborne lead
in adults, by stable isotope dilution. Environ Health Perspect 1998;106:Suppl:1589-1594.
Rabinowitz MB, Kopple JD, Wetherill GW. Effect of food intake and fasting on gastrointestinal lead
absorption in humans. Am JClinNutr 1980;33:1784-1788.
Comments on "Epidemiologic Studies of Human Health Effects Associated with Lead
Exposure" (Chapter 6)
QF1. Are the discussions on the various biomarkers adequate to elucidate their role in assessing
human health effects from lead exposure?
Yes.
QF1. Does Chapter 6 adequately address the issue of which exposure metrics (i.e., peak, average
lifetime or concurrent) are now believed to be most strongly associated with specific health
endpoints and, therefore, should be the focus of exposure and risk assessments targeting those
endpoints?
Yes, at least for the IQ-blood lead relationship. There is less data on various blood lead or bone
lead indices for predicting other outcomes, such as behavior.
D-45
-------
QF2a. New human epidemiologic studies provide evidence for IQ decrements associated with
blood lead levels <10 |j,g/dL. The pooled analysis shows a significant inverse relationship of lead
concentration and IQ measured at school age, after adjusting for common confounders. Due to
the log-linear relationship, the slope of the lead effect on IQ was greatest at the lower blood lead
level range, i.e., below 10 |j,g/dL. Does this chapter adequately address questions regarding
significant neurotoxic effects observed at low blood lead levels (<10
The Chapter could describe one other aspect of the pooled analysis that is directly related to
potential limitations (Lanphear, 2005). There was some criticism that a certain site was driving
the results and that the HOME Score was not measured concurrent with the IQ test (Ernhart,
2006). This commentator also pointed out limitations of other sites (e.g., early blood lead tests
measured with capillary finger stick rather than veni puncture and inclusion of two
geographically distinct villages with disparate levels of childhood lead exposure). In the pooled
analysis, we conducted sensitivity analyses to test whether excluding any one of the sites altered
the results of the pooled analysis. The analyses showed quite convincingly that no single study
was responsible for the estimated relationship of lead and IQ decrements. This finding
diminishes the concerns about unique attributes or potential limitations for any specific sites.
Ernhart CB. Effects of lead on IQ in children. Environ Health Perspect 2006;! 14:A85-6; author
reply A86-7.
QF2a. Is the issue of the influence of model selection on the estimated health effects adequately
discussed?
No. The pooled analysis used a log-linear analysis to quantify the lead-associated IQ
decrements. But it was not explicit in the review that the non-linear relationship observed in the
pooled analysis was not due to the influence of the log-linear model. The following points would
help to clarify this:
1. Using a cubic spline regression analysis, which doesn't assume any particular shape of the
relationship, the lead-associated IQ decrements were greater at lower blood lead levels.
2. The log-linear analysis agreed remarkably well with the cubic spline analysis compared with
other models.
3. Using linear models for blood lead levels above and below 7.5 ng/dL, the slope was
significantly steeper at lower blood lead levels compared with the slope at higher blood lead
levels.
4. The log-linear model will tend to exaggerate the slope at the lowest blood lead levels; as a
result, the authors provided estimates for the lead-associated IQ decrements from < 1 |j,g/dL to
10 |j,g/dL as well as from 2.4 |j,g/dL to 30 |j,g/dL (representing the 5th to 95th percentile)
(Lanphear, 2005).
D-46
-------
QF2b. Does this chapter provide an adequate overview of key lead-related health effects? Are
the key summary statements and conclusions regarding the effects of lead on various organ
systems sufficiently substantiated by the assessed epidemiologic evidence?
Yes.
QF2d. Drawing causal inferences between increased lead exposure and adverse health
effects in epidemiologic studies is complicated by the presence of many potential
confounders that may both affect lead exposure and be associated with the health outcome
of interest. Is the discussion of the various potential confounders of lead health effects
adequate? Given the concern regarding the influence of such confounders on the effect
estimates, are the stated key conclusions regarding lead effects on various health outcomes
appropriate?
Yes. The discussion was clear and appropriate. The conclusions were justified.
One option to further address potential confounders would be to incorporate more of the relevant
toxicological data from animal studies, but these studies should be described in Chapter 5.
One additional point that is worth describing is how other "unknown or unmeasured
confounders" may or may not account for the findings at lower levels of exposure. In general,
children who have lower blood lead levels also have fewer environmental insults (e.g., less
tobacco exposure, more nurturing home environment) that act as confounders than children with
higher blood lead levels. Yet the greatest decrements per unit increase in blood lead levels were
observed in this "low-risk" group. Although other types of confounders may be at play, do the
authors think this minimizes the problem of unmeasured confounders?
QF3. Discussions of epidemiologic studies mainly focus on studies of potential lead effects
among infants, school-aged children, the general population, and occupationally-exposed
populations. Some studies also examined potentially susceptible individuals such as those
with chronic medical diseases and specific genetic polymorphisms. Does Chapter 6
adequately cover key populations that should be considered for present purposes? Are the
discussions of differences in individual vulnerability and susceptibility adequate?
Given that there is limited evidence of individual susceptibility, the review is adequate.
Other Comments:
Page 6-50, lines 25-27: It is also worth pointing out that, in addition to other sociodemographic
factors, the observed lead effects vary because the mean blood lead concentrations differ among
the various cohorts. The two prospective studies with the lower mean blood lead levels exhibited
the steepest IQ-blood lead slope, despite representing two distinct subpopulations (Lanphear,
2005).
D-47
-------
Page 6-68, lines 10-12: As written, the reader might assume that the primary analysis of the
pooled analysis was a random-effects model. The random effects model was done as a
secondary analysis to test the stability of the main (fixed effects) model.
Page 6-91, line 27: There is a typo. "101 ng/dL" should be "10 ng/dL".
Page 6-225, line 27:1 was a bit surprised at the levels the writer considered "relatively low." Is
there any reason to speculate that, consistent with the new findings for the lead-IQ relationship,
that the absence of consistent associations with reproductive outcomes is because most studies
only examined women with higher blood lead concentrations (i.e., there was no true control
group)? This should be considered in the summary.
Page 6-227, line 31: In the review on lead exposure and low both weight, the author concludes
that the existing studies "adequately measured exposure". It may be prudent to consider whether
a single blood sample - whether it was cord blood or maternal blood - is "adequate" for
reproductive outcomes. What are the correlations of serial blood lead levels taken serially during
pregnancy? Is exposure misclassification a potential problem? What about timing of the
measure of lead exposure during pregnancy?
Page 6-233, lines 6-30: The summary should include some discussion about exposure
misclassification, particularly because most of the existing studies relied on a single measure of
blood lead (cord or maternal) to estimate lead exposure.
D-48
-------
Additional Comments of Bruce Lanphear, M.D. [BL] (dated 4/13/2006)
Addressing Comments of Michael Rabinowitz, Ph.D. [MR] (dated 4/11/2006)
Additional Comments on the 1st Draft Lead AQCD by Michael Rabinowitz, Ph.D.
[4/11/2006]
Mike's comments are in bold, followed by my reply.
MR: I want to express my concern that the issue of confounding of lead's effects on child
development, as they appear in the section on the epidemiology of lead's impact on human
health are not fully developed or adequately explored.
BL: I am quite comfortable with Mike's proposal to augment the discussion of confounding, but
it should be balanced. If we discuss the issue of confounding, we should also discuss the
likelihood that we are underestimating lead's true effect by including iron status and the HOME
Inventory. The HOME Inventory is an index of the child's environment constructed using a
variety of variables, including housing condition. Housing condition is highly correlated with
lead exposure.
One additional point that is worth describing is how other "unknown or unmeasured
confounders" may account for the findings at lower levels of exposure. In general, children who
have lower blood lead levels also have fewer environmental insults (e.g., less tobacco exposure,
more nurturing home environment) that typically act as confounders with lead exposure and IQ
scores. Yet the greatest decrements per unit increase in blood lead levels were observed in this
"low-risk" group. Although other types of confounders may be at play, these results reduce the
problem of unmeasured confounders.
Two of the variables that people often raise questions about as unmeasured confounders that
have not previously been accounted for in the lead literature are breastfeeding and mouthing
behaviors. Although one can argue that mouthing behaviors are on the causal pathway, we
explored whether these two variables acted as confounders or altered the results of the Rochester
Lead Study. In unpublished analyses, we showed that breastfeeding and mouthing behaviors
were not confounders nor did they change the results of the study.
Finally, the available evidence suggests that a relatively small number of variables (e.g., HOME
score, SES, birthweight and maternal IQ) are the primary confounders of the lead-IQ relationship
and that including other variables does not appreciably change the estimated lead effect (Tong,
2000; Bellinger, 1992; Canfield, 2003; Lanphear 2005).
MR: Unlike most other environmental pollutants, lead's effects, as we encounter them, are
strongly confounded by other risk factors. The extent of this confounding varies among the
populations studied, depending on the patterns of lead exposure in the particular
circumstances. For example, is lead exposure correlated or not with social class or family
income.
D-49
-------
BL: My impression is that, more often than not, environmental pollutants concentrate among
impoverished communities or minority populations. This is true for biomarkers of tobacco
exposure, mercury and PCBs - all of which are more heavily concentrated in African Americans
(CDC, 2005). The effects of these toxicants on various outcomes would probably be strongly
confounded by other risk factors. In some cases, these environmental pollutants may actually
explain health disparities that are too often blamed on generic surrogate markers such as
socioeconomic status.
It is worth noting that the two prospective studies with the lowest mean blood lead levels
exhibited the steepest IQ-blood lead slopes, but represented two distinct subpopulations (the
Boston Study, which was largely White, middle-class families and the Rochester Study, which
was largely African American, lower income families) (Canfield, et al. 2003; Bellinger, et al.
2003).
MR: It is worth remembering that lead's effects are small, only a few IQ points, typically,
compared to other much stronger determinants, such as parental education or family
income. This is especially true at the lower blood lead levels currently being explored, now
typically below 10 ug/dL.
BL: I am always surprised by the interpretation that the effects of lead are "small." Although I
wasn't involved in lead research during the 1980s, my understanding is that it was commonly
argued that "the effects lead were small or non-existent". Yet during the past three decades
children's mean blood lead levels declined from 15 |ig/dL to 2 |ig/dL. The data indicate that the
impact of this decline in blood lead levels is quite substantial, perhaps by an average of 5 or
more IQ points.
For contemporary children, a 4 to 7 IQ point decrement is associated with a 10 |ig/dL increase in
blood lead concentration (Lanphear, 2005). On a population level, these subtle deficits are
actually quite substantial. Landrigan and others estimated that the annual cost of lead poisoning
in US children was $40 billion per year — which is more than the NIH's annual budget
(Landrigan 2002). Landrigan used a 2.5 IQ point decrement for the first 10 |ig/dL increase in
blood lead concentration, which is an underestimate (see Figures 1-2).
This argument - that the effects of lead are small - also ignores other adverse effects of lead,
from cardiovascular disease to tooth decay and criminal behavior. Although these other effects
are not as extensively studied as lead's effect on IQ, they raise further questions about our
tendency to underestimate the overall effect of lead on human populations. There are two
(unpublished) prospective studies that confirm prior studies implicating childhood lead exposure
with criminal behavior and dental caries. The first shows that childhood lead is a predictor of
criminal arrests and, for males, incarceration. The second study shows that childhood lead
exposure is a risk factor for tooth decay. Thus, while the effects of lead on an individual are
subtle, the existing literature indicates that the effects are quite substantial for several major
public health problems.
D-50
-------
MR: For me, this is especially potentially troublesome because lead is usually considered
as a continuous variable, well measured, while often the other variables may be expressed
as categorical variables, so small differences in these factors, when lumped into the same
category, cause differences in the outcome which could be ascribed to lead. For example, if
the mother graduated only 8th grade or 9th or 10th, they could all be scored as not
graduating high school. I do hope my concern about this is unfounded, and closer
examination will show me worried needlessly.
BL: I doubt that there is anything I can do to convince Mike that he is worrying needlessly. On
the other hand, we could test some of your questions using the pooled data set if it would be
valuable to the Committee. In contrast with Mike, I am actually quite surprised that the studies
show considerable consistency after adjusting for confounders, despite substantial differences in
the sample characteristics and measurement error.
MR: As I tried to explain, I think these circumstances require us to look at each of the
studies to examine the extent of the confounding. For example, in a given study, the model
predicting IQ as an outcome, can be made with and without a lead term; and then seeing if
adding the lead term (and the additional degree of freedom) statistically-significantly
improves the model's goodness of fit. This would tell us if lead is an independent risk
factor. Also, examining the effect of adding the lead term on the regression coefficients of
the other terms would tell us in a quantitative way the extent of the confounding in that
population. For example, does the coefficient for parental education change significantly
when the lead term is added. In this way we can see how free the lead effect is from
confounding, and hence how reliably the lead effect was measured in that study. This
would enable us to put more weight on studies where lead's effects were more cleanly
measured.
BL: As promised, we prepared the attached table at Michael's request (Figure 3). One of the
difficulties with Michael's request is that maternal IQ (which he requested at the meeting) is
acting as a confounder (i.e, alters the lead coefficient by > 10%) in every prospective study. I
also attached a slide with unadjusted estimates and estimates adjusted for HOME Score, maternal
IQ, birth weight and maternal education (Figure 4). The reason we only adjusted for HOME
score, maternal IQ, maternal education, and birth weight in Figure 4 should be evident by
examining Figures 5 and 6. We could explore Michael's new questions using the pooled data set
if it would be useful to the Committee. I could also write up our analysis exploring breastfeeding
and mouthing behaviors as confounders. In the meantime, I attached several tables from the
pooled analysis that we conducted to understand the extent of confounding (Figures 2-6).
References:
Bellinger DC, Stiles KM, Needleman HL. Low-level lead exposure, intelligence and academic
achievement: a long-term follow-up study. Pediatrics 1992;90:855-61.
Bellinger DC, Needleman HL. 2003. Intellectual impairment and blood lead levels.
N Engl J Med 349:500-502.
D-51
-------
Canfield RL, Henderson CR, Cory-Slechta DA, Cox C, Jusko TA, Lanphear BP. 2003. Intellectual
impairment in children with blood lead concentrations below 10 micrograms per deciliter. N Engl J Med
348:1517-1526.
Centers for Disease Control and Prevention (2003) Second national report on human exposure to
environmental chemicals. Atlanta: National Center for Environmental Health. NCEH Pub. No.
02-0716. 257 p. Available: http://www.cdc.gov/exposurereport/2nd/pdf/secondner.pdf
Dietrich KN, Ris MD, Succop PA, Berger OG, Bornschein RL. 2001. Early exposure to lead and
juvenile delinquency. Neurotoxicol Teratol 23:511-518.
Landrigan PJ, Schechter CB, Lipton JM, Fahs MC, Schwartz J. Environmental pollutants and disease in
American children: estimates of morbidity, mortality, and costs for lead poisoning, asthma, cancer, and
developmental disabilities. Environ Health Perspect 2002:110:721-728.
Lanphear BP, Hornung R, Khoury J, Yolton K, Baghurst P, Bellinger DC, Canfield RL, Dietrich
KN, Bornschein R, Greene T, Rothenberg SJ, Needleman HL, Schnaas L, Wasserman G, Graziano
J. Low-level Environmental Lead Exposure and Children's Intellectual Function: An International
Pooled Analysis. Environ Health Perspect 2005;! 13:894-899.
Moss ME, Lanphear BP, Auinger P. 1999. Association of dental caries and blood lead levels. JAMA
281:2294-2298.
Nash D, Magder L, Lustberg M, Sherwin RW, Rubin RJ, Kaufmann RB, Silbergeld EK. 2003. Blood
lead, blood pressure, and hypertension in perimenopausal and postmenopausal women. JAMA
289:1523-1532.
Needleman HL, McFarland C, Ness RB, Fienberg SE, Tobin MJ. 2002. Bone lead levels in adjudicated
delinquents. A case control study. Neurotoxicol Teratol 24:711-717.
Needleman HL, Schell A, Bellinger D, Leviton A, Allred EN. 1990. The long-term effects of exposure
to low doses of lead in childhood. An 11-year follow-up report. N Engl J Med 322:83-88.
Schwartz J. Lead, blood pressure, and cardiovascular disease in men and women. Environ Health
Perspect. 1991:91:71-75.
Schwartz J. Lead, blood pressure, and cardiovascular disease in men. Arch Environ Health. 1995;50:31-
37.
Tong S, Lu Y. 2000. Identification of confounders in the assessment of the relationship between lead
exposure and child development. AnnEpidemiol 11:38-45.
D-52
-------
Figure 1: Estimated Lead-associated IQ Deficits by
Concurrent Blood Lead Concentration
Range of Blood Lead Estimated IQ
Deficit (95% Cl)
<1to30|jg/dL 9.2 (5.7,13.1)
<1to10|jg/dL 6.2 (3.8,8.6)
10to20|jg/dL 1.9 (1.2,2.6)
20to30|jg/dL 1.1 (0.7,1.5)
D-53
-------
Figure 2: Estimated Lead-associated IQ Deficits by
Blood Lead Concentration, 5th to 95th percentile
Range of Blood Lead Estimated IQ
Deficit (95% Cl)
2.4 to 30 |ig/dL 6.9 (4.2, 9.4)
2.4to10|ig/dL 3.9 (2.4,5.3)
10to20|ig/dL 1.9 (1.2,2.6)
20to30|ig/dL 1.1 (0.7,1.5)
Lanphear BP, et al. EHP 2005; 113:894-899.
D-54
-------
Figure 3: Linear Relationship of Full Scale IQ and
Concurrent Lead by Individual Cohort
Cohort
N=1333
Boston
Cincinnati
Cleveland
Mexico City
Port Pirie
Rochester
Yugoslavia
Unadjusted
p(S
-0.97
-0.45
-0.82
-0.02
-0.58
-1.45
-0.11
.E.)
(.34)
(.14)
(.19)
(.23)
(.13)
(.29)
(.06)
P
.006
.004
<.001
.946
<.001
<.001
.046
Adjusted for Maternal IQ*
p(S
-0.85
-0.32
-0.54
0.14
-0.26
-0.94
-0.14
.E.)
(.33)
(.14)
(.19)
(.22)
(.12)
(.27)
(.05)
P
.012
.021
.006
.516
.031
<.001
.006
Figure 4: Linear Relationship of Full Scale IQ and
Concurrent Lead by Individual Cohort
Cohort
Boston
Cincinnati
Cleveland
Mexico
Port Pirie
Rochester
Yugoslavia
p(S
-0.95
-0.39
-0.82
-0.02
-0.62
-1.42
-0.11
Unadjusted
.E.)
(.34)
(.14)
(.19) <
(.20)
(.11) <
(.28) <
(.05)
P
006
004
.001
914
.001
.001
038
Adjusted*
p(S.E.)
-0.61 (
-0.25 (
-0.24 (
0.16 (.
-0.1 3 (
-0.80 (
-0.1 4 (
.34)
.14)
.19)
22)
.13)
.28)
.04)
P
.075
.089
.209
.465
.305
.005
.002
' Adjusted for HOME score, maternal IQ, maternal education, and birth weight
D-55
-------
Figure 5: Univariate Relationship of Full Scale IQ to
Individual Covariates, Controlling for Site
Covariate
HOME Score
Maternal Education
Maternal IQ
Birth Weight
Birth Order
Maternal Age at Delivery
Child's Sex
Prenatal Smoking (Y/N)
Prenatal Alcohol (Y/N)
P
7.31
4.45
7.85
2.01
-1.38
0.99
0.58
-2.20
1.16
p value
<001
<001
<001
<001
<.001
0.008
0.394
0.006
0.20
Figure 6: The Relationship of IQ vs Selected Blood
Lead Indices Adjusted for All Available Covariates
Variable
log blood Pb
HOME Score (z-score)
Maternal IQ
Maternal education
Birth weight (per 100 g)
Birth order
Child's Sex
Marital status
Maternal age
Smoking (Y/N)
Alcohol (Y/N)
Concurrent
(R2=0.630)
p p value
-2.58
4.23
4.77
1.12
1.53
-0.44
0.46
1.15
0.13
-0.08
0.74
<.001
<.001
<.001
0.016
<.001
0.275
0.486
0.259
0.768
0.915
0.401
Peak
R2=0.628)
p p value
-2.79
4.30
4.88
1.14
1.53
-0.44
0.46
1.00
0.18
-0.06
0.40
<.001
<.001
<.001
0.014
<.001
0.275
0.489
0.328
0.675
.938
0.707
Early Childhood
(R2=0.631)
p p value
-2.09
4.32
4.98
1.17
1.48
-0.46
0.44
1.10
0.21
0.03
1.08
0.001
<.001
•s.001
0.012
<.001
0.252
0.511
0.437
0.761
0.974
0.227
Lifetime Average
(R2=0.629)
p p value
-2.97
4.28
4.88
1.11
1.51
-0.44
0.46
0.99
0.17
-0.05
0.81
<.001
<.001
<.001
0.017
<.001
0.275
0.493
0.333
0.701
0.947
0.361
D-56
-------
Dr. Samuel Luoma
February 27, 2006
To: Rogene Henderson, CASAC Chair; Fred Butterfield, CASAC Federal Officer
From: Samuel N. Luoma, US Geological Survey
Subject: Review: Chapter 8
General Comments: The document does cover a wide range of topics with regard to aquatic and
terrestrial effects of lead. But much of the review leaves the perception of a preliminary draft.
