UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                          WASHINGTON, B.C. 20460
                                                             OFFICE OF THE ADMINISTRATOR
                                                              SCIENCE ADVISORY BOARD
                                    May 21,2009

EPA-CASAC-09-008

The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Ave., NW
Washington, D.C. 20460

       Subj ect: Review of EPA's Integrated Science Assessment for Paniculate Matter (First
               External Review Draft, December 2008)

Dear Administrator Jackson:

       The Clean Air Scientific Advisory Committee (CAS AC) and members of the CAS AC
Particulate Matter (PM) Review Panel met on April 1-2, 2009 to review the Integrated Science
Assessment for Paniculate Matter (First External Review Draft, December 2008, the PM ISA)
and two planning documents related to this NAAQS review.  The CASAC PM Review Panel
also held a teleconference on May 7, 2009 to finalize this report. In this letter, we summarize the
CASAC's major recommendations on the ISA. Comments of individual panel members are
enclosed.

       The CASAC commends the EPA Staff for the development of a generally excellent and
comprehensive first-draft PM ISA.  Most chapters were viewed as excellent, particularly for a
first-draft document. The document draws on a massive base of evidence that is usefully
summarized in tables and the appendices. The evidence is thoughtfully synthesized in a
transparent fashion; the framework for classifying the strength of evidence has continued to
evolve, and it provides transparency in documenting how determinations were made with regard
to causation. The CASAC is particularly pleased that the Agency has adopted a uniform
descriptive language for various levels of confidence in making causality determinations. We
support the five-level hierarchy developed for causal determinations, and recommend it as the
model for future ISAs. Chapter 2 ("Integrative Health Effects Overview"), which summarizes
and integrates across the chapters, was thought to be an invaluable component of the ISA.
Below are CASAC's major recommendations for improvement, followed by responses to the
charge questions.

   •   The ISA should provide more context for the reader on the regulatory options considered
       in the last review, and on the divergence between EPA's most recent PM decisions and

-------
   CASAC's advice. This background would help CASAC and the public in identifying the
   key new information summarized in the ISA.

•  In the summary chapter (Chapter 2, "Integrative Health Effects Overview"), building on
   sufficient discussion in subsequent chapters, the ISA should more fully integrate the body
   of evidence related to PM size and composition for both health and welfare outcomes.
   Since Chapter 2 is intended as a summary, we recommend integrating conclusions about
   welfare effects into this chapter, and, in fact, inserting the term "welfare" in the title of
   the chapter ("Integrative Health and Welfare Effects Overview"). In addition, we
   recommend bringing all important findings into this chapter, such as the Agency's
   determinations on the effects of PM2 5 on cancer risk and on the health risks of various
   PM components.

•  Welfare effects addressed in the ISA were too narrowly restricted to visibility. There was
   little effort to apply the criteria of causality laid out for health endpoints, even for
   visibility for which the evidence is strongest among the various welfare outcomes. The
   implications of PM size and composition for climate effects need to be more thoroughly
   explored, both with regard to climate change and the associated health and welfare effects
   of climate change. The document's restriction of the range of welfare effects may
   perhaps reflect a priori conclusions on the "policy-relevance" of various outcomes.  For
   example, rather than focusing exclusively on the most tolerant plants, a search for
   sensitive receptors for specific endpoints might help to better focus the discussion of
   welfare effects and the search for a quantifiable concentration-response (C-R)
   relationship for guiding development of a separately determined secondary standard for
   PM.  It would improve the ISA if the same causality framework were applied to PM and
   welfare effects, as for the health outcomes, emphasizing discrete endpoints and possible
   causal relationships.  Visibility, it should be noted, is not fully separable from health
   outcomes since reduction of visibility may have consequences for quality of life and
   well-being, which are arguably health outcomes.  Additional discussion and emphasis on
   "sense of well being" studies is needed as they relate to urban visibility. The implications
   of these  studies should be highlighted in the summary section for Chapter 9 ("Ecosystem
   and Welfare Effects") as well as in the full ISA summary in Chapter 2.

   As an additional point, visibility effects in eastern and urban areas (non-Class 1 areas)
   deserve far greater attention,  particularly in view of the consideration of an urban
   visibility standard. EPA should also discuss those data needed for developing an
   alternative optical standard for visibility.

•  With respect to concentration-response functions, the ISA needs to set out as clearly as
   possible, the concentrations and their associated averaging times at which health effects
   have been observed in relation to short-term and long-term exposures.  This information
   is needed in anticipation of use of these functions in the REA.  This dimension of data
   display is currently lacking, but needs to be reflected in tables and figures.

•  Other chronic health endpoints deserve more attention in the ISA.  In particular, the
   coverage of lung  cancer is limited as is that of chronic obstructive pulmonary disease.

-------
   •   The document should more fully address implications of particle size and composition for
       risk to health.  At the least, the major gaps in the evidence should be identified.