Chapter 8 has many redundancies and, more important, insufficient synthesis. Nor does it yet
focus on the aspects of Pb in the environment that are most relevant to atmospheric deposition. I
am not sure that Chapter 8 yet meets the objective cited in our charge materials (italics are mine):
"The purpose of the Lead AQCD is to provide a critical assessment of the latest available
scientific and technical information in peer-reviewed published literature ... effects associated
with the presence of lead in the ambient air" "emphasis is placed on interpretative
evaluation and integration of evidence in the main body of the document"
Specifically:
1. The document recites findings in a selected literature. But it is not a critical assessment,
an interpretive evaluation or an integration of evidence. One way to begin to meet these
goals could be inclusion of an explicit conceptual model for how Pb in ambient air might
affect aquatic/terrestrial ecosystems. No integrative concept is presented verbally or as a
simple model. For example, such a model could show how various processes are
integrated: sources>concentrations>speciation/form>pathways of exposure>processes
affecting bioavailability via each pathway>internal accumulation by different
components of the food web>role of detoxification or resistance>toxicity> adverse
effects at the population/community level>roles of food webs.
2. The document often falls short of clearly discussing how behavior and effects of Pb differ
from other metals. Pb is clearly different in terms of sources, processes determining
environmental concentrations, dispersal, pathways of exposure, what organisms are most
likely to be affected etc. Where data is missing for Pb, specifically, it could be
highlighted.
3. The document does not focus in on how atmospheric sources might differ in their
impacts from other sources. Physicochemical form of inputs could be better considered.
Most important, however, is the type of systems that are vulnerable to atmospheric
inputs: lakes (especially pristine lakes) should be of special concern; coastal zones and
even the open ocean are known to show anthropogenic signals from increased Pb use.
The document should help us understand what is known about Pb in those ecosystems,
and where data is missing.
4. The document does not help us understand what types of organisms might be most
affected by Pb and why. It presents an uncritical citation of selected toxicity studies, but
does little to pull these data together with more biologically-specific information. Metal
contamination eliminates some species while others survive. Understanding even a few
basic things about which are which could be quite helpful. For example, are higher
D-57
-------
trophic level organisms more or less vulnerable to Pb exposure than lower trophic levels,
and why (e.g., Settle and Patterson [?] did some work on this in the 1970s). Are
phytopiankton/plants more likely to be affected before invertebrates? Can we judge this
from existing literature, and if not, why not?
5. Less important, Chapter 8 is kind of irritating in the way it is written. Subject areas,
citations and even full paragraphs are repeated, sometimes more than twice. It could be
cut by l/2 with no loss of content.
Some individual topics are well covered, including analysis of limits to the state of knowledge.
Examples are dissolved speciation of Pb and acute toxicity from dissolved exposures. There is
some synthesis of ecosystem effects observed in field studies, although the tone of these sections
is not especially constructive. But important topics are incomplete in their analysis. For
example:
1. The section on Pb concentrations in surface waters is especially incomplete. The authors
of some of the early chapters in the review taught us that Pb concentrations are very low
in natural waters, if exacting sampling/analytical protocols are followed. Most numbers
on Pb in water are wrong, in fact. In Chapter 8, the NAWQA database is the only
literature used to evaluate Pb in natural waters. Lead analyses from its predecessor
(analyses in the same laboratory), the NASQUAN database, were thoroughly discredited
in the published literature in 1991 (15 years ago). As a result the NAWQA analyses have
very high detection limits (as the document notes, 86-88% of data are below detection).
Furthermore these data are all from streams and rivers, where direct atmospheric sources
are not as important as in lakes. NAWQA data (from my own agency, I might add) are
therefore not especially informative (and perhaps even a little suspect) for understanding
relevance of atmospheric Pb to the aquatic environment. There are many analyses from
the sea and some studies of lakes (e.g., Nriagu et al's study of the Great Lakes circa
1996) in the published literature, that are done by reliable laboratories. But literally none
of these are reported. The failure of USEPA to note the analytical challenges of
determining lead in natural waters is also a serious omission. Finally, by mistakenly
leaving the impression that Pb occurs in |ig/L concentrations in natural waters, the report
disguises the important discrepancy between Ambient Water Quality Criteria and real
world concentrations.
2. Three bioaccumulation strategies are cited (first on p 8-121) "for lead". There is no
evidence of strategy 2 or 3 for lead that I am aware of (and none cited in the document).
This concept applies to all metals. Pb is distinct in its typical strategy (type 1). There are
no uptake rate or loss constants given for Pb; so no coherent analysis of such data is
possible from the report. If such data do not exist it should be explained why. In general
the bioaccumulation section needs to be better integrated with toxicity and detoxification.
3. The absence of critical evaluation in key places leaves a perception of unbalanced
analysis of some important issues; and leaves, un-discussed, contradictions that could be
critical to managing Pb contamination.
a. As mentioned above, Ambient Water Quality Criteria for Pb are in the |ig/L
range; but Pb in natural waters probably never reaches such concentrations except
where near mining activities. Does this mean there are not adverse effects in
nature? Are the studies of Pb effects in ecosystems consistent with that? Or does
D-58
-------
it mean we should take a new look at chronic Pb effects in light of what chemical
advances have taught us about Pb concentrations in the real world.
b. The BLM model is mentioned many times in both the terrestrial and aquatic
sections. There is no BLM model for Pb, as the document notes in one place.
There are also serious limitations to application of the BLM for managing aquatic
contamination, at the present state of knowledge. The discussion of this concept
is overly optimistic and incomplete. Some balance is needed.
c. BCFs and toxicity tests are cited, with no discussion of why there is such
immense variance in both. Some discussion of why Pb toxicity varies from 0.45
|ig/L to 1000's of |ig/L warrants some attention. The USEPA Metals Framework
sees only limited uses for BCF's, but that is not mentioned that I saw.
d. There is literally no discussion of dietary exposure to Pb in the aquatic section; or
trophic transfer. The quantitative importance of dietary metal exposure has now
been well established for -10 years. It is mentioned twice in passing but never
taken seriously. If the data is not available that should be noted.
e. Recommended protective values for Pb in soils and sediments vary quite widely,
illustrated in the tables in this chapter. The statistical approach is examined only
to state that bioavailability is not considered, then it is dismissed. This approach
has advantages (and disadvantages) that govern where and when it is useful.
Those are not discussed. The equilibrium partitioning approach, on the other
hand, is discussed, several times, in much more detail, but with no consideration
of any limitations. For example, the behavior of sulfides is critical to this
approach but is never mentioned (an entire issue of Marine Chemistry was
devoted to this in the last few years, but is not cited). There is no indication of
the scientific furor over the applicability of the SEM-AVS approach in natural
settings (e.g. related to instability of AVS, vertical and spatial variability in AVS,
dependence of results on sampling protocols, role of dietary exposure, role of
experimental protocols like homogenization of sediments, difficulty of relating
geochemical sampling to what organisms actually experience). A balanced
document should give the reader some sense of this debate.
In its incomplete state, Chapter 8, as it stands, will be of marginal value in helping consider the
aquatic environment when deriving air quality standards for lead. The chapter needs to be
completed. That should include a balanced analysis for all topics, explicit consideration of
where lead falls in the spectrum of processes that govern metal effects in the environment, and
explicit consideration of how such processes/principles apply to regulating atmospheric inputs of
Pb to the aquatic environment.
D-59
-------
Dr. Frederick J. Miller
Fred J. Miller, Ph.D.
February 21,2006
Chapter 4. Models of Human Exposure that Predict Tissue Distribution of Lead
General Comments
The author(s) of this chapter have captured the basic information needed to understand the
strengths and weaknesses of the various dosimetry models for lead in humans incorporating the
oral, dermal, and inhalation routes of exposure. The text is well written and easy to follow.
Currently missing from the chapter is "a bottom line" as to which model or models the author(s)
feel would be the most appropriate for use in the assessment of potential risks in humans from
exposure to lead. In addition, a lot of information on model parameter values and variables that
is stated to be available would be well suited to include in an Annex chapter. Specific comments
listed below address the need for technical clarifications or the need to address certain topics or
questions.
Specific Comments
p. 4-3,1. 12
This sentence about large uncertainty being expected to remain is an
overstatement. One is better off identifying the uncertainties and then
using sensitivity analyses and probabilistic methods to quantify their
impact.
p. 4-4,1. 25
Personally, I do not like calling models by the developer's name. The
model on this page was developed by Rabinowitz - it is not the
Rabinowitz Model.
p. 4-7,1. 4
Is the assumption that the central compartment is 1.5 times the volume
of whole blood a reasonable one?
p. 4-10,1. 5
The fact that the O'Flaherty and Leggett models do not specifically
state lead exposure patterns is not a drawback. Moreover, the author
states the IEUBK model includes parameters for handling exposure, but
how good are they?
p. 4-10,1. 14
Provide a reference that applications in risk assessment typically
require models that accurately predict blood lead distributions in the
upper tails of the distribution. One is usually limited by the availability
of experimental data to evaluate the reasonableness of model
predictions.
p. 4-13,1. 13
Why assume that 32% of the inhaled lead is deposited in the respiratory
tract? One can use any one of a number of models to estimate this
amount - e.g., ICRP or the Multiple Path Particle Dosimetry Model
available from CUT Centers for Health Research.
p. 4-13,1. 19
The assumption that lead deposited in the alveolar region is completely
absorbed from the respiratory tract is simply not valid. Macrophage
mediated clearance removes a large amount to the G.I. tract absorbs
about 30% of this amount.
D-60
-------
p. 4-13,1.
p. 4-15,1.
p. 4-17,1.
p. 4-18, 1.
p. 4-18, 1.
p. 4-21,1.
p. 4-22, 1.
p. 4-29, 1.
30
13
5
12
26
18
17
27
p. 4-37
A reference should be provided for the reasonableness of the
assumption that blood lead concentration at birth is 0.85 of the maternal
blood lead.
How reasonable is it to assign the two bone compartments identical rate
coefficients for transfer of lead from bone to plasma-ECF?
The authors should describe what percent the value of 0.7 |ig/dL
distance from the geometric mean represents as a percentage of the
mean.
Remove the duplicate "for estimating"
This is a QA exercise and not a model validation or verification.
Where do the t /^ values come from? Estimated from data?
The authors should provide a cross reference to Chapters 2 & 3 where
airborne concentrations of lead are discussed.
Of the EPA All Ages Lead Model is currently under development, why
is it presented here? The model should be published in the peer
reviewed literature before being used in "prime time". Why does the
model stop at 90 years of age?
Do the authors have an explanation for why the Leggett model
consistently predicts higher blood levels than either the O'Flaherty or
IEUBK models? Which predicts are closest to any available
experimental data?
D-61
-------
Dr. Paul Mushak
USEPA's FIRST AQCD EXTERNAL REVIEW DRAFT: PRE-MEETING REVIEW
COMMENTS AND RESPONSES TO THE PANEL'S CHARGE
Reviewer: Paul Mushak, Ph.D.
I have a number of general and specific comments. General comments address the
organization of the draft and its chapters, the rationale for the organization and chapter
composition, and conceptual issues underpinning the updating since 1986 (in some instances
since 1990). The specific comments are assembled with reference to charges to the Agency's
Panel and stay as much as possible within the overall topics contained in the Agency's 2/15/06
Charge questions to the panel.
Comments are guided by perspective and experience with the preparation of numerous
Federal, international and NAS/NRC expert consensus documents as either coauthor or as a
review panel member/chair. Particular guidance for this Pb draft AQCD is provided from
coauthorship of multiple sections of two previous EPA lead criteria documents, those issued in
1977 and 1986. Comments on Chapter 4, the predictive models of body lead burdens in various
age groups, are guided by membership on the SAB review panel for EPA's All Ages Lead
Model, an independent effort.
I. OVERVIEW AND OTHER GENERAL COMMENTS
A. Overall Draft Pb AQCD Quality, Thoroughness and Organization of Chapters
This draft is an enormous and generally well-done effort with respect to an objective
reckoning for the material that's out there. The count is 1743 pages (v. I + II) without the missing
Chapter 7. It is especially noteworthy given the time constraints under which the Chapters were
done. The Agency and its authors plus reviewers are certainly to be commended for the scope
and interpretive depth of the draft.
Lead research and the resulting global lead literature are seemingly continuous processes.
No critical assessment of this ongoing process or "moving target" by any body of scientists and
policy makers can ever be 100% complete in the dimensions of time and space. Rather, one uses
a selective process of choosing the most well done, the most vetted and the most relevant studies
for evaluation. The EPA has generally hewed to this prioritizing in assembling this first review
draft. NCEA/EPA has done this in two ways in term of format. First, there was a focus on the
main data base and its placement in Volume I. Second, more details and data of an archival
nature were placed in Volume II. The evaluation process at a later administrative and
programmatic step is further filtered through the statutory mandates and regulatory needs of the
Agency.
Purpose of the Updating
The updating rationale was reasonably well presented in Chapter 1. However, that text
dealt largely with its focus on the broad statutory and regulatory requirements of EPA. There are
D-62
-------
requirements for updating in any expert consensus document for lead that are peculiar to lead.
This aspect of the preamble could perhaps be an early part of the integrating synthesis chapter,
Chapter 7. Present place-holder Chapter 7 should actually be Chapter 8, with the Environmental
Effects being the wrap on the supporting Chapters. The Environmental Effects chapter should
not be editorially presented as a categorical afterthought to the main document. More on this is
offered below.
Lead as contaminant and toxicant differs from many other contaminant substances within
a clinical medical and risk assessment framework. That is, lead not only produces risk of disease
as some probability within a dose/exposure-driven epidemiological and statistical significance
framework, but also produces actual and identifiable disease which stratifies into symptomatic/
clinical or asymptomatic/subclinical categories.
It is useful to keep in mind that, before the first modern epidemiological studies of lead
impacts were ever published, around 3,000 papers had been published in the global literature on
clinical lead poisoning and clinically toxic occupational and nonoccupational lead exposures by
the early 20th Century (Blansdorf, as cited in Fee, 1990 and analyzed by Mushak and Mushak,
2000). This reviewer also noted (Mushak, 1992) that in 1991 there were over 1200 entries in
Chemical Abstracts on lead toxicity, lead exposure and lead laboratory measurement. There was
in 1991 over 400 entries in IndexMedicus for mainly human lead exposure and poisoning. Over
600 entries for mainly experimental lead toxicity studies appeared in the 1991 Biological
Abstracts.
Fee E. 1990. Public health in practice: An early confrontation with the 'silent epidemic'
of childhood lead poisoning. J. Hist. Med. 45: 570-606.
Mushak P, Mushak EW. A comparative analysis of the evolution of lead and mercury as
public health hazards. In: 11th International Conference on Heavy Metals in the
Environment (J. Nriagu, ed.). CD-ROM: Manuscript No. 1445. University of Michigan,
Ann Arbor, MI, August 6-10, 2000.
Mushak P. 1992. The landmark Needleman study of childhood lead poisoning: Scientific
and social aftermath. PSR Quarterly. 2: 165-170.
The 1977 lead criteria document dealt largely with manifested disease, overt lead
poisoning. That is, the human impact portions of the 1977 document were guided by the
traditional medical model of physician and patient, presenting the upper end of the full
lead dose-toxic response relationship. However, some of the earlier studies in asymptomatic
children, i.e., studies placed lower on the dose-toxic response curve and yielding to more of a
public health risk or preventive medical model, received some discussion. The mix of dose-
response data for purposes of risk assessment was presented in Chapter 13 of the 1977 document.
The 1977 document was the first to identify the fetus and young children as the subsets at special
toxicity risk from lead. Young children were then considered as the target group for protection
from lead exposure in formulating the 1.5 ug/m3 air Pb Primary and Secondary NAAQS in 1978.
D-63
-------
The four-volume 1986 AQCD, the 1986 Supplement and 1990 Addendum collectively
had much to say about clinical lead poisoning but also included much more published data from
the global literature on sub-clinical lead toxicity. The clinical and public health models were
more equally represented. The latter evolved in concert with evolution of more sophisticated
tools to identify increasingly more subtle toxic endpoints. Many of the target organs and systems
for lead toxicity as now understood to comprise the corpus of lead's toxic injury, including the
critical effect of developmental neurotoxicity in young children, were at least in place in the
1986 AQCD, the 1986 Supplement, and the 1990 Addendum. In the case of developmental
neurotoxicity, cross-sectional studies and the earlier epidemiological data generation from the
prospective studies were becoming available.
Lead as contaminant and toxicant also differs from virtually all other toxic contaminant
substances in the nature and amount of solid scientific evidence establishing the substance as a
very lexicologically potent contaminant. The three pillars of evidence are mechanistic,
experimental and epidemiological. Very few environmental contaminants answer in the
affirmative to as many of the criteria questions for entrance to causality posed by A. Bradford
Hill in 1965.
To date, the various mechanistic and experimental studies on lead are in full accord with
the evidence continuously forthcoming from human population studies. To date, the various
mechanistic and experimental studies keep showing more robust dose-response relationships in
lock-step with the most recent environmental epidemiological studies. The direction of findings
continues to be down the dose-response curve. That is, lead continues to be more potent and
more varied in this increased potency than we had assumed. One need only have a comparative
look at the dose-response data for developmental neurotoxicity and the robust data for the newer
field of childhood Pb immunotoxicity in the current Pb AQCD. I am not aware of any body of
scientifically credible empirical studies of the critical adverse effects of lead and done by
credible scientists that has posited the argument that lead is much less potent lexicologically than
we had assumed.
Given the above historical and contextual points, a critical review of the current Pb
AQCD would entail looking for known developments in the lead field relevant to lead toxicology
and lead exposure assessment and regulation:
—Newer data since 1986-90 identifying more precisely the shape of the lead dose-
response curve in both human and non-human populations for known somatic targets of
lead in vulnerable segments of populations;
—Newer data identifying toxic effects of lead in those organs and systems that were only
first being recognized or were not recognized at all as targets for injury in the 1986/90
documents;
—Newer methodological and interpretive tools that permit a fuller and more transparent
elaboration of more effective assessments of strategies within lead regulation and public
health policies. Overall, the updating draft AQCD for lead has done a reasonably good
job in providing the above.
D-64
-------
Organization and Quality of the Chapters
I don't quite understand the placement of "Environmental Effects of Lead" in this
updating AQCD draft as Chapter 8, after any synthesizing-of-the-material. It really ought to go
as Chapter 7 and the wrap Chapter 7 should then become by definition the last chapter. The
synthesizing chapter, when written, should of course include a section for environmental effects.
Environmental effects, ecotoxic effects, or whatever the vocabulary applied here, are not
inconsequential. They provide the wherewithal for the secondary NAAQS and also are
intimately connected to human impacts in a holistic way such as to define lead as a biospheric
pollutant.
EPA and CASAC clearly consider the environmental effects important. The page count
and subject matter coverage in this draft is much more than in the respective sections in the
previous criteria documents. CASAC obviously considers the Chapter as significant, having
assigned more reviewers (five) to this section than any other in the draft. Secondly, a number of
the Chapters serving as prelude to the human impact chapters, i.e., Chapters 2 and 3, are also
critical to environmental effects. For example, lead emissions, lead fate and transport, lead in
environmental media, are equally important to all receptors. Air lead deposition into terrestrial
and aquatic systems covered in current Chapter 8 is simultaneously a major conduit for pathways
of childhood lead exposure via leaded dusts and soils.
Whither Chapter 7?
I understand the role of this CAS AC/SAB panel to be the usual one: to review the
scientific quality and validity of the material as prepared by EPA in this first draft AQCD for
lead. EPA presents and CASAC proposes. That being the case, the absence of a first-draft
integration and synthesis Chapter leaves the panel in the dark as to what is EPA's take on all the
information. I can certainly understand that time was in increasingly short supply as the writing
of this draft proceeded. That was pointed out.
As I understand the planned preparation of a wrap chapter, this will be done post-first
meeting and presented for separate review at a second meeting. There is some logic in having the
Agency prepare a synthesizing integration chapter from individual sections each of which have
already been evaluated by the CASAC panel. There are worse things for agencies nowadays than
paying attention to scientific advice more than once.
Some might argue that a first-draft synthesis could have been attempted between the
December 1 posting on the internet and the circulation of the chapters to the review panel. I
would argue that a hasty effort would represent little additional progress.
Prior incarnations of EPA's Lead AQCDs did not present this kind of problem and some
value may attach to looking at some of the thinking that went into the wrap chapters in those
documents. The 1977 document had a synthesizing Chapter 13 prepared and reviewed with the
other 12 Chapters by the then-extant CASAC committee twice: June, 1977 and October, 1977.
D-65
-------
The four-volume 1986 Lead AQCD had two layers of overall assessment and synthesis,
all presented simultaneously with the other Chapters to the then-extant CASAC committee. The
principal synthesizing Chapter was Volume I, Chapter 1 "Executive Summary and Conclusions"
that consisted of 198 pp. and a 67 pp. Appendix. A shorter, more focused Chapter 13, Volume
IV, 54 pp, contained an overall summary of data "EVALUATION OF HUMAN HEALTH
RISKS ASSOCIATED WITH EXPOSURE TO LEAD AND ITS COMPOUNDS."
Organization, Variability and Thoroughness Across Chapters
There appears to be considerable variability among the Chapters as to what comprises
updating, what's enough information and what's not enough information, what's the level of
detail that's appropriate, etc. Presumably, this partly reflects differences in opinion among the
Chapter coauthors about the nature and extent of required updating. However, it also clearly
reflects the amount of new information. The problem for the latter then becomes one of how
much to put in the main part and how much in the archival Volume II.
In many cases, the length of the chapters in this draft update are much longer than they
were in the 1986 volumes, meaning that this AQCD is not so much an Addendum as a stand-
alone full document. This reflects changes in the size of the data base, even when there is a
prioritizing of data that is the most relevant to the eventual regulatory needs of the Agency. The
overall length of the sections in Chapters 5, 6, and 8 should not be materially shortened. In some
cases, additional text should be added for thoroughness. More about this in specific comments.
In some cases, there appear to be data sets that are both detailed as to clues about
molecular and cellular mechanisms but perhaps empirically somewhat remote from impacts of
low-level lead exposures at the organismal level. On balance, they should remain for their
expository value.
Authors of a number of the Chapters made an effort to clearly interconnect the previous
1986-90 AQCD material with new information, and a number also first provide a summary of
what the earlier documents had to say about lead toxicity in this or that organ or system. I would
recommend that a standardized format be in place to connect old and new across the Chapters
and sections within each Chapter. The proportion old and the proportion new is certainly going
to vary across chapters and within chapters. For example, the sections of Chapters 5 and 6
addressing immunotoxicity were quite new, reflecting major advances in the field. The topics
were clearly explained and the reader was provided a clear blueprint to go from data to such
take-home messages as dose-response relationships and the insidious nature of lead's effects on
the immune system.
In some cases, the nature of the new chapters has changed because the contamination and
exposure realities have changed. The 1977 and 1986 Pb AQCDs had much to say and analyze
about mobile lead sources, i.e., atmospheric contamination from leaded gasoline combustion.
That focus has now switched to point sources, keeping in mind that all primary sources of lead
past or present have added to pathway reservoirs with long half-lives. In the 1970s, at the time of
the preparation of the 1977 Pb AQCD the diet was a significant pathway for lead exposure of
children, via a combination of lead-seamed can use and contaminated dietary categories. By the
D-66
-------
time of the 1986 Pb AQCD the diet portion of human Pb exposures in the U.S. had declined,
based on the FDA surveys such as those of Pennington and Beloian.
The authors of the main human effects Chapters are to be commended for the level of
detail about toxicological mechanisms. They were often presented in a way that allows the reader
to make sense of the highly diverse phenomenology of lead toxicity. It was also gratifying to see
that there was equivalent appreciation among the various authors of the current thinking about
certain basic mechanisms associated with lead and other toxicants across organs and systems,
e.g., oxidative stress.
The information presented in chapters 5 and 6 also allows the reader to draw comparisons
for similarities in responses for lead across organs and systems. For example, it is striking how
immunotoxicity and developmental neurotoxicity are largely expressed not at baseline levels of
activity but only when exposed individuals are challenged cognitively or immunologically.
Some of the major Chapters are so complete they could serve as an up to date and stand-
alone monograph. This specifically applies to Chapter 6, an excellent body of work. All sections
were quite impressive in Ch. 6, but the bookending sections, biomarkers of exposure and
interpretations of the data, were outstanding.
I found the early, biomarker section of Chapter 6 to be especially clear and well done.
The repeated notion here and elsewhere in the draft of an isolated blood lead measurement not
being a fully effective marker of lead exposure for certain types of studies was especially
gratifying to someone who has been an irritated voice in the wilderness about limitations to
single-shot Pb-B screenings on toxicokinetic and other technical grounds:
Mushak P. 1993. New directions in the toxicokinetics of human lead exposure.
Neurotoxicology, 14: 29-42.
Mushak P. 1998. Uses and limits of empirical data in measuring and modeling human
lead exposure. Environ. Health Perspect. 106 (Suppl. 6): 1467-1484.
The virtue of serial Pb-B measurements in certain settings is becoming increasingly obvious.
A very good treatment was also done for XRF measurements. The distinctions between
methodological and statistical measures of such parameters as detection limits are not easy to
explain but a good job was made of it here. There is an interdisciplinary disconnect between the
experimentalists in biology and toxicology who use instrumentation characterized by empirically
transparent and deterministic detection limits and the data crunchers with a statistical bent, who
are not spooked by the notion of calculations from distributions that include negative numbers. A
caveat here is that the latter is harder to grasp by the public at large.
Some of the key chapters have gaps. For example, Chapter 6 should have a summary
section on dose-response Tables, one for children and one for adults. The Tables would be set
forth as they are elsewhere, i.e., thresholds for onset of each adverse effect indexed as Pb-B.
These Tables have appeared in earlier EPA documents and documents of other agencies, such as
D-67
-------
ATSDR and CDC. One could use shading print for new entries into the Tables from those in the
1986 document.
A second gap is to import the information now in Chapter 5, the Neurotoxicology section,
on lead toxicity in individuals versus groups into the appropriate part of Chapter 6. That material
is out of place where it now resides and belongs in Chapter 6. As I read that material, there is
little that is actually challenging to the material in Chapter 6. Before there were environmental
epidemiologists, there were diagnosticians with their clinical takes on bodies of evidence that go
into informing a diagnosis.
There were also case reports, case series and clinical epidemiology. CDC Statement risk
tables, now partly derived from epidemiological results, are used by clinicians to assist in
formulating a diagnosis. CDC does not recommend that an epidemiological study be done for
every neighborhood from which an elevated Pb-B index case appears. I don't see the logic of
having an important section on what the epidemiological data mean for all aspects of human lead
toxicology off by itself.
A third gap at such places as Chapter 6 is insufficient interpretation of some of the
findings. For example, the discussion of which of the Pb-B biomarkers - lifetime average Pb-B,
high risk age average Pb-B, peak Pb-B, concurrent Pb-B - are more correctly classifying as to
the best dose/exposure index needs more elucidation.
For example, it may not be surprising that concurrent or lifetime average are equally
robust as the exposure biomarker if the former is largely being determined by the latter, at least
in terms of rank ordering. We looked at a five-year follow-up of children's Pb-B levels for a
North Carolina exposure group and found highly statistically significant correlation in rank
order. We could not discern if the preservation of rank order reflected historical body lead
burden input to Pb-B or relative immobility of the children as to residence with resulting
continued external lead contacts.
Otto DA, Robinson G, Baumann S, Schroeder S, Mushak P, Kleinbaum D, Baerton C,
Boone L. 1985. Five-year follow-up study of children with low-to-moderate lead
absorption: Electrophysiological evaluation. Environ. Res. 38: 168-186.
Mushak P. 1989. Biological monitoring of lead exposure in children: overview of
selected biokinetic and toxicological issues. In: M Smith, LD Grant, A Sors, eds. Lead
Exposure and Child Development: An International Assessment. Lancaster, United
Kingdom: Kluwers Academic Press, pp. 129-145.
Chapters in the draft should take cognizance of the fact that, while children's bone lead is
quite labile in the first several years of life, bone physiology as described by O'Flaherty at
various places favors net bone lead accumulation post infancy. This post-infancy and early
toddlerhood onset of bone lead net accumulation is certainly enough that by ages 5 to 7 lead has
accumulated.
D-68
-------
Germane to the above and for inclusion for discussion is that body lead compartments are
all in equilibrium with each other. In particular, bone lead and blood lead are in equilibrium, the
k-1 values for lead resorption differing considerably between cortical and trabecular mineral
matrix, and between basal states and physiological stress conditions. Existing equilibria are the
reason we get "washout" contribution from high bone lead stores newly encountering a central
blood compartment with a declining Pb burden owing to reduced exogenous exposures and for
stable isotopic mixing where measurable with exogenous lead intakes. Equilibria are also the
reason that we find that retired lead workers have their blood lead burden driven by bone lead
accumulation, while active lead workers have their Pb-B mainly responding to their real-time
exposures.
One gap in current Chapter 8 is little mention even in historical terms of the huge amount
of effort that the Federal government put into technical evaluation of the acid precipitation
problem and its interaction with toxic metal and metalloidal contaminants. One product of this
was a huge interagency study some years ago on the numerous aspects of the acid precipitation
problem. That resulted in a multi-volume report that I and others peer-reviewed. Other agencies
made periodic reports to Congress, including the National Institutes of Environmental Health
Sciences, using peer review reports:
Goyer RA, Bachmann J, Clarkson TW, Ferris GF Jr, Graham J, Mushak P, Perl DP, Rail
DP, Schlesinger R, Sharp W, Wood JM. 1985. Potential human health effects of acid
rain: Report of a workshop. Environ. Health Perspect. 60: 355-368.
A clear gap in the document is absence of a full section on the various national and other
studies of lead exposures and various correlates thereof. The obvious place for this section is
chapter 6, between the discussion of Pb-B and other biomarkers and the organ/system specific
human studies of human toxicity. Since the 1986-90 AQCD for Pb, there have appeared
NHANES III, Phases 1 and 2, and NHANES IV, with various interim incarnations of the last
named over recent years.
Some sections need to keep things in context as to the larger picture for the benefit of the
ultimate general reader who is more interested in the regulatory and health policy implications.
Chapter 6 describes the international pooled analysis showing a more robust relationship for Pb-
B versus cognitive decrement for Pb-Bs < 10 ug/dl versus what's seen at > 10 ug/dl. This is by
itself a critical finding, and several reasons for this relationship are discussed in the draft. The
curvilinear (downward) relationship of blood-lead and developmental neurotoxicity over the
whole dose-response spectrum does not materially affect the fact that, as body lead goes up so
does the severity of a particular effect and the multiplicity of effects. The finding of a
particularly robust slope size for Pb-B and IQ decrement at Pb-B levels for one segment of the
full curve does not mean higher Pb-B values don't also induce robust toxic harm. Pb-B values >
100 ug/dl are still associated with acute and chronic encephalopathy and a significant risk of a
fatal outcome. That is, within roughly one order of magnitude of Pb-B values, one goes from risk
of subtle neurobehavioral harm to risk of severe, permanent brain injury and death.
D-69
-------
II. SPECIFIC COMMENTS AS RESPONSES TO THE CHARGE QUESTIONS
QA1: To what extent is the draft document format useful and desirable? Can the structure be
further improved?
The device of placing support details for major studies in Volume II is logical and should
be retained. Reviewers and the interested public may well have questions and differences of
opinion from the authors as to what a certain study says or does not say. In those cases, however,
the material in Volume II is still available. The size of Volume I with the present arrangement is
at the edge of being unwieldy for at least quick review.
Further improvement could include a standard format, as I noted above, of interweaving
the 1986-90 documents with the present draft. At present, various authors have dealt with the
pre-existing material in quite different and somewhat confusing ways. A good model for this
would be a set of bullet statements up front for each Chapter section that concisely captures what
was done in the earlier Pb AQC documents.
A second improvement would be more generous use of tables to summarize materials
that have only moderately changed since the last documents. This would be helpful for the
hematotoxicity portion of Chapter 5.
The document will eventually require an Executive Summary as well as a wrap chapter at
the end. Will the panel in the second round of review have an ExSum to look at??
Q Bl. Does Chapter 2 provide adequate coverage of important chemical properties of
lead...pertinent information on sources ofPb emissions...most relevant data sets...Does
the discussion for point and area sources...estimate uncertainties...emissions by key
industrial sectors...long-term impact of lead in soils, dusts...
The sections of Chapter 2 are uneven. Part of this has to do with changing exposure and
contamination realities. In contrast to what the corresponding sections in the 1986 Pb dealt with,
e.g., mobile Pb sources and lead in diet, the current section needs to address point sources, area
sources and industrial sector sources. For example, while diet Pb was a factor in discussing lead
exposure risks in the 1977 and the 1986 AQCDs, input to today's children's body lead burdens is
much less. However, these reductions are for a centralized food supply. The case with ethnic
groups using folk diet components and remedies from outside the country for diverse ailments is
different, and in U.S. locales where large immigrant populations are clustered, idiosyncratic
sources of lead could be more important sources.
The role of lead-laced soils as a continuing localized source is a critical topic for areas
where no effective soil and dust lead abatements have occurred. The Chapter discussed this issue
in a reasonable way but needed to include more data sets. For example, in Sec. 2.3.3,1 don't
understand why the 1985 EPA document dealing with fugitive dust emissions was not included.
The report, by Cowhert and colleagues from the Midwest Research Institute, is reasonably broad
ranging and uses sets of models useful for predicting air lead levels from both aeolian and
anthropogenic mobilization mechanisms, e.g., vehicular tires.
D-70
-------
Also missing from this section are the various studies of fugitive dust emissions and dusts
as a recontamination source at the large Bunker Hill Superfund site. Dr. von Lindern, who has
been involved in most of the Bunker Hill studies, will probably have much more to say about
what the BH results portend elsewhere.
Also missing from this section are data for tailings piles from mining and milling wastes,
which can heavily impact local communities. In the 1970s and 1980s a number of papers
appeared by investigators of waste piles in Welsh mining areas.
Davies BE, White HM. 1981. Environmental pollution by wind blown lead mine waste:
A case study in Wales, U.K. Sci. Total Environ. 20: 57-74.
Other data sets include those of Van Born et al. and citations contained therein:
Van Born W, Keersmaekers T, Adams F. 1988. Characteristics of resuspended soil
particles with high concentrations of Cu, Zn, Cd and Pb as a function of particle size.
Aerosol Sci. 19: 1287-89.
The U.S. EPA should have fenceline U.S. monitoring data for point source lead
emissions for U.S. facilities subsequent to the 1986 document. The Doe Run smelter in
Herculaneum, MO is still operating and several other primary smelters were still operative at the
time of the 1986 AQCD and some time afterward. It is a gap in the data in Ch. 2.
Some of the passages in Ch. 2 are either misleading or incorrect and should be clarified
or dropped. For example, the first paragraph in 2.3.8 notes that "it is often difficult to determine
the original source of an organism's lead burden." It's more often the case that major sources can
be identified. One uses conventional exposure assessment methodology within the larger risk
assessment paradigms to identify lead sources and pathways. An example is inferential statistical
analyses, such as structural equation modeling. One also uses tracers and in some cases ratios of
stable lead isotopes.
There are data gaps in Sec. 2.3.7: Plant Uptake. The discussion is mainly focused on root
uptake of lead when both foliar and root uptake should be considered. In a number of cases, leafy
crops in the proximity of fugitive dust emissions can acquire lead even in the absence of an
operating emission facility. The authors should check Chapter 8's parallel section, where foliar
uptake is considered in some detail.
Q Cl. Does Chapter 3 provide adequate coverage of pertinent available U.S. information on Pb
exposure routes..environmental lead concentrations...Does the chapter adequately delineate
connections among sources and pathways...Does the chapter adequately identify key sources of
information characterizing the level and distribution of lead soil...near roadways... and
characterizing "background" levels in urban, suburban, and rural pristine areas?
The Chapter generally provides a set of summaries of environmental lead levels, the
interrelationships of one medium to another in terms of the nature and magnitude of lead moving
between environmental compartments, etc. However, there are some simplistic statements, some
gaps, and some misunderstandings.
D-71
-------
One data gap is a set of U.S. studies under the aegis of the National Human Exposure
Assessment Survey (NHEXUS) described by Sexton and others in the 1990s. Some of the study
results were presented in the Journal of Exposure Analysis and Environmental Epidemiology and
elsewhere. The authors should check these out.
Sexton K, Kleffman DE, Callahan MA. 1995. An introduction to the National Human
Exposure Assessment Survey (NHEXAS) and related phase I field studies. J. Exp. Anal.
Environ. Epidemiol. 5: 229-233.
Pellizarri E, Lioy P, Quackenboss J, Whitmore R, Clayton A, Freeman N, Waldman J,
Thomas K, Rodes C, Wilkosky T. 1995. Population-based exposure measurements in
EPA Region 5: a phase I field study in support of the National Human Exposure
Assessment Survey. J. Exp. Anal. Environ. Epidemiol. 5: 327-358.
Lebowitz MD, OP'Rourke MK, Gordon SM, Moschandreas DM, Buckley T, Nishioka
MG. 1995. Population-based exposure measurements in Arizona: a phase I field study in
support of the National Human Exposure Assessment Survey. J. Exp. Anal. Environ.
Epidemiol. 5: 297-326.
Is there any reason why soil lead levels near mining and related sites are mainly from
foreign studies? Table 3.7 shows only two U.S. sites out of 14 presented. There are many more
studies out there than given in this Chapter. Furthermore, many of the foreign sites are of more
interest for historical reasons than for more current lead exposure assessments. The Portuguese
site was mined in Roman and pre-Roman times! Some additional studies should be included. For
example, there are those for the Tri-State mining area (MO, OK and KS). Dr. von Lindern can
provide separately his list of the Coeur d'Alene River Basin studies, including the Bunker Hill
Box site.
Lynch PA, Malcoe LH, Skaggs VJ, Kegler MC. 2000.The relationship between
residential lead exposures and elevated blood lead levels in a rural mining community. J.
Environ. Health 63: 9-15.
Malcoe LH, Lynch RA, Kegler MC, Skaggs VJ. 2002. Lead sources, behaviors, and
socioeconomic factors in relation to blood lead of Native American and White children:
A community-based assessment of a former mining area. Environ. Health Perspect. 110:
221-231.
Murgueytio AM, Evans RG, Roberts D. 1998a. Relationship between soil and dust lead
in a lead mining area and blood lead levels. J. Exp. Anal. Environ. Epidemiol. 8: 173-
186.
Murgueytio AM, Evans RG, Sterling DA et al. 1998b. Relationship between lead mining
and blood lead levels in children. Arch. Environ. Health 53: 414-423.
D-72
-------
Casteel SW, Cowart RP, Weis CP, Henningsen GM, Hoffman E, Brattin WJ, Guzman
RE, Starost MF, Payne JT, Stockham SL, Becker SV, Drexler JW, Turk JR. 1997.
Bioavailability of lead to juvenile swine dosed with soil from the Smuggler Mountain
NPL Site of Aspen, CO. Fund Appl Toxicol 36: 177-187.
Lanphear BP et al. 1998. The contribution of lead-contaminated house dust and
residential soil to children's blood lead levels: a pooled analysis of 12 epidemiological
studies. Environ. Res. 79: 51-68.
Succop P, Bornschein R, Brown K, Tseng C-Y. 1998. An empirical comparison of lead
exposure pathway models. Environ Health Perspect 106(Suppl 6): 1577-1583.
The authors should look at some of the studies cited in the Chapter more carefully. For
example, the several studies by Mielke and coworkers purporting to show that the principal
source of lead in old, inner-city urban areas is gasoline lead have some serious flaws. Most of the
flaws have to do with methodology. That is, the study designs favor gasoline lead over paint lead
weathering as the lead input source to soils and dusts for these particular inner-city, urban
settings.
The evidence for lead paint inputs to dripline soils and soils near property perimeters is
solid. When one examines the Mielke studies for stratifying soil lead relative to proximity to the
drip line, drip line soils are much higher than soil leads elsewhere. In other cases, the
methodology was one of using roadside soil lead as the source variable for gasoline lead input
into soil, but using only housing age as the source variable for lead paint. In the Mielke studies,
exterior lead paint measurements were not taken for direct environmental lead source
comparisons. Without use of exterior paint lead levels and a further variable for lead paint
condition, his comparisons are apples with oranges.
Some of the statements in this Chapter with references to inter-compartmental lead
movement need to be tightened up. On p. 3-1, the first par., and p. 3-14, top, the statements need
clarification. For example, inner-city housing with deteriorating lead paint in the interior and on
the exterior can contribute lead to dust by various mechanisms. The series of studies done by the
Cincinnati group over the years indicate that inner cities, unlike extractive industry communities,
will have contributions to dust lead from both sources. This particular issue of urban lead sources
is discussed at some length in EPA's 1996 Integrated Report on the Agency's soil lead
abatement demonstration project.
In mining, milling and smelter communities, the case is quite different. There are
extensive data showing the significant input of air lead via primary emissions or fugitive dust
emissions to soil lead. These yard soil lead inputs produce dust lead, which then determine hand
lead and, eventually, children's blood lead.