   •   Revisions are needed to remove the impression that PMio is a separate pollutant from
       PM2.5 and PMio-2.s.  The ISA handles PMio, PM2.5, and PMio-2.5, as if they were separate
       entities, even though the latter two are components of PMio.  The CAS AC cautions
       against this approach and notes that PMio is a mixture that contains varying proportions
       of particles in the smaller and larger size ranges. The current approach leads to
       inconsistencies in classification of evidence, with for example the evidence for
       respiratory morbidity classified as sufficient for PM2.5, but likely for PMio, which includes
       PM2 5.  When possible, the particle size distribution of the PMio mixture should be
       provided. To the extent possible, the document should take a more integrative approach
       in drawing inferences across the range of PM sizes. Similarly, a discussion of the
       relevance or importance  of particle composition for the different particle size fractions is
       warranted.  The ISA can be useful in identifying gaps related to particle size and
       composition that should be addressed with new research.

Our specific responses to charge questions follow below:

   1.  The framework for causal determination and judging the overall weight of evidence
       is presented in Chapter 1. Is this framework appropriately applied for this PM
       ISA? How might the application of the  framework be improved for PM effects?

       The CAS AC regards the framework for causal determination and judging the weight of
evidence, as presented in Chapter 1, to be appropriate.  The two-step approach, using first a
categorization of the evidence in support of causation and second an evaluation of exposure-
response is a reasonable, logical process. By characterizing the level of certainty with regard to
the presence of causal effects, the ISA sets out a range of adverse health and welfare effects,
along with an implied statement as to the likelihood of benefits following control. The
categorization gives guidance to the Administrator as to the potential for  public health benefits
from the PM (or other) National Ambient Air Quality Standards (NAAQS).  The categorization
reflects the strength of evidence and not the potential magnitude of public health benefits; those
effects for which the evidence is less certain should also receive consideration in the Risk and
Exposure Analysis (REA) and in policy deliberations.

       Because there is concern for susceptible and vulnerable populations, the concepts of
confounding and effect modification need to be sufficiently developed so that readers of the ISA
understand the relevance of the concept of effect modification when considering susceptibility.
We suggest this topic be addressed in this chapter and then more comprehensively in Chapter 8
with regard to susceptibility.

       Publication bias is likely to be relevant in the  assessment of causality, and perhaps even
more  so in the assessment of the evidence characterizing concentration-response relationships.
Discussion is needed on the implications of publication bias; how it is detected and how its
consequences should be taken into account.

-------
   2.  Chapter 2 presents the integrative summary and conclusions from the health effects
       evidence at the beginning of the ISA with the evidence characterized in detail in
       subsequent chapters. (Environmental and public welfare effects evidence is
       evaluated and summarized in Chapter 9.) Is this a useful and effective summary
       presentation? How does the Panel view the inclusion in Chapter 2 of only health
       categories with causal determinations of (a) likely to be a causal relationship or (b) a
       causal relationship?

       The ISA is comprehensive, lengthy, and complex.  Chapters 3, 4, 5, 6, 7, and 8 are
detailed and well documented, but are so extensive as to reduce the functionality of the document
for readers. Consequently, the public and policy makers need the high-level summary provided
in the 26 pages of Chapter 2, which provides an excellent overview of the major acute and
chronic health impacts of PM.  CAS AC commends the inclusion of Chapter 2 and recommends
similar chapters for future ISAs.  Inevitably, there are conclusions that require further
documentation,  but one can easily find the corresponding sections of subsequent chapters.  In
answer to the charge question posed, we are affirmative and find it to be a "useful  and  effective
summary presentation."

       As discussed above, the five-level hierarchy developed for causal determination is
appropriate. This is a reasonable approach to a central issue in the development of the ISAs (and
previously the Criteria Documents). The EPA staff has critically reviewed the relevant literature
on this topic and has made specific recommendations regarding their conclusions with regard to
causality.  We agree with the approach of starting with a list of the key health effects and a firm
statement with regard to conclusions of causality and following this by a brief description of the
key findings supporting the conclusions. The discussion in Chapter 2 should highlight the
evidence leading to the assignments of classification regarding the strength of evidence within
the five-level hierarchy.  This approach should become a model for future ISA documents.

       In general, the CASAC concurs with focusing the summary on evidence with causal
determinations of "(a) likely to be a causal relationship or (b) a causal relationship." But we
make two recommendations. First, the summary should point the reader to appropriate sections
of Chapter 1 where the definitions of the five-level hierarchy of causal determination are given.
Second, CASAC further recommends that the focus on outcomes for which the strength of
evidence for causation is at the higher levels should not lead to exclusion of those  outcomes for
which the evidence is weaker.  Third, emerging issues relevant to the PM NAAQS should also
receive consideration.

       We also suggest that the findings of Chapter 9 for welfare effects be included in this
summary and that the chapter title include welfare effects. Without including this material, the
summary is imbalanced and does not give sufficient weight to the ecosystem and welfare effects.

       Additionally,  CASAC recommends that the ISA  should acknowledge and address the
broader relationships  between the ecosystem and welfare effects considered in Chapter 9 and
human health. Climate change has diverse implications for human health and reduced visibility
                                           4

-------
is a potential environmental stressor. We also suggest addition of a section entitled "Emerging
Issues." This section could address health effects for which the evidence has not reached the top
two categories of strength of evidence and point to research needs. It could address topics of
likely concern in the future, such as epigenetic changes or influences of PM on premature birth
and low birth weight babies. It could also highlight other topics that have been identified
through the extensive review process underlying the development of the ISA.

   3.  To what extent are the atmospheric chemistry and air quality characterizations
       clearly communicated, appropriately characterized, and relevant to the review of
       the PM NAAQS?  Does the information on atmospheric sciences and exposure
       provide useful context and insights for the evaluation of human health effects of PM
       in the ISA?

          a. Is accurate and appropriate information provided regarding PM source
             characteristics, techniques for measuring PM and its components, policy-
             relevant background PM, and spatial and temporal patterns  of PM
             concentration? Are the analyses and figures presented in Chapter 3 effective
             in depicting ambient PM characteristics?

       Chapter 3 generally describes the relevant atmospheric chemistry, air quality
characterization and exposure  assessments that are relevant to the NAAQS review. The chapter
was well done and the length is justified, given the new data available. There are changes to be
made that will provide improvements in the presentation and completeness of the material that
has been introduced.  For example, although the chapter indicates it will provide information on
particle size, composition, and shape (line 23,  page 3-3), there is no subsequent  discussion of
particle morphology or surface area with respect to the characteristics of the ambient aerosol.

       The section on measurement methods needs to be improved.  There needs to be a more
complete discussion of PM mass measurements and the serious limitations of the current Federal
Reference Method (FRM) for PM. The current FRM does not provide complete and adequately
time-resolved concentration data,  nor does it provide an accurate indicator of mass concentration,
given known losses of semivolatile constituents and retention of particle-bound  water. There
needs to be discussion of the quantification of PMio-2.5, and a justification of the use of PMio as
an indicator of coarse thoracic particle exposure.

       There is a lack of information on the presence of chemically reactive species, e.g.,
reactive oxygen species, associated with particles (particularly SOA and their formation by
atmospheric chemistry) and of the chemical composition of coarse thoracic particles.

       The emission inventory data needs to be better integrated with source apportionment
results and the quantification of contributions from primary and secondary sources needs to be
better described.

       Changes from previous reviews in the rationale and methodology for estimating the
policy relevant background should also be addressed.

-------
          b. Is the evidence relating human exposure to ambient PM and errors
             associated with PM exposure assessment presented clearly, succinctly, and
             accurately? Are there PM exposure issues that should be expanded,
             shortened, added or removed?

       The current exposure section should be reorganized to flow more logically and revised to
discuss additional PM issues. Correspondingly, the section on PM exposure and socioeconomic
status (SES) should be better integrated with the other exposure related issues. Presently, the
section focuses primarily on 24 hr PM2.5 exposures and corresponding ambient concentrations. It
should be modified to include discussions of shorter-term exposure (e.g., hourly) as well as
exposures of longer duration (e.g., annual), and of PMio, PMio-2.5., and the components, in
addition to PM2.s. Further, the section should discuss how the relation between ambient PM and
personal exposure varies under different scenarios (e.g., populations measured at single points in
time and intra- or inter-community comparisons).

          c. To what extent does the Panel find Annex A appropriate, adequate and
             effective in supporting the ISA?

       The balance of data between the Chapter and Annex is good, although the Annex needs a
table of contents to make it more navigable. CAS AC generally appreciated the completeness of
the material presented in Appendix A. Using the  "search and find" approach, it does prove
useful in supporting the material in Chapter 3.  Annex A would be more useful if text references
to the Annex were more specific (e.g. by referring to Table a-2, etc.) rather than just a generic
'see Annex A'.

   4.  The dosimetry of PM is discussed in  Chapter 4. The primary focus is on factors that
       might lead to differences in deposition and clearance between individuals, species,
       and as a function of the physicochemical properties of particles. Is the review of
       basic dosimetric principles presented in sufficient detail? Are the new particle
       translocation data adequately and accurately described?  Recognizing an overall
       goal of producing a clear and concise chapter, are there topics that should added or
       receive additional discussion? Similarly, are there topics that should be shortened
       or removed? To what extent does the Panel find Annex B appropriate, adequate
       and effective in supporting the ISA?

       In general, the review of basic dosimetric  principles is well written and presented in
sufficient detail.  It is a needed component of the  ISA. Information about particle translocation is
largely up to date.

       There is one evident gap to be addressed.  A section is needed under "4.2.4 Biologic
Factors Modulating Deposition" that addresses exercise, increased ventilation, and route  of
breathing (nose vs. mouth). For both children and adults, exercise has a strong effect on
deposition and retention and hence on dose. First, greater amounts of PM are inhaled with
exercise, the amount inhaled being proportional to minute ventilation. Second, deposition
mechanisms are a function of linear velocity of the inhaled air, of residence time, and of other
factors related to ventilation.  Third, as ventilation increases, most people switch from nose

-------
breathing to mouth breathing.  The elimination of the nasal filter has a substantial impact on
deposition, particularly the deposition of larger particles in the airways.  For example, coarse
thoracic particles, filtered by the nose under sedentary conditions, have a greater probability of
reaching the large airways and being deposited there under oral breathing.

       The chapter also would be enriched by additional discussion of some specific PM
components. For example, what are common teachable components, and how do they differ in
their clearance mechanisms? What is the fate of metals coming from soluble particles? What
organs do they reach and what are the implications of retention of PM components in the brain,
heart, kidneys, liver, or other critical  organs?