The authors need to make sure their information is current in all cases. On p. 3-23, 2nd
full par., the statement about the primary plumbing solder being 50-50 tin-lead solder seems to
be in error. The previous amendments to the Safe Drinking Water Act banned the 50-50 solder
D-73
-------
for all public supplies, although presumably rural households on private water sources could still
use the 50-50 formulation.
Q Dl. How well does chapter 4...characterize key information on...approaches to modeling of
external lead exposures and impacts on internal Pb body burdens...status of model evaluation
efforts...Does Chapter 4...characterize ability of models to handle the significant factors...
Furthermore, does Chapter 4 identify alternative models to the AALM...
The need for modeling of lead exposures as a risk assessment modality of use to the
Agency is clear from the potential multi-media nature of lead at particular contamination
settings, especially with reference to dust and soil lead sinks that have historically accumulated
from historical ambient air lead deposition over the years.
The need for modeling approaches became clear to EPA and its consulting scientists after
the promulgation of the 1978 NAAQS for lead. Furthermore, a measure of the amount of
progress and scientific sophistication that has occurred in the lead field over the past 28 years
since the last primary and secondary NAAQS for Pb in 1978 were promulgated can be readily
discerned in considering how the 1978 standard was derived by staff of the OAQPS using the
1977 Pb AQCD. That methodology was based on a mix of empirical exposure data, health
guidelines and ad hoc statistical relationships.
The all-important averaging time and form of the averaging was: a maximum arithmetic
mean averaged over a calendar quarter. The original proposal in late 1977 was for an averaging
time of 30 days, based on considerations of then-known blood lead toxicokinetics in children and
adults [42 FRNo. 240, 63076-63094, 12/12/77]. However, the final ruling entailed a selection of
a calendar quarter [43 FR No. 194, 46246-46277, 10/5/78]. The shorter the averaging time and
the more stringent the form, the less in the way of exceedence extent or frequency is permissible.
The chapter does a reasonable job of describing the construction of the models, their
ability to deal with predictive modeling requirements in actual use, etc. The Chapter also does a
good job of model comparison. However, I do not understand what the Agency is actually
supposed to do with this modeling chapter. The Chapter offers descriptions and model
anatomies, but there remains the question of what happens with whatever the review produces.
The historical models are not currently in use and are not likely to be used. The IEUBK
model, one confined to childhood lead exposure risk assessment, is in routine use and has been
the standard risk assessment tool of EPA for numerous risk assessment and risk management
decisions at many Superfund sites.
The IEUBK model has been reviewed by two CAS AC committees, in 1989 and 1991-92.
I was one of the ad hoc advisory members of both of these committees. Those committees found
the IEUBK model sound for use and they approved applications for use in communities with
lead point sources and use at hazardous waste sites, such as Superfund sites. Since the findings of
those committees, the IEUBK model has been heavily-evaluated as to source code validity and
field applications. In fact, a whole EPA symposium in 1996 with a monograph (EHP, v. 106
Suppl. 6, December 1998) was given over to modeling of exposures, with a significant focus on
D-74
-------
the IEUBK model. The IEUBK model has also been enshrined in numerous Agency policy
decisions over the years, such as those issuing from OSWER.
There is arguably not a lot the present CASAC committee can or should do at this point
regarding assessment of the IEUBK model beyond acknowledging the reasonably satisfactory
performance of the IEUBK model in the field. Arguments have been made over the years for
revising the model to make wider use of probabilistic rather than central-tendency deterministic
values in the exposure module. A battery of probabilistic components obviously only works with
a lot of hard data available, otherwise uncertainty reduction remains problematic. One cannot
even begin to think about probabilistic modeling of each of the numerous biokinetic components
in the biokinetic portion of the IEUBK model. At present, there is some provision for
accommodating inter-child variability in the Pb-B output.
I don't believe at this time that the direct use of the O'Flaherty or Leggett models as
stand alone entities are feasible operationally or computationally in terms of evaluation and
validating for wide use in risk assessment of lead exposures. Neither is ready for prime time and
these two models are currently little more than computational and conceptual templates for the
All Ages Lead Model (AALM). The relative status of these two models as stand-alone systems
has been addressed by the SAB AALM panel.
The SAB AALM panel was charged with evaluating the first full iteration of the AALM
model and its associated manual. It was never presented with the question that is being presented
to this panel of modeling reviewers: If the AALM model is not ready to go, which extant models
would serve for use until that model is user-ready? Had the question been posed, I fully expect
that the panel would have responded. Although some of the SAB AALM panel members are also
on this CASAC committee, I'm not clear on how or why the present panel would be any better
equipped to answer the charge questions than the AALM panel would have been.
Turning to the details of Chapter 4, Table 4.1 material does not match its text, on p. 4-13,
1 st full par.
There appears to be some confusion about whether it is erythrocytes or whole blood that
begin to show curvilinearity with lead uptake at a lead level of 15-20 ug/dl. The literature is clear
on this point, as pointed out in the 1986 AQCD: Chapter 10. Whole blood lead is linear versus
daily Pb uptake up to about 40-50 ug/dl whole blood (RBC level of 100 to 125 ug/dl, assuming a
Hct of 40%). See the cites DeSilva, 1981; Manton and Cook, 1984; Marcus, 1985a,b,c in the
1986 document.
QE1. Have any important new animal toxicology studies been overlooked in Chapter 5?
The Chapter's sections on lead and experimental animal organs and systems seems to be
reasonably complete in terms of the studies that have appeared since the 1986 AQCD for Pb.
Generally, the same can be said for the in-vitro studies cited. I believe the question of
completeness in data should be mainly reserved for the principal systems that show the most
robust dose-toxic response relationships, since those are the systems that will principally figure
in the eventual use of the data by the Agency in consideration of changes in the NAAQS: that is,
the developmental neurotoxicological, the immunological, and reproduction and development.
D-75
-------
Some of the Chapter 5 sections are relatively new compared to 1986, e.g.,
immunotoxicity, while others have had much of their basic toxicological data presented in 1986,
e.g., the hematological effects.
One format change that would make it easier to sort out thoroughness of data gathering is
to have a standardized way for presenting human, in-vitro and animal data. As it is now, there's a
range of organizations that are quite confusing. In some cases, human and animal data are mixed
within subsections. A second recommendation is to make use of as many tables as possible and
discuss in text only the data most useful for highlighting.
The major studies that deal with multi-system toxicological mechanisms that are critical
for understanding the toxicological phenomenology are well represented in the
neurotoxicological, renal, cardiovascular, immunological and other sections.
QEla. Discussions in the neurotoxicology section focus mainly on lead effects on...Are there
other pertinent areas missing or not adequately covered?
One topic within toxicological mechanisms that could be included in the neurotoxicology
section is the whole multi-organ/multi-system issue of oxidative stress as a potent injury
mechanism in multiple organs and systems including the brain. For example, Vaziri and Ding
(2001) showed significant decrease in NO availability in the brain, kidney, aorta and heart of
rats. They also showed NOS was increased in cerebral cortex and brain stem. The anti-oxidant
enzyme Cu, Zn-superoxide dismutase was significantly increased in brains of Pb-Treated rats
(Vaziri et al., 2003).
QElb. To what extent does the...literature provide evidence for developmental lead toxicity
having a permanent impact on bone and teeth...and for these tissues serving as storage pools...
Beyond infancy and toddlerhood, when bone lead is less labile toxicokinetically, net bone
formation with lead deposition occurs. We know from diverse data that in older children and in
adults, bone lead is a major store whether we are speaking of non-occupational or occupational
populations. Bone lead release occurs as a consequence of lead following, as Aub noted in 1925,
"the calcium stream." It is not necessary that there be deranged mineral metabolism. Bone lead
release can occur in quite young individuals with any history of lead exposure. For example,
Markowitz and coworkers showed that immobilization of children during hospitalization
produces significant elevations in Pb-B, concomitant with the expected demineralization of
skeleton and lead resorption. Numerous studies now document releases of bone lead under
physiological or life cycle conditions, such as those of Gulson et al. using stable Pb isotope
mixing for study of pregnant and lactating subjects and the postmenopausal women with
osteoporotic changes by Silbergeld et al., 1988.
Is lead osteotoxic as well? Does Pb osteotoxicity in turn affect bone's role in body lead
storage? The former Q is clearly answerable in the affirmative, and in the case of experimental
animal studies, there are long term effects from chronic lead exposures. But it is not clear what
level of persistence or actual irreversibility of osteotoxic effects in humans occurs. One body of
older data we have for lead-induced skeletal mineral derangement in humans are numerous
D-76
-------
reports in the time of much higher lead exposures of the presence of "lead lines" in older
children. These are not lead deposits, but are multiple banded areas of dense mineralization at the
metaphyses of the growing long bones.
Elc. Are the animal studies with chelation agents relevant to studies in humans?
There is potential virtue in studying the use of chelating agents in animals in terms of
relevance to humans, but there are some practical considerations. First, we know from the multi-
center TLC studies of the use of Succimer as an intervention agent to attempt reversal of
neurobehavioral deficits that Succimer intervention does not prevent (apparently irreversible)
deficits (Rogan et al., 2001; Rosen and Mushak, 2001).
Different agents have different modes of action and it would be difficult to model lead
exposure changes in the presence of chelant use.
Ele. Does the oxidative stress theory appear plausible for lead toxicity and represent a common
mode of action across species and organs?
The oxidative stress theory is not only applicable to lead toxicological mechanisms but a
number of the sections of Chapter 5 appropriately discuss the elements of injury from, e.g., ROS.
Elf. Concentrations ofPb compounds used in animal toxicology...often appear high...what
advice can the panel give...to identify a cut off value...?
That's a peculiar Q, albeit an interesting one. First, doses for animal testings are high
because one wishes to first characterize the presence of an overt effect at the high end of the
substances's dose-response curve. This has been a dictum of toxicology for decades. One then
creeps down the dose-response curve, in the case of noncarcinogens, until one hits LOAEL and
NOAEL values, with or without enough data to generate Benchmark Dose values.
Second, different organs, systems, and species have different sensitivities to lead
dose/exposure. It's not possible to readily answer the question because there's not a one-
experimental-size-fits-all answer.
Thirdly, we have plenty of epidemiological data for human risk populations as
extensively set forth in Chapter 6, so that it's not clear what we want at this stage of the field
from animal data in terms of QUANTITATIVE DOSE-RESPONSE yardsticks as compared to
QUALITATIVE MECHANISTIC INSIGHTS. The Lanphear et al. international pooling results
show little evidence for a threshold in neurotoxic effects, especially those evaluated in that
analysis. The association extends down to around the PRACTICAL QUANTITATIION LIMIT
(PQL) of a reasonably good clinical laboratory, 1 ug/dl or so.
Fourth, there are potential inherent limitations to extrapolation of effects. In the case of
inorganic As (As-i), for example, there are few good animal models for dermal and internal
organ carcinogenesis in humans, even though we know As-i is a potent human carcinogen, with
extremely low measured water levels of As-i associated with cancers (e.g., NAS, 1999 and
2001). Some have argued that the absence of a good animal model for various As-i effects raises
doubts about its human toxicity. This ignores the obvious point that animal exposures are mainly
D-77
-------
used as surrogates in the absence of data for human exposures and not vice-versa. What's more,
EPA is usually in the business of using animal data to protect humans, and not vice versa.
Q Fl. Does Chapter 6 adequately address the issue of dose/exposure metrics for use in human
population studies?
Chapter 6 did a very good job of describing the pros and cons of using the various
exposure biomarkers. I have already made comments elsewhere about the dose metric portion.
One point for further discussion, beyond the differences in using the various metrics, is
their relative status in quantifying dose-response relationships. For example, Pb-B is the
dose/exposure gold standard for dose-response relationships whatever its toxicokinetic and
interpretive quirks.
What type of bone lead measurement does one do for bone lead-toxic response
relationships? What level in which type of bone corresponds to a scaled series of toxic
responses? The use of bone lead has to spread to beyond its use in a research setting. The VA
Normative Study has produced a number of studies but the technique will require much more
quantitative characterization and diversity of use to be accepted as a routine tool.
Q F2a. Does this Chapter adequately address questions regarding the significant neurotoxic
effects observed at low blood Pb levels...Also, is the issue of model selection adequately
discussed?
Yes, the Chapter adequately addresses a number of Qs regarding the lowest threshold for
neurotoxic effects. The authors have presented several explanations for the overall curvilinear
nature of the dose-response curve, with the lowest segment of Pb-B having the most
comparatively robust effect.
Does the shape of this segment of the overall D-R curve reflect a more biased loading of
misclassification of exposure into this lowest segment compared to higher Pb-b ranges?
On toxicokinetic grounds and the available literature on the temporal behavior of Pb-B in
significantly exposed children, any explanation for the more robust lower segment is not readily
forthcoming if the definition of misclassification is an erroneously assumed brain lead burden
based on its surrogate, Pb-B. The lower the Pb-B, the lower the starting level of brain lead
burden relative to brain Pb burdens in other children in the study with higher Pb-Bs, assuming
that brain Pb half-life of clearance does not change with absolute levels in the tissue. That
applies as well with cumulative lead burdens.
The lower the starting Pb burden in brain, for a given assumed half-life of decay, the
more rapidly Pb-B rain reaches a potential sub-effect threshold value. Succop et al. (1987)
reported that for the Cincinnati study cohort with quite elevated Pb-Bs in their earlier years of
testing, i.e., > 10 ug/dl, the half-life of Pb-B was 10 months, compared to the more typical value
for either low chronic or acute exposures of 20-30 days. The difficulty of ascribing this extended
half-life to body lead burden is the inability in this and other cases to stratify the more persisting
Pb-B elevations to endogenous or "washout" lead and simply ongoing elevated exogenous lead
D-78
-------
intakes. If however, the half-life of lead clearance from brain is not independent of absolute Pb-
brain level, e.g., dose-dependent induction of Pb nuclear inclusions, then things are even more
complicated.
The question of model selection as a determinant is a biostatistical one, and I am not able
to comment.
Q F2b. Does the summary section of Chapter 6 adequately and correctly reflect the
content of the Chapter?
The answer is generally yes but some fleshing out of the summary would help. In fact,
the language of summary is quite conservative, i.e., does not overreach. It is clear from this
chapter and its reasonably correct reflection of the extant literature that lead injures multiple
systems and does so through a multiplicity of mechanisms. Some of these mechanisms are
insidious. This certainly characterizes those subtle effects that only become apparent when the
organism is stressed in some way, whether it's an infectious episode, a mathematically complex
test question at school, etc. More should be said about the need for test paradigms of lead
neuroepidemiology and lead immunotoxicity to make sure challenges of the resting system are
robust enough to see things start to fall apart.
Q F2c. There is concern as to what level of change for various endpoints constitute
adverse effects. What are the Panel's thoughts?
The argument about what defines an adverse effect has very long legs. The matter has
come up in every lead document preparation and review and every other consensus document
that I have been involved with. There are various answers.
First, one can use the definitions set forth by the National Academy of Sciences back in
1977. In paraphrase, adversity of effect is defined by either some impairment of optimal function
or reduction in reserve capacity to sustain optimal function. Note the operative term is optimal,
not minimal. For all practical purposes, the definitions find themselves in the realm of low-level
toxicity or sub-clinical effects in asymptomatic children.
Secondly, there is the matter of diagnostic and assessment tools to evaluate lead's toxic
effects and the conclusions to be drawn from there. If one asked what is the principal adverse
effect of lead in children in 1910, the answer would be "death" (Blackfan and Thomas, 1914). If
one asked the Q in 1925, the answer would be "encephalopathy with coma and seizures"
(McKhann, 1926). In the 1940s, the answer would be "recovery from acute poisoning but with
severe cognitive and behavioral sequelae" (Byers and Lord, 1943). And so forth.
Thirdly, there is a societal dimension, i.e., societal expectation, to the definition. If one
told a group of parents, at a meeting to discuss the hazards to their children of a lead operation
next door, that their children would sustain a large drop in IQ but not enough to be considered
clinical retardation, i.e., there's no problem, one can imagine the resulting response.
The notion of adverse effect also figures in the use of biomarkers of effect. European
researchers and clinicians are much more cautious, on medical ethical grounds, that routine
D-79
-------
biomarkers of effect be markers of earliest, biochemical effect without the biomarker being an
index of an adverse effect. This has been due to the heavy influence of Lars Friberg and
colleagues at Karolinska Institute, who produced a 1983 monograph on what are and what are
not appropriate biomarkers.
Q F3. Are the discussions of genetic and other susceptibilities adequate?
They are generally adequate in that for cases of genetic susceptibility there are still Qs
about what markers are in place to show such heightened sensitivity to lead. The ALAD allele
population distribution case is telling. Elevated Pb-Bs in people with the ALAD 1:2 or 2:2 vs. the
1:1 variant may not mean more risk of toxicity but less. That is, what's significant is the
tightness of Pb binding and removal from the vulnerable target organ pools until erythrocyte
turnover occurs.
Mention should be made of the opposite direction in two phenomena: Pb-B is declining
while the incidence and prevalence of childhood asthma is increasing. If lead is a factor, why is
this? I believe that one matter that should not be overlooked is that those subsets of children with
the highest increases in asthma rates are those whose numbers have been most refractory to Pb-B
declines: inner-city, low-income minority children whose economic straits have been getting
worse in the last number of years. Children forced into housing of ever worsening quality, for
example, encounter more asthma risk factors.
Q G. Comments and suggestions for Chapter 7
It may be difficult for the panel to come up with a list for Chapter 7 until the review of
the content chapters is finished. This may be something that could be discussed with a scheduled
teleconference between Rounds 1 and 2. See, also, earlier comments about the synthesis sections
in the previous AQCDs for Pb.
Q H. Does Chapter 8 cover the most relevant issues for environmental effects?
I have made a number of comments on 8 earlier in this pre-meeting review.
POST-MEETING COMMENTS - EPA PB AQCD
Reviewer: Paul Mushak, Ph.D.
March 6, 2006
I have a number of post-meeting comments. They cover both the assigned Chapter 4 and
the remaining Chapters.
In addition to these post-meeting comments, I will be providing a short set of bullet lines for
consideration for the letter to the Administrator.
D-80
-------
Chapter 4 Comments
The panel's discussion of the models generally indicated that, for the foreseeable future,
the IEUBK model is the one farthest along of the four non-historical models discussed. The All
Ages Lead Model (AALM) is still early in development and has little near-term readiness for use
in EPA health risk-based assessments of Superfund sites and other sites of concern. The AALM
model is also currently being reviewed by a separate panel of the SAB. The O'Flaherty and the
Leggett or Pounds Leggett PB-PK models as stand-alone Pb exposure predictive models still
need work. This leaves the IEUBK model for the near term.
A number of panel discussions occurred about the desirability of modeling dust Pb as
lead loadings, given that studies have shown that dust Pb as loading is a more robust predictor of
dose for dose-response relationships than dust Pb as concentration. I would also note that even
moderate levels of lead in air can produce a surprisingly large impact on lead exposure pathways.
This occurs for the very young child via dust lead loadings onto interior hard surfaces and also
Pb loadings onto exterior hard surfaces, e.g., toy surfaces, outside picnic tables, etc.
I have some comments about the issue of lead loadings and any subsequent use in
modeling. It is important to note that in the present context we are speaking of dust lead loadings
to hard surfaces contacted by children. Dust lead loading and loading rates can also be applicable
to soils, but that intermediate medium is not directly at issue for the loading question in terms of
dose robustness in dose-response relationships.
All three of the models at issue are constructed to have their biokinetic components
crunch on Pb input, which presently is the product of Pb daily intakes multiplied by uptakes
(absorption, bioavailability) summed across media. To anthropomorphize the models, they don't
care what pre-biokinetic calculations in the exposure portion occur that eventually produce
intakes of lead per unit time. Lead intakes from dusts for subsequent uptake and biokinetic
disposition, for example, can theoretically take the form of intakes derived from Pb as a
concentration term multiplied by intake mass or as a lead loading per unit area multiplied by one
or more parameters that govern transfer of amounts of lead from a confined hard surface area to
the mouths of children for subsequent ingestion.
None of the three more finished models are currently constructed with an actual way to
handle dust or soil Pb loadings, i.e., Pb level-per-unit-area, as a means to compute eventual daily
dust Pb intakes into children's blood. In the IEUBK exposure module, in particular, intakes of
dust Pb are derived via concentrations x intake masses. How does one derive a Pb loading rate
and then get from a hard surface Pb loading, or Pb loading rate, to a daily lead intake? There are
calculations one can make in several stages.
EPA's 1996 Integrated Report for the three-city soil Pb abatement demonstration project,
on p. 2-25, provides the specific and simple calculation (Eq. 2-3) for obtaining dust Pb loading
rates per annum for an air Pb level of 0.1 |ig/m3 and a reasonable estimate of the deposition rate
in this setting of 0.2 cm/sec. From this specific expression, one can obtain the daily dust Pb
loading rates in the usual metric of m2 for any Pb-air value where the deposition rate would be
reasonably valid:
D-81
-------
1 x 10 7 jig Pb/cm3 * 0.2 cm/sec * 3.15 x 107 sec/year = 0.6 jig Pb/cm2 * year
U.S. EPA. 1996. Urban Soil Lead Demonstration Project. Volume I: EPA Integrated
Report. Report No. EPA/600/P-30/001aF.
For a hard surface, the result is 6,000 |ig/m2/year, or 16 |ig/m2/day daily dust Pb loading rate for
an air level of 0.1 units. In units of ft2/day, for cumulative loading comparisons in the units of the
EPA Lead Hazard Rule standard for floor loading, 40 |ig/ft2, this is 16.0/10.8 = 1.5 |ig Pb/ft2/day.
The EPA dust lead loading standard is for interior floor surfaces. The standard relationship of
outdoor to indoor air lead level is:
Pb-air, interior = Pb-air, exterior * 0.3
(see p. 2-28) U.S. EPA. 1994. Guidance Manual for the Integrated Exposure Uptake
Model for Lead in Children. Report No. EPA 540-R-93-081).
U.S. EPA. 2001. Lead; Identification of dangerous levels of lead; Final rule: 40 CFR 745
[66 FR 1205-1240; January 5, 2001].
An interior floor dust loading rate requires a calculation for interior Pb-air when ambient outside
Pb-air is X by simply multiplying X by 0.30
The attached Table 1 presents the daily Pb loading rate for interior and exterior hard
surfaces as a function of Pb-air along with loadings for a sample time interval, 90 days. The dust
Pb loadings over 90 days are quite significant. Table 2 provides the relationship of daily dust Pb
loadings to the EPA floor dust lead loading standard with respect to time periods (days) required
to reach the standard.
The current draft of the AALM model has, in its exposure module for dust Pb, the choice
of using dust Pb as either concentration or loading. The model also has a parameter selection for
"contact rate" as a function of child age if one selects the loading option. The contact rate as
specifically derived for this model apparently links the loading per unit area to an ingestion total
or rate by incorporating estimates of such physical and behavioral parameters as fractional
transfer of surface Pb to hand per pass, as fractional transfer of hand Pb to the mouth via
mouthing per event, and the frequency/day of these passes for a locale with the indicated surface
Pb loading. This is surmised from the current draft of the AQCD, Chapter 4, on p. 4-31, which
notes "...the lead ingestion rate being calculated as the product of loading and contact rate." In
the actual AALM model, I found that this calculation in the exposure module appears to be
disabled.
The contact rate in the AALM model differs with child's age, as would be expected. The
drop-down dust alternative menu in the model is presented as Figure 22 in the AALM draft
manual.
I would note that the question of transfers of contaminants from surfaces to mouths of
young children has been of considerable interest to various agencies. For example, the Consumer
D-82
-------
Product Safety Commission (CPSC) has developed protocols for testing Pb intakes from a
combination of surface wipe data and Pb transfer to children's hands and subsequent ingestion.
EPA's pesticide regulation program and the CPSC have also been looking at children's hand
contact and transfer data for such exposure settings as contact with play products built from
chromated copper arsenate (CCA)-treated wood.
The IEUBK model is only applicable to lead exposure modeling for children out to
84 months of age. It does, however, have the virtues of extensive evaluation and calibration and
having a current routine risk assessment role for lead at such locales as communities impacted by
Superfund sites. All 10 EPA Regions routinely use the model in their Superfund programs and
have been doing so for over 10 years. A quick scan of completed Superfund action documents
involving the model's uses can be done via the Regional web sites. They are numerous.
I and a number of others, both on the Ch. 4 subgroup and other members, had considerable
difficulty as to (1) what bodies of empirical data or combinations of empirical data and modeling
approaches should weigh proportionately more significantly in the thinking of EPA with respect
to addressing NAAQS review, and (2) how would the panel's time at the meeting and afterward
be best used for this particular purpose.
One area of focus with respect to tandem use of modeling and epidemiological data
should be the downstream step where distributions above and below a target Pb-B level would be
determined so as to comply with a selected percentile protection cutoff. In the IEUBK model, the
current Pb-B value to be avoided for 95% of affected children is 10 |ig/dl. That is, EPA's current
practice is to use a protection percentile, 95%, whereby remediation actions will be such that
95% of all children will be below the criterion value of 10 |ig/dl.
The thrust of the current data is that the target Pb-B for applying the protection criterion
should be well below this current level of 10 |ig/dl.
The question of CDC views on sub- 10 |ig/dl thresholds for effects and the relative
feasibility of acting on those findings are noted for Ch. 6.
Consideration should also be given to the dust Pb impact, where the dust Pb loading is
related to air Pb.
Other Ch. 4 Comments
Some specific concerns were expressed about Chapter 4. Dr. Cory-Slechta questioned the
source of the half-life of lead in the brain being two years. That figure is from the Leggett, 1993
paper, which cites earlier data in experimental animals. I also found this half-life to be perhaps
longer than expected. This matter of brain Pb Ti/2 is important because it may influence the
interpretation of such outcomes as persisting effects, the most robust indicator of neurotoxic
injury, etc.
One factor in assessing Pb levels and Pb changes in tissues over time and dosings is a
methodological one. We found some years ago in our animal studies of lead distributions in
tissues of rats and other animal models that tissue levels showed changes with perfusion of the
D-83
-------
animals to avoid occluded blood. Blood of high Pb content retained in target tissues would affect
both measured absolute levels as well as rates of change with changes in exposure.
Dr. Miller was concerned about the source of the pulmonary compartment deposition
figure in the IEUBK model of 32% for children. The lung deposition figure of 32% is based on
the data in the 1989 Staff Paper of EPA-OAQPS:
U.S. EPA. 1989. Review of the National Ambient Air Quality Standards for Lead.
Exposure Analysis Methodology and Validation. U.S. EPA Office of Air Quality
Planning and Standards, Research Triangle Park, N.C. Report No. EPA-450/2-89/011.
That source noted child lung deposition of Pb as 25-45% for children living in non-point source
areas and 42% for those living near point sources.
This document badly needs a section where the whole issue of bioavailability, i.e.,
uptakes from the receiving body compartments of both ecological and human risk populations
can be presented. There is the expertise among the authors to do so. The topic is sprinkled
throughout the document with little emphasis. Chapter 4 is the place to do it. The material is
largely present, just scattered. I would especially urge that a section of Chapter 4 be given over
to the various determinants of Pb bioavailability: biochemical, physico-chemical, anatomical,
physiological, thermodynamic, etc.
Mushak P. 1991. Gastrointestinal absorption of lead in children and adults: Overview of
biological andbiophysico-chemical aspects. Chem. Speciation Bioavail. 3: 87-104.
Comments on Other Chapters
Ch. 6
I like the notion of discussing further the distinction between the evidence of sub-10
|ig/dl effects and the conundrum of what to do about these findings. CDC has had the problem in
its most recent statements of saying there are low-level lead exposures that produce effects below
10 units in Pb-B, while simultaneously and passively holding that the reduction of such
exposures is problematic. This is an odd approach for a health agency. CDC should simply and
actively argue for primary prevention approaches to remediate low but still potent lead
exposures.
Dr. Schwartz, in the quite appropriate mercantile vernacular, likened agencies using lead
exposure data to marketers of data and actions of several types. CDC can be considered as a
retailer and EPA as a wholesaler of not only health data but their disposition. This is critical to
keep clear. I would note that we now have such low levels of lead exposures causing adverse
effects, as evidenced in current studies, that what is required for intervention in hazard reduction
are primary prevention modalities, not secondary ones.
Rosen JF, Mushak P. 2001. Primary prevention of childhood lead poisoning: The only
solution. N. Engl. J. Med. 344: 1470-1471 (editorial).
D-84
-------
EPA, by statute, is a regulatory agency and therefore has the means to implement primary
prevention modalities, i.e., it wholesales primary prevention to others. CDC, by charter, is a
health advisory agency with no primary prevention powers and the burden of having to deal with
secondary prevention realities. That is, CDC concerns itself with children already lead-exposed
enough to have elevated Pb-Bs. Such children also are recognized by CDC as justifying primary
prevention without having the authority to dictate primary prevention. EPA, along with the U.S.
HUD, the U.S. FDA, and U.S. CPSC are empowered to implement primary prevention
approaches to dealing with sub-10 jig/dl effects. Just because CDC has to rely on regulatory
agencies to implement primary prevention approaches to reducing child lead exposures is no
reason it cannot aggressively provide and push the rationales for primary prevention to such
regulatory agencies as EPA to begin with.
Chapter 8
The section on amending soils with phosphate to immobilize Pb should be expanded to
include a discussion of limits to phosphate amendments where lead co-occurs with arsenic (As)
in contaminated soils. It is often the case that extractive industry emissions of lead to soils are
accompanied by co-deposition of highly carcinogenic inorganic As (As-i). Aerated soils contain
As-i as the pentavalent form, arsenate. Phosphate would mobilize arsenate through competitive
binding, creating the risk of moving arsenic down into groundwater. As-i in drinking water is a
particularly hazardous combination, in that As-i in water is a potent human carcinogen. See the
1999 and 2001 NAS/NRC reports on the topic.
Synthesizing Chapter
This chapter is where the massive amount of data in the draft chapters is winnowed down
to those key aspects of the total data base in the previous chapters and the key data for further
Agency analysis that are relevant for the low-level toxic effects and exposures associated with
them are selected. The guidelines for doing the synthesis chapter spring from the various critical
comments made by the panel members at the meeting.
First, the most holistic synthesis of the quantitative and qualitative realities of the current
environmental compartmental lead inventories (in terms of multi-media exposures of children
and other risk groups) should be done and made clear. That is, the extent of permissible further
Pb inputs to human exposures must account for media-specific Pb accumulations.
The legislative history of the control of lead and other contaminants has typically made
for piecemeal or media-specific controls. However, toxicology and biology integrate total intakes
into a collective measure of overall risk.
Mushak P, Schroeder C. 1980. Multi-media Environmental Pollutants. A Report to the
National Commission on Air Quality, Washington, D.C., December, 1980. Summarized
in: To Breathe Clean Air, Report of the National Commission on Air Quality,
Washington, D.C., March, 1981.
D-85
-------
Schroeder C, Mushak P. 1980. The Legislative History of Multi-media Pollutants. A
Report to the National Commission on Air Quality. Washington, D.C. December, 1980.
Summarized in: To Breathe Clean Air, Report of the National Commission on Air
Quality, Washington, D.C. March, 1981.
Then, the most reliable data from mechanistic, experimental and epidemiological studies
that define the lowest levels of Pb-B associated with adverse effects should be retained for
further assessment. This should be made easier by using chapter-specific Executive Summaries.
Some summary statements should be used to wrap up the rest of the studies.
Again, the synthesis chapter should be placed after the environmental effects material,
and include summary analysis of those effects. The recommendation for use of both document-
wide and chapter-specific Executive Summaries is a good one and should be adopted. Policy
wonks and the generally interested public typically read only the ExSums. If the essence of the
effort is not there, it will go unrecognized.
TABLE 1, Relation* ip at Ambient air P» to Dual Pb Loading awl Loading Rates"
Cut doc r AJr P&
:'***.'
3,13
3,2=
3,53
3,75
*,Q3
1.2E
1,53
Irriip-ter Air P6
ive*?}
,03
,03
.1S
23
,33
,33
1 if
Oi/xloaT PuEt =t<
Leasing "(aie
•u&tfty
16,3
40.0
30,3
133.C
16Q.C
2DQ.O
24Q.D
Irdoor Duit _ea-3
Loading las
f^W.-ty
4,3
12,3
24,3
36,3
48,3
E0,3
72,3
Cutotfflr 93-Day
PE Lsadhq
W'^T'I
1,*iD
3.&3D
7,200
13,330
14.i30
13.030
21.630
Indoor 90- 3: ay
PB Leading
fp&^j
432
1,KD
2,'5D
3,2^D
i.32D
5.4QD
5,430
Calauls:arE or EFA '=936 and EPA 1&94.
D-86
-------
TABLE 2. RilaltenBhlp or Arabian! Air Pb to Dial: Pti Load Ing. Loading Rates.
and He EPA Leal Hazard Rule for intertof Oust Pb"
outdoor Ar
agf-m':
D.1D
D.25
D.50
D.7E
1.QD
125
': .50
Irdoor AJr
f**1.!
O.E-3
01*3
0.15
0.23
0.30
0.3-3
0.45
irrfcKf DJEI Pb Loaarq =?aie
ig>',TiV:?
43
f "">
24.
26.
48.
EO.
f Ł...
fjgtip/o
E-,44
-.11
2.20
3,33
4,44
5.55
6.EO
Da^s To Ream
4C-tjQPb«P"
1CO
36
18
12
L
-f-
i
6
teors or EFA "916 3f>3 =PA 1594.
EPA doar ejsl PC Etardard: 40 CFR 745.
ADDITONAL POST-MEETING COMMENTS - EPA PB AQCD
Reviewer: Paul Mushak, Ph.D.
March 9, 2006
Additional Chapter 4 Comments
[CASAC Lead Review Panel Member] Fred [Miller]'s e-mail and my original comments need to
be juxtaposed in a common context. The context is simply the scientific interplay of models and
empirical data vis-a-vis the AQCD draft and its required connection to evaluation of the primary
and secondary NAAQS. We are not at all concerned in this chapter whether someone can get
either the ACSL or the C++ Basic Desktop versions of the O'Flaherty model to do a simulation
of a Pb-B. In fact,... I've used the desktop version of O'Flaherty to make runs over the years.
I'm not averse to use of O'Flaherty. The problem is broader than this.
D-87
-------
The modeling chapter cannot escape its linkage to risk assessment use for population assessment,
whether that assessment is for air regulations or some other purpose. The O'Flaherty model still
needs more evaluations with field data; the O'Flaherty model, to the best of my knowledge still
needs a discrete exposure model for handling batch runs; and the O'Flaherty model needs a
statistical module for descriptive and inferential statistics. Cutting to the chase, when I say a
model is ready to go, I don't mean someone can run it for a single output in Pb-B, but rather that
it can handle the risk assessment needs for lead-impacted groups, however large or small they
may be.
Secondly, we're also in a pickle, as I noted in my pre-meeting comments, about harmonizing
what this sub-group is saying about the modeling issues versus what the SAB [AdHoc All-Ages
Lead Model] AALM [Review] Panel (all of whose members were focused only on modeling),
have to say about modeling and the models. As [CASAC DFO] Fred [Butterfield] can attest,
there are a number of places in the comments and reportage in the AALM panel's report now
being finalized about not only the AALM model, but the O'Flaherty, Leggett (Pounds Leggett)
and IEUBK models. Seeking consensus among us four sub-group members would probably be
relatively less turbulent than the U.S. EPA having two parallel advisory panels having divergent
things to say about the same topic.
Secondly, there is the use of slope factors versus modeling. Again, the role of regression
analysis-derived slope factors is in the context of utility for subsequent usefulness for risk
assessment of lead-impacted groups. The panel was not convened to apply a slope factor to a
single community or a discrete study.
In the [February 28-March 1, 2006 CASAC Lead Review Panel] meetings, my recollection is
that [CASAC Lead Review Panel Member] Bruce Lanphear was not arguing in any adversarial
way for slope factors to be the choice rather than modeling. He mainly had questions about
modeling per se. They are not the same thing. [Panel Member] Joel Schwartz pursued at more
length the use of slope factors. Now, consider what is being said. What's being said is that one
can presumably take some central tendency expression of the slopes from 12 pooled studies,
some in inner urban and some in isolated mining/milling/smelting communities and come up
with a slope factor that can be applied anywhere else that EPA wishes to use this value.
Consider using this slope factor versus using a model, any model, that permits discrete inputs to
the exposure module from each and every of the multitude of specific sites where an air lead,
say, is measurable via either primary emissions or fugitive dusts. No one is proposing using any
of the models to come up with a national Pb-B based on single national snapshots of
environmental lead inputs. But one presumably would be proposing using the slope factor
equivalent if this approach were employed.
There is also the problematic nature of Pb-B measurement. A number of the studies contributing
data to the pooled study at issue for dust/soil Pb versus Pb-B were "single shot" Pb-B
measurements. Consider that a significant point being made in Chapter 6 and elsewhere in the
draft AQCD for Pb is that single-shot lead measurements are highly limited (unless, of course,
one wants to use the single measurements for deriving slope factors). The matter drips with
inconsistency.
D-88
-------
The 12 or so pooled studies of dust/soil vs. Pb-B, keep in mind, are centered on children. These
pooled studies drew on existing studies that were based on kids, not the entire range of ages. So,
no particular disadvantage attaches to models limited to kids when slope factors would be limited
to kids.
Alternatively, if one proposes to do an ad-hoc study using measured Pb in various media and
measured Pb-B for each and every site to come up with a specific slope factor where a model
might be used, one then has the problem of single-shot Pb-B measurements and more overall
resource costs.
Finally, I would note that the model vs. ad-hoc slope factor issue is, to quote Yogi Berra, "deja
vu all over again." Slope factors were the first approach, many years ago, but this strategy had
so many problems in terms of broader relevance, that mechanistic models were developed. My
collective sensibility after 38 years in the lead and other toxic metal field is taken aback by some
having us regress back to regressions.
[Note: Excerpted from Dr. Paul Mushak's e-mail message sent 03/09/2006 at 02:36 PM]
D-89
-------
Dr. Michael Newman
Dr. Michael Newman
Virginia Institute of Marine Science
February 28, 2006
In general, Chapter 8 appears to be well written and comprehensive. The authors should be
complimented for their efforts. In the spirit of enhancing their work, I offer the comments
below. Many of them may reflect personal opinion but are offered just in case they help.
The figures in the manuscript are clean but often what exactly is being plotted or communicated
is not straightforward. This makes some figures difficult to grasp without effort, e.g. what
exactly is the x-axis of Figure 8-1.5.1 or 8-1.5.2? Is it cumulative proportion of ranked values?
If so, the scatter in the plot is confusing.
Page 8-76
The rationale behind using the highest bounded NOEL that was lower than the lowest bounded
LOEL for survival, growth and reproduction for the TRY is difficult to understand as the
paragraph is currently written. Although it is important to be balanced about providing too much
detail for the many issues discussed in the manuscript, I would ask that the authors please
provide more justification. This is particularly key in my mind as the TRV and bioavailability
estimates are so important.
Page 8-81, lines 50-51.
The statement is made that growth and reproduction were considered because they are the most
ecologically relevant [compared to biochemical, physiological, pathological or survival]. I do not
understand exactly what is meant as the other factors certainly do influence the ecology of an
organism. The most obvious is survival: a dead organism certainly has a different ecology than a
living one. Behavior is relevant to fitness within a community so it is also ecologically relevant.
Page 8-91, lines 1-11 of text describe the selection of study results based on consistency with the
overall set of results. Because the observation of "significant" effect is usually made in each
paper with a statistical test, it is odd that this discussion does not include any consideration of
design/test power. If one test shows an effect in contrast to many that do not, that does not mean
it should be disregarded. It may be that that study was the most powerful one. I would ask that
some consideration of power differences due to design and testing issues be included in such
selection of studies for use.
Page 8-91. LC50 values are reported with estimates of uncertainty. It is customary to use 95%
fiducial (confidence) intervals but these values seem to be presented as one would a standard
deviation or standard error. Which is being expressed here?
Page 8-98. The transition is made from considering effects to individuals to those to ecosystems
without any separate discussion of population or community level effects. This is confusing to
me because EPA states in numerous documents that protection of population persistence/
viability within natural communities is their intent when assessing ecological risk. As the
D-90
-------
manuscript stands, the effects are discussed as either to individuals or to other all other levels of
ecological organization. I would prefer to see separate discussions of population, community,
and ecosystem effects because they are equally important. Combining them as done here makes
it too easy to not mention an important effect at one of these levels. (This blending of population,
community, and ecosystem effects into "ecosystem" effects after discussion of effects to
individuals is done again on page 8-127.)
Page 8-117/118
The AVS/SEM approach described here is sound within the limits of the discussion. For
organisms that can be exposed via particulate ingestion, it can be insufficient. This has been
documented in several publications, including those by Luoma.
Page 8-130
This is an accurate description but the EPA method has shortcomings that should be discussed
more relative to the purpose of this document. For example, it is biased toward metrics of effects
to individuals and the associated calculations do not consider crucial ecological interactions,
keystone species, or critical roles. Also, the selection of 5% is somewhat arbitrary.
Page 8-132
The FIAM is discussed as distinct - a "conceptual" model - from the BLM, and perhaps not
sufficient as a consequence. This distinction should not be made because the BLM is also a
conceptual model, i.e., it is not a physical model. The FIAM is a computational model when
applied with a speciation computer program as is the BLM. I find the distinction being made
unconvincing. It seems to push the BLM into the spotlight awkwardly. I suggest that the FIAM
and BLM be discussed as models emerging from the same general vantage and then just discuss
the details using the BLM. The discussion on page 8-167 seems balanced in this way.