   Annex B is appropriate, adequate, and effective in supporting the dosimetry chapter.

   5.  Chapter 5 is intended to support the evaluation  of health effects evidence for both
       short-term and long-term exposures to PM.  Some potential modes of action may
       underlie a number of health outcomes and may  contribute to health effects of both
       short- and long-term exposures. Thus, the potential modes of action are described
       briefly in Chapter 5, and some specific study findings are discussed in more detail in
       the relevant sections of Chapters 6 or 7. What are views of the Panel on this
       approach and on the characterization of potential modes of action for PM-related
       effects in Chapter  5?

       Chapter 5 ("Possible Pathways/Modes of Action")  is essential to establishing a
foundation for interpreting the evidence on the health risks of PM.  It outlines basic mechanisms
of action, and thus contributes to understanding the plausibility of the findings presented in
subsequent chapters. The common fundamental toxic mechanisms,  limited in number, are well
described, particularly the role of reactive oxygen species (ROS). The chapter is short, focused,
and incorporates informative illustrations that describe the interplay of pathogenic mechanisms.

       Each section provides a clear summary of biological effects. However, it is important to
establish whether these effects are similar across species and under what conditions.  This could
be true for paths of activation, species sensitivity and/or tolerance, gender-based differences in
response.

       CAS AC recommends expansion of section 5.1.3 on inflammation.  Inflammation is an
important pervasive mechanism, and additional highlights  of neutrophil biology and the
inflammatory process should be delineated. There should  also be a discussion of acute
inflammation as well as chronic inflammation.  This will undergird sections on acute responses
to air pollutants (e.g., bronchoconstriction, thrombosis and cardiovascular effects such as heart-
rate variability) versus responses influenced by chronic inflammation (e.g. proteolysis,
emphysema, fibrosis, atherosclerosis, and other persistent anatomic changes).

       CAS AC notes the omission of one important mechanism—epigenetic effects of PM
exposure.  We recommend the addition of a new section after 5.1.10 on this topic.  There is
increasing evidence of PM-induced epigenetic changes in DNA, particularly DNA methylation
and changes in histones.  These epigenetic changes  may be far more common and possibly more

-------
important than PM-induced changes in the primary structure of DNA.  Such epigenetic changes,
now readily measurable, can have a profound effect on gene expression and cytokine levels.
There are both animal studies as well as epidemiologic studies that demonstrate significant
epigenetic changes in response to PM exposure.

       We welcome the section entitled gaps in knowledge. It is an excellent addition to the
chapter.  Other gaps could also be listed:

          a.   Changes in effects due to timing of exposure and timing post-exposure
          b.   The anatomic spatial distribution of retained particles and the implications of the
              distribution, e.g., airways vs. the deep lung and retention at carinal ridges
          c.   Deposition, uptake, translocation, and clearance of ultrafine particles.

    6.  To what extent are the discussion and integration of evidence on the health effects of
       PM from the animal toxicological, human clinical, and epidemiologic studies,
       technically sound,  appropriately balanced, and clearly communicated? Does the
       integration of health evidence focus on the most policy-relevant studies or health
       findings?

          a.   Are the tables and figures presented in Chapters 6 and 7 appropriate,
              adequate and effective in advancing the interpretation of these health
              studies? To what extent does the Panel find Annexes C, D and E
              appropriate, adequate and effective in supporting the ISA?

          b.   In Chapters 6 and 7, toxicological studies were included in the PM ISA text if
              they were conducted at PM concentrations <2 mg/m3. The toxicological
              focus in these chapters was on inhalation studies, with intratracheal
              instillation studies and in vitro studies included only if they contributed
              significantly to the understanding of health effects from exposure to PM.
              The toxicological studies excluded from the text are presented in Annex D.
              What  are the Panel members' thoughts on this approach and the selection
              criteria?

       Chapter 6 does a reasonable job of integrating toxicology, human clinical studies and
epidemiological studies for each of the exposure scenarios discussed as well as for  each of the
outcomes of interest.  However, the summary text sections need to be edited for consistency in
approach and also to assure that the discussions are grounded and consistent with the antecedent
sections that describe the evidence. Generally, the text relating to evidence characterization was
satisfactory, but the review of details of the evidence was sometimes uneven. A discussion of
the basis for assigning a level of causality needs to be presented for each of the pollutant-
outcome causality statements.  Further discussion is also required for both Chapters 6 and 7 on
the level of causality for the PM components. Justification needs to be presented as to how
evidence related to PMio might have a lower level of certainty than PM2.5 (or PMio-2.s) when
     is comprised of both.

-------
       There is an appropriate discussion of the time-series studies, but this section needs to
have an explicit finding that the evidence supports a relationship between PM and mortality that
is seen in these studies. This conclusion should be followed by the discussion of statistical
methodology and the identification of any threshold that may exist.

       In general the tables and figures of Chapter 6 are used effectively to communicate the
copious information in a succinct manner. Annexes C, D and E are thorough, appropriate,
adequate and effective in supporting the goals of the PM ISA.