Page 8-133, line 13-17
The movement of free ions through cell membranes seems to be presented immediately before
suffocation and disruption of ion regulation in a causal sequence here. There are too many steps
in between membrane transport and these consequences, such as oxidative damage, epithelial
layer lifting, inflammation, excess mucus production/sloughing, and chloride cell changes, for
this sentence to accurately represent the process. Please reword.
The discussion of the BLM having the advantage of using the "biotic ligand" instead of the gill
to allow one to model the site of action in direct contact with the water does not seem clear to
me. The gills are in direct contact with the water. The "biotic ligand" is a conceptual tool that
likely is - but is not necessarily - in contact with water.
Page 8-134
The AVS/SEM approach is likely a good one for a metal such as lead, but may not be as
appropriate for other metals with less "b metal" natures. The AVS/SEM depends heavily on an
interstitial water exposure route and does not fully include uptake from ingested solids. A
qualifying sentence might be good here.
D-91
-------
Page 8-146
The preponderance of
-------
Mr. Rich Poirot
Review Comments on 1st Draft Lead Criteria Document, R. Poirot, 3/7/06
I found Chapter 8 to be generally well organized and clearly written. The proposed approach to
use the summary sections on terrestrial effects (8.1.1) and aquatic effects (8.2.1) as the main
body of the chapter, with more detailed information intended to be appended as annexes, should
improve the readability. The brief summarization of "pre-1986 knowledge" followed by more
detailed focus of "new results" was effective. Some of the "conclusions" sections in the more
detailed "annexes" were clearer and crisper than what was provided in the summary sections,
and some of that language might be used in the new main body of the chapter in the next draft.
One general suggestion is to consider conducting the proposed "Integrative Synthesis" chapter
after, and to include some information from, the Environmental effects chapter. I think this
might be especially warranted for Pb, given the multimedia nature of lead exposures and the
large historical burden of Pb distributed to many compartments of the environment (i.e. current
human exposures are partly a function of what's been stored in or is being released from the
environment). Since there is a strong tendency (or at least a tradition) of avoiding real
consideration of secondary standards (typically and simply set equal to primary standards), an
important set of questions that might be included in such a synthesis is whether there are
environmental effects that occur at lead concentrations lower than, or for indicators, forms, or
averaging times different from, those which effect human health.
There were several potential applications such as use of the Biotic Ligand Model (BLM) and the
development of Critical Loads (CL) for lead deposition that appear to be promising ways to more
carefully consider environmental effects and pollutant limits in the future, but which might not
be quite ready for use at the current time. Staff can be commended for this eye toward the
future, even as some caution is encouraged to assure that various bugs are worked out of these
relatively new approaches. The critical loads approach - in relatively widespread use in Europe
and Canada - seems like an especially appropriate approach for considering environmental
impacts of pollutants like Pb, for which effects are more related to long-term cumulative
deposition than to current ambient air concentrations. However, the caution on page 8-239 (that
neglect of a factor for Pb transfer from the terrestrial catchment could result in a 5-fold
underestimate of Pb concentration in aquatic systems) indicates that a CL approach for aquatic
ecosystems needs to be further refined.
A conceptual question that might be raised in considering critical loads for lead or other
pollutants is whether a secondary standard based on such CL calculations should most logically
be expressed in units of deposition, or whether some approximate "conversion factor" could
relate assumed deposition rates to measured ambient air concentrations. Most of the Chapter 8
discussion on environmental Pb measurement methods (8.1.2.1 - 8.2.1.6 and 8.2.2.2 - 8.2.2.6)
focuses on analytical methods. This is useful information, but there could and should be better
coordination with, and to the extent possible better linkages to (Chapter 2) methods used to
measure Pb in the ambient air, precipitation, or dry deposition. At some point a direct linkage
needs to be drawn between the many various effects of lead in the environment and the relative
contributions to those effects from lead currently present in or deposited through the ambient air.
D-93
-------
It's not a topic specific to the environmental effects chapter, but I think more discussion of
particle size distributions in relation to environmental deposition and to ambient air sampling
methods is needed somewhere. This may be of increasing importance if relatively coarse re-
suspended roadside dust represents an increasing fraction of airborne Pb emissions. The FRM for
atmospheric lead (atomic absorption on high volume TSP samples, with fiberglass filters) is not
much in use any more, in part because ambient concentrations are typically so far below
standards, but also because the relatively "dirty" and fibrous glass filters (with high and variable
blank values for many elements) are not well suited for multi-pollutant analysis, and especially
by inexpensive surface beam techniques like XRF or PIXE. There is a general movement in "air
toxics" sampling programs to use PM-10 samplers. Meanwhile, fine particle Pb measurements
by XRF on Teflon filters are relatively abundant in both urban (STN) and rural (IMPROVE)
networks. If a revised secondary (or primary) standard were considered for Pb, how should it be
measured?
It was interesting to note the substantial discussion - in chapter 8 and elsewhere throughout the
document - of the potentially large emissions and environmental contamination of Pb (typically
in association with, and with effects often difficult to distinguish from, other injurious metals like
As, Cd, Cu, Zn) from various mining activities. A quick word search indicates about 170
instances of the words "mine", "mines" or "mining" in the Pb CD. Atmospheric emissions from
such mining activities would tend to be primarily coarse mode particles, characterized by
relatively high concentrations of multiple toxic metals. Yet such mining industry emissions are
specifically exempted from consideration in EPA's recently proposed PMio-2.5 standards. Some
additional explanation for this apparent substantial inconsistency seems warranted.
Specific comments on Chapter 8:
8-3, line 24: "RBA" needs to be defined.
8-12, line 30: Can we have some additional information on the "three different soils" that might
account for such a wide range (0 to 52%) of decrease in nitrogen reductase activity?
8-14 through 8-30: Generally it would be useful to better relate some of the analytical methods
commonly employed for analyzing Pb in different environmental compartments vs. Pb in the
ambient air. (i.e., better linkage to chapters 2 & 3). Also for methods like ICP-MS requiring
sample extraction, are there any issues with extraction efficiency.
8-15, line 11: should be "factor ... is" or "factors... are"
8-24, Figure 8-1.2.5: What is the shaded rectangle in this figure?
8-46, line 22: This "dry deposition may account for anywhere between 10 to 40% of total" is the
only estimate of dry deposition I found (in this chapter), is a rather wide range, and is quite dated
(1982). This also contrasts with other estimates provided in Chapter 2 (page 2-52). I would
expect that the substantial source controls have likely reduced (fine particle and) wet deposition
of Pb proportionately more than coarse particle Pb (since re-suspended roadside dust now
D-94
-------
accounts for a high fraction of atmospheric Pb emissions). So likely the dry/wet deposition ratio
has increased since 1982.
8-55, line 27: It might be informative to plot equations 8-1, 8-2 & 8-3 on a single graph - to
illustrate log:log relationships of different biota to soil Pb.
8-146, line 19: Its unfortunate that the only surface water quality data you present has such a
high frequency of non-detects. Are there no other useful data?
8-150, Figure 8-2.3.6 is not very clear.
8-157, line 2 and several preceding figures & pages: There seems to be a fairly consistent
pattern of general Pb increases from West to East, but with several notable Western "hotspots".
Can you provide any explanation for the cause of the hotspots?
8-239, line 1: should the CL units be mg/m2 (rather than mg/m3)?
Specific Comments on Chapter 2:
2-11, line 27: Over what spatial domain were these 19,000 tons emitted? Does not agree with
the 12,000 tons in Table 2-6?
2-15, lines 11 & 16: These specific reported fractions of total US Pb emissions from primary &
secondary smelters provide useful information, but do not seem to be reported in a similar way
for other subsequently discussed source categories.
2-21, line 8: Not clear what "6.1% of all Pb in the United States" really means?
2-26, line 29: I assume you mean that the emission rate per liter of fuel is 50-100 times higher
and not that total Pb emissions from waste oil are this much greater?
2-41, line 6: I think you mean "smaller than 2.5 microns" and not "smaller than
D-95
-------
Dr. Michael Rabinowitz
Comments on Chapter 3 of the First Draft By Michael Rabinowitz [02/23/2006]
Section 3.1 Air
This section on indoor air lead includes a lot of information about dust lead. Might the document
be more usefully organized if dust and air were treated in separate sections? Dust particle sizes
of 50 to 75 micron as an upper limit may not be respirable, but may well find there way into to
digestive tract for absorbtion there (page 3-1 line 10).
Also on page 3-1, on the subject of removal of old lead paint perhaps add the following example.
In the course of following a cohort of over 200 Boston children, it was found that about 68
percent of the households reported some sanding, scrapping, or other indoor surface refinishing
at their residence in the previous six months, and in 34 percent of the homes there was significant
levels of lead paint (more than 1.5 mg/sq cm) on at least one surface. There was an increase of
1.4 (SE 0.7) |ig/dl in the child's blood lead among those where there was refinishing activity,
compared to no activity. Furthermore, with increasing amounts of paint lead in the house, there
was a larger increment of blood lead: 1.1 (SE 1.4, when lead paint was < 1/2 % by dry weight) to
2.8 (SE 2.3) for paint lead levels of 1/2 to 3%, and to 4.8 |ig/dl (SE 2.2) when paint lead over 3%
was detected, a 70% rise in blood lead levels. This demonstrates that household refinishing
activity in the presence of lead paint can significantly raise the child's blood lead, at least
temporarily (Rabinowitz et al.,1985b). (Could also be placed on page 3-33 line 18)
On the subject of indoor dust lead, it might be worth mentioning the strong correlations among
indoor dust, indoor air and outdoor soil lead levels. For example, in Boston in a study of the
homes of 248 young children, the Spearman correlation between indoor air and floor dust was
0.22 (p<001), air and soil 0.18 ( p<01), and between dust and soil 0..43 (p<.0001) (Rabinowitz
et al. 1985a). Similar patterns of correlations have been reported in Cincinnati and elsewhere.
(Perhaps page 3-2, line 9 might be a good place to put this)
Section 3.2 Soil and Road Dust
On the subject of the response time of soil lead (page 3-13), two items might be added. Soil lead
levels have been found to decline not only from "removal," strictly speaking, but also from
mechanical mixing of surface soil with lower, less contaminated soils by the action of
earthworms, tilling, and landscaping. This results in a dilution of the soil lead. This intentional
landscaping by human activity is continually reshaped many urban and sub-urban environments.
(Page 3-13, middle of page)
Despite this, soils from lead smelter sites have demonstrated the remarkable persistence of
elevated soil lead, many decades after the suspension of industrial activity, particularly where the
land has been left unused. For example Rabinowitz (2005) surveyed soil lead in the vicinity of
old industrial sites to examine how the stable isotope patterns vary among the sites according to
the sources of the lead ore processed at each site. Lead smelters and refineries, which closed
down decades ago, are the basis of this investigation. Samples were taken from near five old
factory sites in Collinsville and Alton (Illinois), Ponderay (Idaho), East Chicago (Indiana) and
D-96
-------
Omaha (Nebraska). At every site visited, remnants of the old factories, in terms of soil lead
pollution, could be found. (Maybe place page 3-18, line 9 or 10)
Page 3-18 soil lead near smelter- perhaps add:
Many old lead smelters go unrecognized, particularly secondary, battery recycling smelters. The
current residents may not know that smelters operated in their area. For example, Eckel et al.
(2001) identified approximately 430 defunct smelter sites in the United States by examining city
registers and telephone directories from many decades ago. That compiled list was compared
with government registries of known hazardous sites. Soil samples were taken from 10 sites.
Only 5 of 319 sites, chosen from the top 8 states, were known to authorities. Nine of the 10 sites
sampled exceeded residential standards for soil lead level. So, the location of most old smelter
sites may be unknown to current inhabitants.
Section 3.4 Food
It may be of historic interest to note (perhaps on page 3-18) that food lead levels were
significantly higher during the 1960's and early 1970s. Values of well over 100 |ig/day in
composite diets were found. Values in canned goods were exceptionally high because the
packaging methods of that era consited of three part steel cans with soldered seems. These seams
were soldered with a brush application of molten solder which introduced microscopic particles
of "solder-splash" into the food. These methods have been superseded with welded and
aluminum cans as well as other food packaging technologies. A recent Canadian Food Inspection
Agency report bears on this matter:
(http://www.inspection.gc.ca/english/anima/fispoi/manman/canboi/ch7sec3/numl5e.shtml)
Do you want to mention the occasional instances of intentional unscrupulous adulteration of food
items with lead to increase the weight. Occasionaly batches of paprika are reported laced with
red lead (J Toxicol Clin Toxicol. 1996;34(5):507-11). Lead intoxication epidemic caused by
ingestion of contaminated ground paprika. Kakosy T, Hudak A, Naray M. From the National
Institute of Occupational Health, Budapest, Hungary, for example. And another example from
Germany (Dtsch Med Wochenschr. 1994 Dec 16; 119(50): 1756). [Lead poisoning caused by red
lead in paprika powder] [Article in German] Lohmoller G.
Section 3.5 Other Routes
Section 3.5.3
Glazes perhaps add: Lead glazes are a particular problem when low temperature fluxes and
glazes are used. This is often the case, for example, when traditional charcoal kilns are used,
rather than gas fired kilns. Tea cups and dishes from the People's Republic of China, sold in the
United States, supposedly only for decorative uses, frequently is in violation of the lead
standards.
Section 3.5.6 page 3-34
Indeed some folk remedies intentionally contain lead. The toxicity of the lead may be
responsible for the intended potency of the product. For example, such lead-containing products
include alarcon, alkohl, azarcon, bali goli, coral, gliasard, greta, kohl, liga, pay-loo-ah, rueda and
surma. These are of Mexican and Latin American, Middle Eastern, and Indian origins.
D-97
-------
(Morbidity and Mortality Weekly Report 1999;48:27-29.) The lead content of these products can
be more than fifty percent by weight. In other cases, lead is an unintentional contaminant, for
example ayurvedic metal-mineral tonics, a fertility drug called Deshi Dewa, and a Chinese
clamshell powder known as hai gen fen, which is a calcium supplement.
Section 3.6 Measurement Methods
It may be worth mention 2 points here. Firstly, the lead levels reported in this section range from
percents (parts per hundred) to parts per billion. Consequently, concerns about contamination of
the sample during sample collection and laboratory analysis vary greatly. Special care needs to
be exercised when the same laboratory process samples across a wide range of lead
concentrations.
Secondly, lead levels in food are often to low, that levels are below detection limits, and reported
values are inferred rather than measured. Care must be exercised when these low levels are
extrapolated back up to daily intake numbers.
S ecti on 3.7 Summary
Perhaps add here of in the section on measurement something about bioavailability: The lead
concentrations discussed here are usually total lead content, or lead measured by strong acid
extraction. Although this does yield values for lead concentrations, it may not be the most useful
approach. For the lead to be toxic it must be absorbed, and to be absorbed it must be
bioavailable, or soluble in physiological fluids. In some situations, such as waste smelter slag
this might be only a small fraction, while for tap water, it might be essentially all the lead. For
that reason, care must be exercised in co-paring lead levels among the various rotes of exposure.
Additional Citations
Eckel W, Rabinowitz M, Foster G (2001) Discovering unrecognized lead smelting sites by
historical methods. Amer J Public Health.91:625-7.
Rabinowitz M, Leviton A, Needleman H, Bellinger D, Waternaux C. (1985a) Environmental
correlates of infant blood lead levels. Environmental Research 38: 96-107.
Rabinowitz M, Leviton A and Bellinger D (1985b) Residential refinishing and elevated infant
blood lead levels. American Journal of Public Health 75:403-4.
Rabinowitz M (2005) Lead Isotopes in Soils near Five Historic American Lead Smelters and
Refineries. Science of the Total Environment; 346: 138-148.
D-98
-------
Additional Comments on the 1st Draft Lead AQCD by Michael Rabinowitz, Ph.D. [4/11/2006]
I had offered some suggestions at the time of our meeting, mostly about chemical forms and
sources of environmental level lead.
Also, I want to express my concern that the issue of confounding of lead's effects on child
development, as they appear in the section on the epidemiology of lead's impact on human
health are not fully developed or adequately explored.
Unlike most other environmental pollutants, lead's effects, as we encounter them, are strongly
confounded by other risk factors. The extent of this confounding varies among the populations
studied, depending on the patterns of lead exposure in the particular circumstances. For
example, is lead exposure correlated or not with social class or family income.
It is worth remembering that lead's effects are small, only a few IQ points, typically, compared
to other much stronger determinants, such as parental education or family income. This is
especially true at the lower blood lead levels currently being explored, now typically below 10
Hg/dL.
For me, this is especially potentially troublesome because lead is usually considered as a
continuous variable, well measured, while often the other variables may be expressed as
categorical variables, so small differences in these factors, when lumped into the same category,
cause differences in the outcome which could be ascribed to lead. For example, if the mother
graduated only 8th grade or 9th or 10th, they could all be scored as not graduating high school. I
do hope my concern about this is unfounded, and closer examination will show me worried
needlessly.
As I tried to explain, I think these circumstances require us to look at each of the studies to
examine the extent of the confounding. For example, in a given study, the model predicting IQ
as an outcome, can be made with and without a lead term; and then seeing if adding the lead term
(and the additional degree of freedom) statistically-significantly improves the model's goodness
of fit. This would tell us if lead is an independent risk factor. Also, examining the effect of
adding the lead term on the regression coefficients of the other terms would tell us in a
quantitative way the extent of the confounding in that population. For example, does the
coefficient for parental education change significantly when the lead term is added. In this way
we can see how free the lead effect is from confounding, and hence how reliably the lead effect
was measured in that study. This would enable us to put more weight on studies where lead's
effects were more cleanly measured.
These concerns, and the suggested approach to addressing them, can be developed more as the
human epidemiology synthesis chapter is reviewed.
D-99
-------
Dr. Joel Schwartz
Comments of Joel Schwartz
In general, I like approach of interpreting epidemiology in the light of toxicology. It is important
to cross-connect between these results to interpret the data in its entirety.
We need to be careful in the discussion of measurement error in outcomes, and the implication
of such error. Assume that the outcome is normally distributed, like cognitive function. Then in a
linear regression with a set of p covariates, and n observations, the predictors of the outcome are
an nxp matrix, which is conventionally denoted X. Let the true outcome be y (an nxl matrix).
Assume that y is mis-measured, and we instead observe Z, with
Z=y+u
And u is random measurement error. Then recall that in a linear regression the vector of
regression coefficients is estimated as
3 = (XXy1 XZ
1 x'y + (x'xy1 x'u
Assuming that the measurement error in outcome is uncorrelated with exposure the last term has
expectation zero. Hence, measurement error in the outcome induces NO BIAS in the estimate of
effect of lead, unless e.g. the error in assessing blood pressure is correlated with lead exposure.
Now consider the precision of the estimate. In a linear regression, we have
Var(J3) = (X'XylVar(Z)
= (X'XylVar(y) + (X'X)-lVar(u)
Since the last term in this case has a positive expectation, the standard error of J3 will be biased
upward. Hence the implication of measurement error in the outcome, except in the unlikely case
that it is correlated with lead, is to reduce the power to detect a significant effect, but not to
produce any bias in the estimate of effect.
P. 6.4 There is some leftover language from the PM document still in here. For example, the
discussion of whether the exposure marker reflects the geographic differences in exposure seems
PM-related. Presumably, we are asking whether it reflects the interpersonal differences in
exposure, for a specific averaging time.
Pp. 6.6-6.7 Emphasis should not rest on studies combining results in different cities, this seems
left over from the PM document. For lead, the issue is studies reporting combined results for
different cohorts, or, more generally, different study populations. In general replace the phrase
multi-city study everywhere it appears. Again, meta-analyses are of different study populations,
not "various cities." Similarly for the discussion of heterogeneity.
D-100
-------
P. 6.14 The conclusion that a single blood lead can be expected to be a relatively poor index of
body lead is too broad. It is derived from the simulation where there was a sudden, large, and
temporary change in blood lead levels due to a sudden, large, temporary change in lead intake. In
other circumstances, it may do much better. Also, remember, in a regression analysis it is the
variation in exposure across subjects that is correlated with the variation in outcome. So, as long
as the blood lead accurately reflects the relative ranking of subjects, all is not lost.
P. 6.14 Be clear that a better measure of the long term lead body burden is not necessarily what
is wanted for an epidemiologic study. If the effects of lead on a particular outcome are lasting
and cumulative, it is the preferred measure. If the effects of lead on the outcome represent the
acute effects of current exposure, than long term body burden is not the preferred exposure
metric. In the absence of clear evidence from toxicology as to which averaging time is most
relevant to a particular outcome, we must look at both.
P. 6.22 needs to make clear in the discussion of "methodological" vs. "epidemiologic" methods
for dealing with "detection limits" that:
a) The use of the phrase detection limit is somewhat misleading. We are really talking about
measurement error. The instrument detects values less than 3 SD of background or
measurement error. The values are merely quite imprecise. This is measurement error. On
average, values of e.g. 1SD above background are truly less than values of 2 SD above
background, although both are "less than the detection limit". Hence the real issue is
measurement error.
b) To assign values below the "detection limit" a value of, for example, half the "detection
limit" results in biased estimates of the association of bone lead with outcome. Assigning
the measured values, even when negative, does not lead to biased estimates of the slope
of the association of bone lead with outcome.
Pp. 6.28-29 The sentence "In contrast to using urine as a proxy for serum and measuring lead
isotopes ..." needs to be rewritten to clarify what Gulson did that was different, and better.
P. 6.40 Again, the discussion here seems to assume a priori that we want to measure body
burden, and all biomarkers of lead exposure, in this case urine lead, are discussed in the context
of how well they perform as surrogates for body burden. However, this is not the case. Lead can
have immediate effects. For example, it is lead concentrations in plasma, and not body burden
(except to the extent it provides endogenous sources of plasma lead) that is responsible for
inhibiting ferrochelatase. For this outcome, plasma lead is clearly superior to body burden.
Plasma lead is notoriously hard to measure, and urinary lead may serve as a less difficult marker
of plasma lead, and therefore is a potential surrogate of the immediate effects of current
lexicologically active lead. This needs to be highlighted in the discussion of urinary lead, along
with a discussion of the difficulties.
If ferrochelatase were the only endpoint that is dependent on plasma lead, or short-term
lexicologically available lead, that would be a minor point. However, for many outcomes, we
simply do not know what the appropriate averaging time for lead is. For example, lead results in
increased concentration of calcium in the endoplasmic reticulum, resulting in increased calcium
release in response to alpha adrenergic stimulation. To the extent that this is responsible for the
D-101
-------
effects of lead on blood pressure, that part of the response should be proportional to plasma lead
concentration. Recent studies suggest that some cognitive outcomes may depend more strongly
on current blood lead than previously hypothesized. The pooled results of Lanphear, strongly
suggest this, for example. Hence discussions of lead biomarkers should not be limited to their
ability to proxy for long term body burden.
P. 6.52 It is important to note that because of the way the sample was selected, cord blood lead
was not associated with lower SES in the Boston study. This is a key result, which has
substantial implication for interpreting the study. It is always better to control for potential
confounding by sample selection rather than statistical control.
P. 6.53 The CD follows the outdated approach of reporting that there was a negative association
with tooth lead, but not significant, and cautioning that reduced sample size suggests caution in
interpreting the lack of significance. I agree that the small sample size makes significance less
important. Indeed, given the large number of other tooth lead studies, it is almost irrelevant. The
key question is effect size. In this longitudinal study, how did the effect size for tooth lead
compare to the effect sizes from the cross-sectional studies that have used that metric of
exposure? The statement that there was a null finding should be dropped as misleading. There
was not a null finding. The investigators found a negative effect of tooth lead on cognitive
function. They also found wide confidence intervals for that effect. Show us both.
P 6.53. Please mention that chance is another possible explanation for the finding that the
strongest association between lead and full scale IQ was with exposure at age 2. Exposures to
lead showed moderate tracking in this cohort. Random chance can decide which measurement
shows the strongest association.
P. 6.55 Again, the most likely reason that the strongest association was with lead exposure at age
6 in the Cincinnati cohort, but age 2 in Boston, is chance. Given the correlations among the
measures, we have poor power to discriminate among the exposure times or intervals.
Pp. 6.56-7 similarly, in discussing the Cleveland study, the CD repeatedly says lead biomarkers
were or were not associated with some outcome. Presumably, they mean the association was or
was not significant. Given that conclusions are going to be drawn from all of the studies
collectively, it would be more useful to have the effect sizes and confidence intervals. At a
minimum, you should tell the audience the direction of the insignificant associations.
In addition, alcoholism among inner city residents is associated with use of other, potentially
fetotoxic drugs, such as cocaine and marijuana. Since these drugs, unlike alcohol, are illegal, it is
difficult to obtain good measures of use and control for them. This would be quite important in
this cohort. Please tell us if illegal drug use was controlled for, and how, in the analyses
presented.
Why does the document state that the authors reported that the covariates explained more of the
outcome than lead? Isn't this obvious? Of course they do. So what? Age explains more of lung
function than ozone, for that matter. But EPA neither regulates age nor maternal IQ. Why do we
care which other covariate (e.g., HOME score) had the greatest incremental R2, or "most strongly
D-102
-------
contributed" to child's IQ? And what does most strongly contributed to mean? Highest effect per
interquartile range change in exposure? Highest incremental R2?
P. 6.57 What does the statement "The estimated lead effect increased as a function of the level of
measurement error in dentine lead variable" mean? Did adjustment for measurement error in
exposure produce a steeper slope for the effects of dentine lead? This is necessarily true. Or does
it mean something else?
P. 6.60 It would be useful to convert the correlation coefficients to slopes and compare them.
P. 6.61 Clarify what is meant by saying the Gomaa study is related to the Schnaas study. Just by
being in the same city, or in some other way?
It is interesting to note that the effect sizes in Kosovo and the other high lead cohorts seem
smaller than the effect sizes in Boston and other low lead areas (although no effect sizes are
reported for Sydney). This is consistent with the steeper slopes at lower levels observed in
several of the studies.
While it is fair to point out that the NHANES analyses did not control for HOME score, it is also
fair to point out that unlike most other studies, it controlled for income, as indexed by the
poverty-income ratio.
P, 6.91 reports that the Inuit study population had a range of blood lead concentrations of 0.8—
27.1 |ig/dL. It then reports results excluding the 10% of the population with blood lead levels
above 101 jig /dL. I think a decimal slipped somewhere.
P. 6.107 Discussion of confounding. Here it would be useful to note that because of the middle to
upper class sample drawn at Boston Lying In Hospital, and the tendency for upper class
participants to live in older housing, the Bellinger study had a non-significant positive
association between SES and cord lead, and a negative association between SES and postnatal
blood lead that was considerably weaker than in most other studies. Hence the potential for
confounding by SES in this study is considerably less than average, yet it reports similar or even
larger effects than average.
P. 6.108 Mention here that the proper way to address the possibility that lead may be on the
causal pathway of the association between social class and IQ, but that there may be other
pathways that need to be controlled, is to use structural equation models, but this has not
generally been done (except for Christchurch).
P. 6.112 Why is the WRAT resistant to lead? The literature in children suggests that reading is
not independent of lead. This would seem to make the recommendation to control for reading
test results in examining other cognitive results in adults questionable.
Given the age of the NAS participants, their high bone lead levels may reflect general population
exposure in the 1960's and 1970's when environmental lead levels were higher and not some
D-103
-------
deviation from a "general population" sample. Very few of the participants had employment that
results in high occupational levels.
The NAS is well characterized for chronic medical conditions, through an every-three-year
physician interview and all-day medical examination. Education is measured and controlled for
in NAS studies when significant. It should be noted that for a medical condition to confound the
association of lead with cognition, it must not only be correlated with cognition, but with lead.
P. 6.128 The statement that there is no consistent evidence of effects in adults if competing risks
are taken into account is unsupportable. It was earlier hypothesized by the author that
uncontrolled covariates might confound the associations. Hypotheses are not facts. No evidence
is presented that there is confounding, and to have a statement conclude that confounding must
exist in the absence of evidence is at odds with the scientific method.
P. 6.164 The word race is here in parenthesis, whereas it is not in the discussion of IQ effects. Be
consistent. Also, given the proven racial differences in sensitivity to salt and to certain
cardiovascular mechanisms, the implication of the quotation marks that the term has no meaning
is clearly false.
P. 6.165 It should be mentioned that the British Regional Heart Study had very little power to
detect an association of lead with heart disease of the magnitude that would be expected given
the magnitude of the association between lead and blood pressure.
P. 6.165 While it is true that blood lead reflects both exogenous and endogenous lead, until
recently; environmental lead levels were high enough so that exogenous lead dominated blood
lead, except under certain circumstances, such as pregnancy, etc. Hence, for studies before the
mid 1990's or in countries where leaded gasoline was phased out later, the cautions on this page
are overstated. For newer studies, this is a fair point, but many recent studies have used bone lead
levels as exposures.
P. 6.168 states that all models included age, age-squared, BMI, race, family history of
hypertension, cholesterol, and tricep skinfold, and stratified by sex. As a comment it is stated that
the stepwise regression procedure used may have deleted potential confounders because of the
well known problems with stepwise regression. This is conceptually true, but in the absence of
even speculation as to a confounder (correlated with blood pressure and lead) seems rather
theoretical to warrant such caution.
P. 6.168 The insistence on reporting that 'no model diagnostics were reported' for most of these
studies is inconsistent with the rest of the CD, where most of the studies of most of the outcomes
do not report the model diagnostics discussed here, and are not singled out for such attention. Be
consistent. Also, some of the studies report results of multiple different models (e.g., Den Hond
et al., Pirkle et al.), and the sensitivity of the results. Such approaches are alternatives to model
diagnostics.
General Comment: While the section on cognition in children repeatedly makes mention of the
animal toxicology that supports the epidemiologic findings, this is totally absent in the discussion
D-104
-------
of lead and blood pressure/cardiovascular disease. But lead has been shown experimentally to
effect multiple cardiovascular endpoints, and these findings need to be cross mentioned here.
P. 6.280 Marcus and Schwartz published a nonlinear model relating lead to EP based on a
toxicokinetic model, which should be included in Table 6.9.1
P. 6.316 The Schaumberg study is described as being in middle-aged males, with ages from 60-
90.1 agree with this definition of middle age, but my younger colleagues may differ. Of course,
they are children.
P. 6.322 Exposure misclassification. This discussion needs to be more focused on epidemiologic
implications and not exposure implication. Remember that in a multiple regression it is the
variation in lead about its mean that is correlated with variations in outcomes about their means.
Hence, a downward bias in absolute levels, such as the stated underestimate in true brain lead
levels, is irrelevant. Mean differences disappear, because it is only variations about the mean that
contribute to the association. The real issue is how variable the exposure error is. If blood lead is
an estimate of brain lead with a great deal of variance, there will be substantial downward bias in
the estimated effect of lead. In general there needs to be more discussion of the errors in
variables theory from statistics, to clarify the effect of measurement error. Calculation of likely
attenuation factors for a plausible range of error variation, to show the possible extent of bias,
would be useful here. Some discussion of whether variations in instantaneous blood lead about
"true" lead exposure are Berksonian errors or not would also be relevant.
P. 6.325 repeats the error I noted earlier of questioning the validity of a reported association
based on the measurement error in the outcome. Since this measurement error can only bias
downward the t-statistic for the association, it is not a possible explanation of why a significant
effect was found, and therefore cannot call the association into question.
P. 6.348 Very nice application of basic risk assessment for the cardiovascular effects. I would
suggest one further step. Given the reduction in risk for a 5 |ig/dL drop in blood lead per 100,000
population, on can easily calculate, based on the latest NHANES data, the reduced number of
cardiovascular events per year if we lowered blood lead levels a further 1 |ig/dL. I believe this
will be quite a large number.
Joel Schwartz
Comments on Chapter 4
Chapter 4 represents the transition between the description of sources, i.e. emissions, and the
description of the association between biomarkers of dose (blood lead, bone lead, etc) and health
in Chapters 5 through 7. As such it represents the key issue of how we get from estimated
emissions to estimated blood lead, etc. There are important issues in this transition that are
completely missing from the chapter.
D-105
-------
First, how do we get from air emissions to concentrations in dust, soil, air, food, etc? Most air
lead is absorbed through these pathways, and not direct inhalation. This involves deposition
models, dispersion models, etc, which are not discussed. How much does a given amount of air
lead contribute to food lead, dust lead, soil lead? This needs to be remedied. This problem must
be solved, no matter how we proceed to estimate the effect of lead in these media on lead in
blood.
Second, the description of how we get from lead in media to blood lead is restricted to the use of
models. There is a large literature using empirical data on lead in various media and blood lead,
and regression models that estimate the blood lead/media lead slope for each media, controlling
for the other media. There is also an international pooling project of such studies, which needs to
be described and discussed. EPA simply cannot ignore the empirical studies and rely solely on
the models. The argument that such approaches are "limited to the situations where they were
estimated" is no more valid than the argument that the IUEBK model is limited to the situations
where it was validated. The regression models are estimated from a wide variety of locations,
probably wider than the situations used to validate the IUEBK model. In addition, unless there
are interactions between the effects of exposure in different pathways, or severe undetected
nonlinearities, the regression models should generalize just fine. The blood lead levels of interest
are near those where the models were estimated, and low enough that nonlinearities (which were
examined in the pooling project) and interactions are unlikely to be important.
Finally, the current models have been validated for children, whereas EPA will need to do risk
assessment for adults as well. There is apparently a crude model for adults used by Superfund.
This should be discussed, because the finding of associations between lead and blood pressure
makes health effects in adults an important question.
As I noted in oral commentary, the chapter needs to better describe the literature on water lead
exposure.
Additional Comments on Chapter 6
I disagree with the comment that vocabulary or reading scores need to be controlled for in the
studies of adult cognition. Controlling for them controls for a measure of baseline cognitive
performance that should improve the model R2, and hence power. This is presumably why it is
recommended. However, if these scores are uncorrelated with bone lead, their omission will not
result in confounding. If these scores are correlated with bone lead, that correlation may
represent the effect of lead. The CD argues that these are conserved resistant properties. But this
ignores the literature that tooth lead was associated with reading disability in high school from
the follow-up of Needleman et al. Reading disability in high school likely translates into reading
disability later in life. In general, examining the summary in this chapter about cognitive effects
on children one sees evidence of association of lead with a variety of cognitive outcomes that are
recommended as controls for adult cognition.. These outcomes have been associated with lead
exposure, and hence it is inappropriate to control for them as potential confounders.
D-106
-------
Flegal and Patterson's study of bone lead in pre-industrial native Americans would be useful to
cite in the discussion of thresholds and "low level" effects, since they indicate that even the 2
|ig/dL levels now seen are not truly low.
Threshold Argument: Following the discussion of all the factors (genetic, iron deficiency, etc.)
that modify susceptibility to lead, it would be worth pointing out that the implication is that if
there were thresholds in individuals, the distribution of those thresholds in the population would
be expected to be wide, and hence one would not expect to see a threshold in the population
average dose response curve.
D-107
-------
Dr. Frank Speizer
Lead CD, December 2005, Draft: Chapter 6 General Comment & Charge Questions
Discussant: Frank E. Speizer
General Comment:
The chapter is well organized and progresses well, in detail through the biologic
measures of exposure through the wide variety of health effects. Of concern is that because such
an abundance of biologic markers of exposure exist the authors seem to have simply "left out" in
this chapter any discussion of sources of exposure. For most readers it is obvious that lead is a
multimedia pollutant, and this is well discussed in Chapters 2-4; however, it seems to this reader
if one were to focus only on this chapter (and maybe the Toxicology Chapter 5) the sources of
exposure are missing. I would suggest that a very brief discussion, perhaps taken from or
expanded with words from the end of Chapter 2, Figure 2.6, be considered as part of the
introduction to this chapter.
Although I indicate that the chapter is well organized, there is a problem. Because it
appears that different writers seem to have written on the different disease outcomes there is
considerable overlap on the discussion of the biomarkers in each section. In addition, the writers
are clearly using many of the same studies in which multiple outcomes have been considered.
The studies are therefore being mentioned in each section as though they have not been
discussed before. In fact, this might be handled by one discussion and cross-referencing.
There are a number of places where reference is made to the Annex Tables particularly in
the biomarker discussion and in the neurobehavioral sections. In large part I agree with this
approach. However, there are a number of places where a summary graph, analysis, or table in
the text would be useful. This is particularly in contrast to the later sections on Renal and
Cardiovascular effects where this is done.
Charge Questions:
Question Fl: Differences in Biomarkers
Although the discussion of each marker seems thorough, although earlier work dating
back to the 70's is not discussed, which is ok; and each marker is summarized individually, what
is missing from the end of the section (at about page 6-47) is a more generally summary. The
Charge question asks whether the chapter adequately addresses the issue of which exposure
metric is appropriate. A summary table of each of the biomarkers, where it has been used and an
estimate of its adequacy might be appropriate. I think this could easily be done and might have
the added effect of pulling the various outcomes together. (Maybe this is a function for Chapter
7)?
Question F2a: Neurotoxic effects at low blood Pb levels and adequacy of discussion of influence
of model selection
D-108
-------
The effects seem to be well characterized. The consistency and heterogeneity discussed.
Having the Lanphear pool analysis significantly advances the understanding and goes a long well
to confirming the effects at low levels of exposure. Model selection seems adequate to me but I
would defer to others more sophisticated with interpretation of log-linear with moderately sparse
data.
Question F2b: Adequacy of non-neurotonic effects
The discussion of the health effects of lead seems thorough. However, it is not clear that
the discussion has focused adequately on environmental (and most particularly air sources) Most
of the sections read more like a chapter in a textbook of medicine, although the tables (and text
on occasions) really describe some important epidemiological studies (NHANES data). Again,
the issue seems to not being adequately tied to environmental exposures.
Question F2c: Adverse or clinically significant effects.
The argument that lead is a toxic pollutant was settled over 20 years ago. Much of the
discussion at that time was settled by reducing environmental exposure (lead out of gasoline) and
the resulting drop in blood lead in large population samples. In adult chelating studies of
individuals at "toxic" clinical levels proved that neurobehavioral effects were significant and
could be reversed. The more recent studies, to this reviewer, confirm the association between
relatively very low levels of biologic markers of exposure, particularly in young children, result
in toxic effects. Minor changes in population averages in IQ addressed in many of these studies
represent important adverse health effects and the reason for these effects need to be considered
(see below as answer to question F2d). In adults operating at significantly higher levels of
exposure changes in blood pressure and cardiovascular effects, again on a population basis,
represent significant adverse effects. Operating as an important mechanism for these later
effects, the effects on the renal system become an important pathway of effect. In addition the
primary clinical effects on renal function is harder to judge until it becomes clinically significant,
presumably because of relatively excess reserve in clinical function that is normally present.
Question F2d: Confounders and causal inference:
Although the discussion of confounders appears appropriate in those sections where it is
discussed, it is still somewhat incomplete. Part of the problem is that in the more clinical
sections (renal, cardiovascular, endocrine, etc.) there are simply inadequate data to discuss.
These are mostly clinical studies in which potential confounders are not adequately considered in
the primary publications (nor are the primary studies designed to take all these factors into
consideration). The discussion of the neurotoxic effects is far better and the important
confounders are considered. However, because of the nature of most of the studies the potential
for residual confounding remains. One must therefore turn to the consistency and coherence of
the data to made causal inferences. One other factor that also contributes to the causal inference
is that by the public health effects of reducing exposure in the 80's levels of lead biomarkers
have also been reduced, and this cannot be explained by unresolved confounding. This point
could be brought out more, again perhaps in Chapter 7.
Question F3: Discussion of susceptible populations:
Discussion seems adequate
D-109
-------
Dr. Ian von Lindern
Memorandum
To: Rogene Henderson, Chair, Air Quality Criteria Document, SAB Committee
From: Ian von Lindern, Panel Member
Subject: Comments Charge Questions Section 1-4 of the AQCD First External Draft
Overview
Charge Questions Al. To what extent is the document format (i.e., main chapters of the
1st draft AQCD focused on evaluative/interpretive aspects, with descriptive materials and tables
presented in annexes) useful and desirable? Can the structure be further improved? If so, how?
Overall, the EPA has done an impressive job in organizing, drafting and producing the first
external draft of this substantial document. Both the content and editorial quality is
commendable, especially considering the topic addressed; the breadth of material that must be
considered, reviewed, evaluated and integrated in a single document; and the time constraints.
Section 1 lays out a cogent discussion of the purpose of the document and the process and
procedures that are ahead for EPA in meeting its regulatory obligations. In that sense the
document has done a good job in presenting and evaluating new information from the scientific
literature that has accumulated since the last AQCD in 1986-90.
There seems to be some debate as to whether the document should be a compendium of all
available information, or be limited to that data that directly addresses modifying the NAAQS.
Both the 1977 and 1986 AQCD and addenda were seminal documents that, in combination with
other programs, supported and led to effective regulation of lead in the environment. These
efforts, in turn, substantially bettered the health and quality of life for millions of children,
adults, workers, and the biosphere, both in the U.S. and globally. During the years that these
environmental and public health improvements were being accomplished, these documents were
utilized, referenced, critiqued and practically applied on an almost daily basis. This occurred
both within and outside the inherent regulatory sphere. In that sense, these documents provided
an invaluable framework for society, industry and the scientific community to develop and
implement strategies to meet public health and environmental needs.
The current document will continue to serve this broader need and will be looked to and applied
by those addressing lead-related and associated issues for the next decade. In deference to those
uses, the EPA could, with relatively modest efforts, improve the document to serve those needs.
Currently there is variability in the level of detail, presentation, discussion and nature of the
several Chapters. Some Chapters provide significant evaluation and presentation of both historic
and new information that has evolved since the last document. Others refer to the previous
documents, particularly, where the level of data collection and sponsored research has decreased
significantly since the 1980's. As a result, in some Chapters, there are long presentations of new
D-110
-------
information that reflect the degree of new research and studies that have been published in the
peer-reviewed literature. Others, particularly Chapters 2 and 3 tend to rely on older information.