       With regard to the judgment of consistency of results, significant effects that move in
opposite directions (HRV effects) do not necessarily imply inconsistency, but rather the lack of
understanding of potential underlying mechanisms.  Results presented need to incorporate
specific issues of exposure, including concentration and duration, variability of exposure and PM
composition when possible. In reviewing the evidence, consideration also needs to be given to
the characteristics of the population studied.

       Inclusion of toxicological studies utilizing PM concentrations < 2 mg/m3 serves to limit
the inclusion of non-specific toxicological effects. Yet, even at these concentrations non-specific
or off-target effects can be observed. It is recommended to include high dose inhalation and
instillation studies only when the findings inform as to mechanisms leading to the biological
effects of PM. Effort should be made to distinguish inhalation from instillation studies
throughout the ISA.

   7.  What are the views of the Panel on the conclusions drawn in the draft ISA
       regarding the strength, consistency, coherence and plausibility of the evidence for
       health effects of PM? In evaluating the evidence to draw preliminary judgments on
       causality, EPA carefully considered evidence from the various scientific disciplines
       for the PM indicators and general health or environmental effect categories.
       Examples of a few specific health categories are listed below that were particularly
       difficult in reaching a causal determination.  We would appreciate CASAC
       comments on all of the causal determinations presented in this first draft ISA.

       •  Short-term exposure to PM2.s and cardiovascular and respiratory morbidity.
       •  Short-term exposure to PMi.s and cardiovascular and respiratory mortality.
       •  Short-term exposure to PMi0-2.s and respiratory and cardiovascular morbidity,
          and mortality.
       •  Long-term exposure to PMi.s and cardiovascular and respiratory mortality and
          lung cancer

       This  chapter provides a reasonably balanced presentation of most of the outcomes of
interest. However, there are several aspects of the presentation of findings that need to be
addressed in the second draft.  First, there needs to be specificity in interpreting associations that
come from cross-sectional studies as reflecting exposure differences across regions and not
increases in PM levels. Second, there needs to be greater balance in the presentation and
discussion of study results, without undue weight being given to positive findings or
characterization of estimates with confidence intervals that include the null as "positive".  This

-------
lack of balance is most evident for intermediate outcomes related to cardiovascular disease
(CVD). The summary statements at the end of the sections are more balanced, but the text needs
to be more even.  Further in the respiratory section there is a tendency to report primarily
positive endpoints from studies with multiple endpoints by drawing on the individual authors'
interpretations of their findings. In spite of this limitation, the overall conclusions regarding
causality in the respiratory section are reasonable.

       Studies support a link between long-term exposure of airborne particle pollution to
progression  of intermediate cardiovascular endpoints such as surrogate measures of
atherosclerosis, including coronary artery calcium and carotid artery internal medial thickness.
These studies must be interpreted in the context of their insensitivity to detect small changes in
either IMT or coronary artery calcium. In addition, the relation of chronic health impacts to
acute health impacts should be discussed.  In particular, this discussion should examine whether
chronic health impacts occur through similar or different biological pathways.

       Regarding the assessment of cancer mortality, particularly related to lung cancer,
CASAC recommends that the cohort study findings on mortality be incorporated in the cancer
incidence section rather than placing that discussion in the following section on mortality. For
lung cancer, because of the poor survival, mortality and incidence are quite close. The literature
search should be strengthened and better reflect the most relevant, earlier studies as well as
relevant occupational studies. The section needs to better consider the plausibility of lung
cancer, given the presence of known carcinogens in PM mixtures.

    8.  What are the views of the Panel on the definitions of susceptibility and vulnerability
       in Chapter 8? Are the characteristics included within the broad susceptibility and
       vulnerability categories appropriate and consistent with the definitions used?

       CASAC generally supports the inclusion of the material in this chapter as part of the ISA.
However, the title of the chapter is not descriptive of the content and should be revised, perhaps
to "Public Health and Welfare Considerations". In general, there was strong support for having
synthesis material such as that found in Chapter 8 to try to tie together in a succinct and policy-
relevant way material presented in more detail  in earlier sections. The introduction should more
clearly motivate the need for and implications of characterizing susceptible and vulnerable
subgroups. The policy-relevance of "vulnerable"  subgroups that tend to have higher exposures
suggests a need to better quantify inter-individual variability, as well as spatial and temporal
variability, in exposure, and to consider multiple PM sizes and components in the context of
exposure. The policy-relevance of "susceptible" subgroups is that they may have different
concentration-response functions than the general population or that "effect modifiers" need to
be considered.  The concentration-response function material would have a better context if
presented after the susceptibility material.

       Furthermore, impaired visibility has generally been treated as an aesthetic effect without
health-related consequences. In fact, most if not all visibility preference and valuation studies
have emphasized in their participant instructions, explicit guidance to value only the aesthetic
impact of impaired visibility and not to value any perceived health, psychological, or other
effects. The committee questions this separation.  It is well documented that impaired visibility
                                            10

-------
or haziness can cause feelings of gloominess, confinement, loss of freedom, and in some cases
depression.  Conversely, good visibility can cause one to have feelings of well-being and
freedom and can be linked to positive health outcomes.