There are also differences among the Chapters as to how the information is presented. The
format of summarizing the understanding and knowledge through the last document in 1986,
then reviewing the new studies and data, followed by a discussion of the significance in terms of
regulatory needs, seems most appropriate. In several instances the analyses refer to or rely on
information presented in earlier Chapters. The document would benefit by better coordinating
the Chapters in line with synthesis chapter when this developed.
Chapter 2
Charge Questions Bl. Overall, does Chapter 2 provide adequate coverage of important chemical
properties of lead and concise summarization of pertinent information on sources of Pb and Pb
emissions, especially in relation to the United States? In particular, how well does Chapter 2
identify the most pertinent available datasets that contain information on emission rates for point
and area sources? Also, does the discussion of available data adequately address issues such as
the spatial distribution of point and area sources and emissions estimate uncertainties?
Furthermore, does the discussion satisfactorily address emissions by key industrial sectors? Does
Chapter 2 adequately address other important issues relating to the dispersal and/or accumulation
of Pb in the environment, e.g., resuspension of roadside dust or the potential for Pb to accumulate
in some media, like soils, due to its relatively low mobility? (The latter fact means that fairly low
air Pb concentrations have the potential to produce elevated soil concentrations over time due to
wet and dry deposition.) In addition, does the chapter adequately discuss key chemical and
transport related factors that should be considered in evaluating long-term buildup of Pb in the
environment? Finally, are the discussions of the leaching of Pb from soil and sediment into
surface and groundwater sufficiently complete for this chapter?
Relatively less new and detailed information appears in Chapter 2 in comparison to other
chapters, and in relation to the types of emission and source data and environmental chemistry
and transport presentations in the earlier AQCD documents. However, there have been
tremendous decreases in both the magnitude of the sources and curtailing of environmental
migration and transport that have effected significant reductions in exposures, absorption,
biological markers, etc. These, in turn, have led to human populations with levels of lead lower
than was conceivable thirty years ago. Clinical and research activities involving this evolving
population has led to greater understanding of the mechanisms and adverse health effects at
lower exposure levels. Sampling, monitoring and diagnostic techniques; exposure and biokinetic
models; and cleanup, remediation and preventative health responses have all evolved as a result
of these actions.
References are made in numerous places in later Chapters to the tremendous source, exposure
and blood lead reductions achieved and the significance of these accomplishments in analyzing
new material in almost every discipline. Presumably, the synthesis Chapter will make similar
references. This story, however, is not obvious in the current document, particularly in some of
sections of Chapter 2. It seems to indicate that emissions are of lesser concern, when the message
should emphasize the reductions that have been achieved, to set the tone for the remainder of the
document.
D-lll
-------
A major concern is that EPA has limited the review to "... where information is available in the
peer-reviewed literature. " Unfortunately, the best information for production data, emission
information, industry transition and economic indicators is found in the trade literature and
government agency records. Much of the practical knowledge that has, and continues to be
developed, with regard to applying scientific findings and methods to remedial and regulatory
activities is being generated in programmatic activities. These are the actions that have resulted
in the dramatic reductions in releases and exposures that have been achieved since the gasoline
phase-down. For the last decade, this information is accumulating within programs and
professional literature at a rate probably an order of magnitude greater than that reaching the
peer-reviewed j ournals.
It seems this information limitation would affect the ability of the Agency to assess impacts of
implementing any major revision to the NAAQS. Perhaps this will be handled in the Staff Paper.
But it leaves a void in the base information about sources that makes it difficult to put review of
the remainder of the document in the context of— How much lead is out there? Where is it?
What's it doing? Where is it going?
More of this contextual information is found throughout the other Chapters, but doesn't seem to
be appropriately introduced or supported by data in Chapter 2. Answering these contextual
questions seems vital to assessing the significance of atmospheric lead and the discussions and
analyses in the remainder of the document.
Section 2.1 Physical and Chemical Properties
This Section is well developed and presented in an interesting manner. It could benefit from
providing a few additional examples of the significance of these chemical and physical
mechanisms in environmental and biological processes that could be tied to material presented in
later Chapters. In particular, these physical and chemical properties could be tied to the transport
discussions later in this Chapter.
Section 2.2 Sources of Lead
Section 2.2.1 Natural Sources
Most of the concentration data indicated for "background" is globally averaged. A good deal of
information has been developed at various mining sites around the country and world regarding
background concentrations in mineralized areas. These data are available for host rock, soils
sediments. Inclusion of these data could provide perspective regarding the variability of
"background".
This is a general comment that goes beyond the natural sources. Much of the presentation in this
Section provides gross estimates but puts little in context of local, site-specific or industry
categories that might be affected.
The last sentence on line 15, page 2-12 and first sentence on page 2-13 should provide some
reason for the observed differences.
D-112
-------
The discussion for lead 210 on p 2-14 seems disproportionate to the relative significance of this
isotope in natural systems, as opposed to its significance as a tracer.
Section 2.2.2 Stationary Sources
This section would benefit greatly by providing perspective for the various discussions
undertaken. The magnitude, values, percentages, etc. should be put in context. This should
include an historic and national context, especially for the last 40 years that EPA has been active
in effecting the changes. With regard to future users of this document, it should also provide
global context that relates how other countries' emissions and releases compare to ours and a
local perspective acknowledging problem areas in the U.S and elsewhere. Summary Tables to
support the discussion would be useful.
A more developed historic summary of primary and secondary production for both the U.S. and
overseas would provide perspective regarding the shift in lead-related industrial activities, major
stationary source behavior, capacities, emission factors, etc. Summary Tables to support the
discussion would be useful.
More information regarding secondary production, in particular recyclers and battery recovery
operations, would be beneficial. This is one area where lead usage and production is increasing
and the prognosis for responsible life-cycle management is not indicated in this Chapter. Is this a
potential problem that should be flagged as a research need, or would a review of industry
practices show little need for regulation? Conversely, neighborhood battery reclamation is an
increasing danger around the world, as more people become car owners in many developing
countries. Summary Tables to support the discussion would be useful.
The Mining and Processing section seems limited to emissions from active mines. Among the
largest sources of lead release in the U.S. today are abandoned mining operations and waste
repositories. There is significant data accumulated in CERCLA and RCRA programs
inventorying the amount of lead waste being managed in the U.S. today. Presenting these data
would provide perspective regarding where regulatory resources should be allocated. Many of
the major mining related sites, in both programs, and at unregulated sites on private and public
lands represent significant sources of wind-blown or vehicle induced dusts and waterborne
erosion.
This introduces the question of contaminated soils that, together with the inventory of lead paint
in the U.S. housing stock, represents the largest sources of lead requiring management in the
country today. These topics likely deserve a sub-section in this Chapter as it relates to their role
as potential sources of lead.
Lastly, it would be more enlightening if the document were to develop and discuss a summary of
lead releases (as opposed to emissions) to the environment. This should supplement a more
detailed emission inventory that should also be developed. Comparing releases from the various
sources, as opposed to emission inventories from point and mobile sources - in a historical
summary- would provide considerable perspective regarding environmental lead.
D-113
-------
It is important that the source categories be presented in a format that demonstrates the relative
significance of the emissions and releases. This should be coordinated with the importance that
these sources play in the synthesis analyses and eventual regulatory needs. The various Tables
presented should be developed to support this theme.
Section 2.3 Transport Within the Environment
This Section was well developed with regard to the physics of various transport mechanisms that
can result in significant migration of lead in particulate or dissolved form. This section would
benefit if the discussions were prefaced with a few additions that tie the transport processes to
later Chapters that discuss mechanisms, sources and receptors in the traditional contaminant
migration format. For example, paniculate lead generally trans-locates either as dust in the
atmosphere or particles adhering to people, pets vehicles, objects etc.; or as suspended sediments
in water. These particles are transported to receptors where key exposure media are house-dust,
street dusts, soils etc. These particles generally end up in the soils where they reside for long
periods. Water- borne parti culate most often reach a sediment sink, where it is subject to both
physical and dissolved phase transport.
Again, there have been numerous studies conducted in CERCLA and RCRA programs that could
provide perspective on the significance of these factors in the U.S. Contaminant transport from
industrial and mining sites continues to present significant exposure to large populations in
developing countries. The migration of lead-based paint to house dusts and soils is an important
mechanism for which a substantial amount of data has accumulated.
Summary:
This summary to this Chapter could be made more cogent by editing and making additions that
preface how these data and the story it portends are used in the following Chapters. The Figures
should also be selected to illustrate the same important points. Figure 2.5 should be updated as it
shows that U.S. production almost returning to peak levels by 1990, where it ends, despite the
phase-down. What has happened since then, how much goes to batteries today, are there releases
in the life-cycle? Do we know, etc.?
Chapter 3
Charge Questions Cl. Does Chapter 3 provide adequate coverage of pertinent available
information (especially as it pertains to the United States) on Pb exposure routes, as well as
environmental Pb concentrations, including those in air, drinking water, food, soils, and dust?
Also, does the chapter delineate adequately interconnections between airborne Pb and its
potential contributions (via secondary deposition) to Pb in other media (e.g., indoor dust)?
Moreover, given the potential importance of historical deposition of Pb from mobile sources,
does the chapter adequately identify key sources of information characterizing the magnitude and
distribution of lead soil concentrations near roadways in urban, suburban and rural areas? Also,
given the importance of characterizing "background" Pb concentrations in conducting
health/ecological impact analyses (where background refers to both natural and generalized
anthropogenic contributions as distinct form specific point sources), does the chapter adequately
denote key sources of information characterizing existing "background" Pb levels in urban,
suburban and rural/pristine areas?
D-114
-------
Chapter 3 suffers some of the same shortcomings noted for Chapter 2. EPA has confined the
review to the peer-reviewed literature when the most advanced work accomplished in defining
exposures, pathways and estimated doses from environmental sources has occurred in regulatory
applications in CERCLA and HUD remedial activities. For the AQCD and follow up standard
setting process, EPA should either take advantage of the wealth of information available from
these regulatory activities, or point out the scarcity of peer-reviewed information suitable for
decision-making. These data sources and experiences could be considered in either the AQCD or
Staff Paper. However, it would not be prudent to limit the combined analyses to the relatively
obscure and less-representative studies that have reached the peer-reviewed journals. In the event
that EPA is restricted from utilizing this extensive experience in crafting effective regulations,
the scarcity and short-comings of information available from journal articles should be noted.
It also seems that that the key message Chapter 3 should leave with the Administrator and critical
reviewers is not clearly articulated. Chapter 3 should take the source, release and environmental
transport information from Chapter 2 and emphasize and quantify those pathways that lead to
critical exposures to humans and environmental receptors. In regard to human health; residual
soil and paint, and (as noted for Chapter 2) a poorly-defined industrial sector contribution are the
largest sources in this country today. The scientific consensus seems to be that dust is the key
environmental media offering excess exposures today. Airborne lead plays a key transport role in
effecting dust concentrations and loading through deposition. The intermediate air lead
relationships between the source and the dust variables are the key component that a new
standard will have to address. Understanding and quantifying the accumulation rates of
deposition is a critical unknown. How much source control will have to be exercised will be
determined by the relationship between dust loadings and blood lead levels (and, as noted below,
other contributors to lead intake).
Chapter 3 would benefit from additional structuring that illustrates the connections to the source
descriptions in Chapter 2, then follows the critical pathways through to exposure and intake in
Chapter 3, thus setting up multi-media input analysis through modeling that is presented in the
following Chapter 4. These analyses have been accomplished in internally peer reviewed, or at
least heavily critiqued, procedures at dozens of CERCLA and RCRA sites. The EPA should not
disregard this wealth of experience in addressing risks associated with airborne lead.
On the question of dust lead measurements, there does not yet seem to be a clear consensus on
the most effective methods to measure dust lead and relate it to blood lead levels. Lanphear, et al.
have conducted and compared dust collection methods in a side-by-side approaches and noted
that lead loading (ug/ft2), as opposed to lead concentration (mg/kg), showed higher correlation
with children's blood lead levels. Yiin et al. (2002) noted six studies of hard surfaces (i.e., floors
and window sills) where loading was found to be better correlated with blood lead and two
studies were cited where dust lead concentrations collected by vacuum sampling of carpets were
better. The latter was also noted at the Bunker Hill Superfund Site in Idaho (von Lindern, et al.
2003a, and von Lindern, et al. 2003b). There is apparently some confusion in citing the Bunker
Hill findings in these journal articles as both should be referenced, one focusing on dust:soil
relationships and the other on dust/soil:blood lead relationships.
D-115
-------
Chapter 4
Charge Questions Dl. How well does Chapter 4 concisely characterize key information on: (a)
the evolution and key features of important available approaches to the modeling of external Pb
exposures and their impacts on internal Pb body burdens; and (b) the status of model evaluation
efforts, e.g., PBPK model code verification and comparisons of model-predicted versus observed
impacts on blood or bone Pb distributions of particular lead exposure scenarios for affected
population groups? Also, does Chapter 4 sufficiently characterize the ability of different models
to handle key factors related to lead exposure modeling, including: temporal variation in external
exposure profiles; low level lead exposure; multi-pathway lead exposure; and the contribution of
historical/artifact lead exposure in influencing blood lead levels?
Furthermore, given that the October 2005 SAB review of the AALM suggested that further model
validation and verification was needed before the AALM should be used in support of regulatory
development, does Chapter 4 clearly identify which alternative models (e.g., IEUBK, O'Flaherty)
should be used for adult and/or child modeling instead of the fledgling AALM? In addition, does
Chapter 4 adequately identify the strengths and weaknesses of the recommended models in
modeling adult and child populations? Finally, overall, how can Chapter 4 be improved without
notable extension of length?
Two main points should be more clearly articulated in Chapter 4. Those points are inter-related
in an important context. The first point is that lead is a multimedia contaminant as illustrated in
Chapter 3. The second point is that later Chapters indicate that there is no apparent threshold for
deleterious effects and adverse irreversible outcomes are likely at low levels. As a result, to
evaluate, develop and implement reasonable interventions and regulations, EPA will need to
effectively analyze exposures and blood lead responses at extremely low levels in multiple
media. At present, this can only be accomplished through modeling. This Chapter should
rigorously critique the available models, as it does. It should also point the regulators to the most
effective and useful models with the best track record. That is not clear in the current version.
In Chapter 4 the difficulties of relying on the experience reflected in peer-reviewed journal
articles needs repeated. The IEUBK model has been employed at dozens of CERCLA/RCRA
sites and EPA has maintained internal review groups to monitor and critique these applications.
The AALM is not currently, and likely will not be in the context of the AQCD timeframe,
developed to a point sufficient to use in regulatory activities. The other models do not have the
depth of application to real-world situations, nor have they been relied upon to make significant
health decisions and resource commitments. The EPA should make an effort, in either the AQCD
or Staff Paper, to convey this experience and lessons learned to reviewers and decision-makers in
this process.
The conflict between peer-reviewed literature and practical experience in utilizing scientific
information to effect environmental policy has been an enduring difficulty. Much of the front-
line and more innovative work with regard to reclamation, remediation exposure reduction and
source management is being accomplished in programmatic activities. Concerns have been
voiced that this information is not appropriately disseminated, is not subject to peer-review, and
may or may not apply the best scientific methods. At the request of Congress, the National
Research Council of the National Academy of Science recently completed an extensive review
of the adequacy of scientific methods employed at the Bunker Hill Superfund Site in northern
D-116
-------
Idaho. The NAS was specifically charged with addressing the adequacy of the IEUBK model
and EPA's use of the model. That report entitled Superfund and Mining Megasites-Lessom from
the Coeur d'Alene River Basin, (NRC 2005) evaluates this problem with respect to a major lead
site that has played a historically significant role in the development of the NAAQS over the last
thirty years. (The study Yankel, et al. 1977 was one of the those selected to define the blood
lead:air lead ratio for the current NAAQS. The NAAQS was subsequently challenged in federal
court on the basis of the scientific adequacy of Yankel, et al. and was upheld). Some of the
lessons learned and references utilized in the NAS review could be applicable to the preparation
of this document. Von Lindern, et al. 2003a summarizes the blood lead: soil/dust relationships,
applies the IEUBK model to more than twenty years of remedial and health response activities,
and compares mechanistic and quantitative models in explaining those relationships.
Lanphear, B. P., M. Emond, et al. (1995). "A side-by-side comparison of dust collection methods
for sampling lead- contaminated house dust." Environ Res 68(2): 114-23.
Lanphear, B. P., T. D. Matte, et al. (1998). "The contribution of lead-contaminated house dust
and residential soil to children's blood lead levels. A pooled analysis of 12 epidemiologic
studies." Environ Res 79(1): 51-68.
National Research Council of the National Academies (NAS) (2005). Superfund and Mining
Megasites-Lessons from the Coeur d'Alene River Basin.
von Lindern, I, S. Spalinger, et al. (2003a). "Assessing remedial effectiveness through the blood
lead:soil/dust lead relationship at the Bunker Hill Superfund Site in the Silver Valley of
Idaho." The Science of The Total Environment 303(1-2): 139-170.
von Lindern, I. H., S. M. Spalinger, et al. (2003b). "The influence of soil remediation on lead in
house dust." The Science of The Total Environment 303(1-2): 59-78.
Yiin, L. M., G. G. Rhoads, et al. (2002). "Comparison of techniques to reduce residential lead
dust on carpet and upholstery: the New Jersey assessment of cleaning techniques trial."
Environ Health Perspect 110(12): 1233-1237.
Yankel, A. J., I. H. von Lindern, et al. (1977). "The Silver Valley lead study: the relationship
between childhood blood lead levels and environmental exposure." J Air Pollut Control
Assoc 27(8): 763-7.
D-117
-------
Dr. Barbara Zielinska
Comments on the EPA NCEA-RTP Air Quality Criteria Document for Lead, First
External Review Draft
Chapter 2: Lead Chemistry, Sources and Transport
Barbara Zielinska
In general, Chapter 2 is very well written and summarizes adequately pertinent
information regarding chemistry, natural and anthropogenic sources and transport of lead in the
environment. However, the information regarding lead emissions from important industrial
sources seems to be somewhat outdated. For example, Pb emission data from coal combustion,
fuel oil combustion and some metallurgical processes relay mostly on older references (such as
Pacyna, 1986). These data are limited and may no longer be applicable. If adequate peer-review
literature data do not exist, the use of publicly available reports and reliable compilation of data
is justified.
The chapter identifies available sources of information on emission sources and emission
inventory and points out their uncertainties. However, although the limitations of AP-42
guidelines and other EPA emission inventories are mentioned in general, the reader is not
informed as to why these data are not applicable. Usable emission and source characterization
data are critical and the need to have them updated has to be emphasized. In addition, although
data concerning lead particle sizes are scattered through different sections, a summary section or
table would be useful.
The transport of lead within the environment is covered very thoroughly. I find the
discussion regarding the accumulation of lead in soil and sediments adequate and informative in
term of the long-term build-up of Pb concentrations in the environment.
Specific comments:
1. Page 2-1, lines 23-25: this sentence needs revision - perhaps ".. .that forms a
protective..."
2. Page 2-1, line 29: amount of oxygen
3. Page 2-9 and 2-10: the terms used in equations 2-2 and 2-3 need to be explained
4. Page 2-15, lines 29-30: the range of emission rates from the blast furnace seems to be
very high. Also, the reference is rather outdated (1989).
5. Page 2-14, line 28 and page 2-68, line 6: I find the terms like non-human animals and
human animals rather strange...
6. Page 2-25-2-26: the information concerning Pb emissions from fuel oil combustion relay
mostly on older reference (Pacyna, 1986).
7. Page 2-48, line 2: Equation 2-1? I suppose 2-5 is meant.
8. Page 2-52, line 6: is the dry deposition more important than wet deposition in marine
areas?
9. Page 2-56, Figure 2-3: the figure capture mentions three resuspension rates, but the figure
legend shows four
10. Page 2-72, Figure 2-5 is rather poor quality and difficult to read.
11. Page 2-74, Figure 2-7. I suppose wet and dry deposition is also important for water
surfaces.
D-118
-------
NOTICE
This report has been written as part of the activities of the U.S. Environmental
Protection Agency's (EPA) Clean Air Scientific Advisory Committee (CASAC), a
Federal advisory committee administratively located under the EPA Science Advisory
Board (SAB) Staff Office that is chartered to provide extramural scientific information
and advice to the Administrator and other officials of the EPA. The CAS AC is
structured to provide balanced, expert assessment of scientific matters related to issue
and problems facing the Agency. This report has not been reviewed for approval by the
Agency and, hence, the contents of this report do not necessarily represent the views and
policies of the EPA, nor of other agencies in the Executive Branch of the Federal
government, nor does mention of trade names or commercial products constitute a
recommendation for use. CASAC reports are posted on the SAB Web site at:
http://www.epa.gov/sab.
D-119
-------
|