       Several panelists expressed concern that the term "vulnerable" might not be the best term
to use in capturing the range of population characteristics that affect the risks to health from air
pollution exposure. CASAC recommends that key terms be carefully defined, such as
"susceptible," vulnerable," and "characteristics," and that a stronger conceptual framework be
developed for understanding the implications of these factors for risks to health.

       CASAC recommends that Table 8-1 be revised to be a more complete "road map" that
not only lists the "characteristics" of susceptible and vulnerable groups, but provides an
indication of the weight of evidence and strength of association for each. These can be indicated
qualitatively.  EPA is encouraged to either further expand Table 8-1 or create additional tables
that would more clearly lay-out how the "characteristics" map to health effects endpoints with
respect to time frame of the effect (long-term, short-term), PM size ranges (PM2.s, PMi0, PMi0.
2.5, and ultrafine), PM components of particular concern (e.g., black carbon, organic carbon,
elemental carbon, sulfate, nitrate, and transition metals), and health effect(s) of particular
concern (e.g.,  specific morbidity and mortality outcomes).

       CASAC recommends greater consistency in the content of Table 8-1 and the explanatory
text of the chapter, and for a consistent approach to defining and using terms and
"characteristics." For example, some "characteristics" are evaluated using  surrogates rather than
direct measurements (e.g., for socioeconomic status). Furthermore, there may be some
unavoidable overlap among characteristics and between susceptible and vulnerable
subpopulations.  For example, socioeconomic status may be associated with conditions that lead
to higher exposures (e.g., because of locations or activity patterns of the subpopulation) but also
to higher susceptibility (e.g., because of health care history and pre-existing conditions). For
purposes of transparency, EPA should identify and explain such overlap and the basis of a
judgment to assign an issue to a particular subgroup or category.

    9.  How useful and complete is the scientific evidence presented and summarized in
       Chapter 9 regarding the effects of atmospheric PM on the environment, including
       (a) effects on visibility, (b) effects on individual organisms, (c) direct and  indirect
       effects on ecosystems, (d) effects  on materials, and (e) effects on climate?  To what
       extent do the discussions and integration of evidence correctly represent and clearly
       communicate the state of the science?

       A summary of the Chapter 9 welfare effects should be added, and integrated with the
health  effects summaries in the Chapter 2  integrative overview. CASAC found that the Chapter
9 summary of welfare effects of PM (on visibility, individual organisms, ecosystems, materials,
and climate) was complete, if uneven.  The clarity of the chapter could be improved by adding
more detailed  explanations of fundamental concepts like urban vs. rural visibility and providing
better definitions of acronyms and specialized technical jargon. Some of the conclusions on
causality could be stated with more specificity, in ways that might provide a better indication of
                                           11

-------
which effects might be most responsive to changes in PM concentrations and exposure durations,
size fractions or chemical components.

       The section on PM effects on climate needs significant editing for clarity and focus and
there are some mistakes that need corrections. Potential benefits and harms of each PM
component should be summarized along with the uncertainties.

       This added specificity would be particularly useful for the section on climate effects,
which would benefit from more detail on differential effects of different PM components, the
relative contributions of PM components originating in the US to global aerosol and radiation
budgets, and discussion of potential health effects of PM-induced climate changes.

       The Chapter 9 discussion of visibility effects is substantially more detailed than the
discussion of other public-welfare effects. To a large extent, this is justified by the strong and
reasonably well-understood relationship between PM2,s concentration and visibility impairment.
However, the detail with which direct impacts on organisms and ecosystems are addressed is
inadequate relative to the amount of information available and the potential importance of the
issue. Important steps can be taken at this time toward causality determination, although it is
recognized that quantitative findings of causality  may not be possible from available data. In the
well-defined process of causality determination, it is clear that the intermediate levels imply
uncertainty that should drive future research to inform the next review cycle.

       Chapter 9's focus on "recent" visibility information,  available since the 2004 Criteria
Document, relies heavily on information generated through the Regional Planning Organizations
(RPO's) to support State Implementation Plans (SIPs) under EPA's Regional Haze Rule.  One
limitation of these data is that relatively little information is provided on visibility in the
suburban and urban areas, which are not protected by the Regional Haze Rule, and for which a
possible secondary PM standard could provide benefits complementary to those provided by the
Regional Haze Rule. Another limitation is that the PM/visibility association is described
exclusively in terms of the chemical-species-specific reconstructed extinction approach that
forms the basis of the Regional Haze Rule. While a similar  approach could conceivably be used
as an "indicator" for a secondary PM standard in urban areas, Chapter 9 could do a better job
laying the groundwork for considering other potential visibility indicators, such as the
relationship  with sub-daily PM2.5 indicator recommended in  the last NAAQS review cycle, and
the potential use of optical indicators.  There is strong support for using  a nephelometer and
aethalometer combination fitted with size selective inlets to routinely and continuously measure
fine and coarse particle scattering and absorption. In addition to measuring a variable directly
related to visibility, the measurement also yields an estimate of fine and coarse particle mass on a
semi-continuous basis. Additional discussion of the  strengths and limitations of the different PM
and optical measurement methods and their potential use as indicators for a secondary standard
would be very helpful.

       Panel members also suggest that additional discussion would be useful on the importance
of (and ways to include measures of) coarse particle contributions to visibility impairment which
are especially important in urban areas in western states.  Added discussion on the value of
continuous particle size distribution measurements would be useful, as would some discussion of
                                            12

-------
the importance of particle densities and differences between internally and externally mixed
aerosols.

       Chapter 9 includes literature citations that relate to the effects of impaired visibility on
psychological wellbeing, but it does not summarize the findings of those studies, nor does it
discuss the relevance of these findings to development of a more appropriate secondary standard
that is different in form from the present identical primary and secondary standards for PM.

       With respect to the eventual formulation of proposal(s) in this review cycle for the
secondary PM NAAQS, the scientific literature demonstrates visibility impairment from PM at
mass concentration levels below those of the current PM primary and secondary NAAQS and at
averaging times shorter than those of the current PM primary and secondary NAAQS. As the
Clean Air Act requires review and advice to the Administrator on secondary NAAQS by the
CASAC that specify a level of air quality requisite to protect the public welfare from any known
or anticipated adverse effects,  CASAC would expect to review policy-relevant summary results
describing EPA analyses relating the proposal(s) for light extinction indicator(s) for secondary
PM NAAQS to the proposal(s) for the primary PM NAAQS.  These analyses should also address
the variation and ranges of PM mass concentrations, chemical composition, size distributions,
and contributing sources across the nation relevant to the secondary PM NAAQS in order to
assist CASAC in providing advice to the Administrator.

   10. This first external review draft PM ISA is of substantial length and reflects the
       copious amount of research recently conducted on PM.  EPA has attempted to
       succinctly present and integrate the policy-relevant scientific evidence for the review
       of the PM NAAQS. Does the Panel have opinions on  how the document can be
       shortened without eliminating important and necessary content?

       We recognize the length of the document reflects the very extensive literature and data
that were considered. Some shortening might be accomplished by removing descriptions of
individual studies, unless highlighting of strengths and weaknesses of particular studies is needed
because of their importance. The document also needs a listing  of the acronyms to assist readers.

       We thank the Agency for the opportunity to provide advice early in the PM NAAQS
review process, and look forward to the review of the  Second Draft ISA in October 2009.

                                         Sincerely,

                                              /Signed/

                                        Dr. Jonathan M. Samet, Chair
                                        Clean Air Scientific Advisory Committee
Enclosures:  CASAC Paniculate Matter Review Panel Roster
            Compendium of Individual Comments
                                           13

-------
                                       NOTICE
       This report has been written as part of the activities of the EPA's Clean Air Scientific
Advisory Committee (CASAC), a federal advisory committee independently chartered to
provide extramural scientific information and advice to the Administrator and other officials of
the EPA. CASAC provides balanced, expert assessment of scientific matters related to issues 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 within the Executive Branch of the federal government. In addition, any
mention of trade names or commercial products does not constitute a recommendation for use.
CASAC reports are posted on the EPA Web site at: http://www.epa.gov/casac.
                                          14

-------
Enclosure A
                    Clean Air Scientific Advisory Committee
                         Particulate Matter Review Panel
CHAIR
Dr. Jonathan M. Samet, Professor and Chair, Department of Preventive Medicine, University
of Southern California, Los Angeles, CA

CASAC MEMBERS
Dr. Joseph Brain, Philip Drinker Professor of Environmental Physiology, Department of
Environmental Health, Harvard School of Public Health, Harvard University, Boston, MA

Dr. Ellis B. Cowling, University Distinguished Professor At-Large Emeritus, Colleges of
Natural Resources and Agriculture and Life Sciences, North Carolina State University, Raleigh,
NC

Dr. James Crapo, Professor of Medicine, Department of Medicine , National Jewish Medical
and Research Center, Denver, CO

Dr. H. Christopher Frey, Professor, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC

Dr. Donna Kenski, Data Analysis Director, Lake Michigan Air Directors Consortium,
Rosemont, IL

Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering ,
Georgia Institute of Technology, Atlanta, GA
CONSULTANTS
Dr. Lowell Ashbaugh, Associate Research Ecologist, Crocker Nuclear Lab, University of
California, Davis, Davis, CA

Prof. Ed Avol, Professor, Preventive Medicine, Keck School of Medicine, University of
Southern California, Los Angeles, CA

Dr. Wayne Cascio, Professor, Medicine, Cardiology, Brody School of Medicine at East
Carolina University, Greenville, NC

Dr. Douglas Crawford-Brown, Professor Emeritus, Department of Environmental Sciences and
Engineering, University of North Carolina at Chapel Hill, Chapel Hill, NC

Dr. David Grantz, Director, Botany and Plant Sciences and Air Pollution Research Center,
Riverside Campus and Kearney Agricultural  Center, University of California, Parlier, CA
                                         15

-------
Dr. Joseph Helble, Dean and Professor, Thayer School of Engineering, Dartmouth College,
Hanover, NH

Dr. Rogene Henderson, Senior Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM

Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Potsdam, NY

Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York
University School of Medicine, Tuxedo, NY

Dr. Helen Suh Macintosh, Associate Professor, Environmental Health, School of Public
Health, Harvard University, Boston, MA

Dr. William Malm, Research Physicist, National Park Service Air Resources Division,
Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort Collins,
CO

Mr. Charles Thomas (Tom) Moore, Jr., Air Quality Program Manager, Western Governors'
Association, Cooperative Institute for Research in the Atmosphere, Colorado State University,
Fort Collins, CO

Dr. Robert F. Phalen, Professor, Department of Community & Environmental Medicine;
Director, Air Pollution Health Effects Laboratory; Professor of Occupational & Environmental
Health, Center for Occupation & Environment Health, College of Medicine, University of
California Irvine, Irvine, CA

Dr. Kent Pinkerton, Professor, Regents of the University of California, Center for Health and
the Environment, University of California, Davis, CA

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. Sverre Vedal, Professor, Department of Environmental and Occupational Health Sciences,
School of Public Health and Community Medicine, University of Washington, Seattle, WA
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, EPA Science Advisory Board Staff Office,
Washington, DC
                                          16

-------
Enclosure B

                           Compendium of Comments
                   CAS AC Particulate Matter Review Panel on
     PM Integrated Science Assessment (First External Review Draft, Dec. 2008)

Comments from Dr. Lowell Ashbaugh	 2
Comments from Mr. Ed Aval	 4
Comments from Dr. Joseph Brain	 9
Comments from Dr. Wayne Cascio	 11
Comments from Dr. Ellis Cowling	 35
Comments from Dr. James Crapo	 39
Comments from Dr. Douglas Crawford-Brown	 41
Comments from Dr. David Grantz	 45
Comments from Dr. Joseph Helble	 49
Comments from Dr. Rogene Henderson	 51
Comments from Dr. Phil Hopke	 52
Comments from Dr. Donna Kenski	 55
Comments from Dr. Mort Lippmann	 60
Comments from Dr. William Malm	 65
Comments from Dr. Kent Pinkerton	 69
Comments from Dr. Robert Phalen	 73
Comments from Mr. Rich Poirot	 76
Comments from Dr. Ted Russell	 86
Comments from Dr. Frank Speizer	 89
Comments from Dr. Helen Suh Macintosh	 100
Comments from Dr. Sverre Vedal	 104

-------
Comments from Dr. Lowell Ashbaugh

Review of Chapter 9 - Ecosystem and Welfare Effects

       Charge questions
    1.  How useful and complete is the scientific evidence presented and summarized in Chapter
       9 regarding the effects of atmospheric PM on the environment, including (a) effects on
       visibility, (b) effects on individual organisms, (c) direct and indirect effects on
       ecosystems, (d) effects on materials, and (e) effects on climate? To what extent do the
       discussions and integration of evidence correctly represent and clearly communicate the
       state of the  science?

    2.  This first external review draft PM ISA is of substantial length and reflects the copious
       amount of research recently conducted on PM. EPA has  attempted to succinctly present
       and integrate the policy-relevant scientific evidence for the review of the PM NAAQS.
       Does the Panel have opinions on how the document can be shortened without eliminating
       important and necessary content?

The authors of this chapter have done an excellent synthesis of the relevant science regarding the
effects of atmospheric PM on the environment.  Some sections require technical editing for
grammar and spelling, but the technical evidence is complete and is presented well. The
presentation could be made more useful in several ways.
The spatial distribution of aerosol extinction and contribution of species to extinction in Figures
9-7 through 9-23 would be more useful if the contour intervals were standardized for groups of
related figures. As they are presented they show the  spatial patterns well but standardized
intervals and ranges would also illustrate the temporal changes. If it makes sense to standardize
the different species it would also better illustrate the relative impact of the different species in
different time periods. If the levels for each species are too different, it would still be helpful to
standardize the scales for each time period of a given species. It's also not clear on these figures
why the highest contour level has a different spacing than the others. That makes the red
(highest) contour levels misleading.
The sentence on lines 3-5 of page 9-31 is not clear. It's either a run-on sentence or is missing  a
crucial piece that would tie it together.
Because Figures 9-12 and 9-13 (also 9-14 & 9-15  and 9-16 & 9-17), are compared in the text, it
would be helpful to place each pair on the same page or on facing pages so the reader can
visually compare them more easily.
In the discussion comparing IMPROVE and CSN data, there should be some mention of the
different methods of handling blank subtraction. IMPROVE subtracts a blank from each sample
for ions and carbon analyses, but CSN does not. This makes comparing the concentrations
between the networks a little more difficult; not impossible, but not straightforward.
The discussion of IMPROVE and CSN carbon data should comment on the different methods
used to obtain the fractions. In fact, the discussion of carbon measurements in any network
should include the method used to make them and how that may affect the results. Comparison

-------
of total carbon between networks is less sensitive to the method used, but comparison of the
fractions is highly dependent on the method.
The statement on lines 7-8 of page 9-40 regarding fine soil and coarse mass is not entirely
correct. The different ratio suggests a different composition of coarse mass but does not
necessarily suggest a difference in the size distribution of suspendable soil materials. The
statement is valid only if the assumption is true that fine soil is a good surrogate for coarse mass.
In my experience that is just not true at all.
There is a mention on page 9-40  of the confounding effect of meteorologically driven interannual
variability, but I would like to see an explicit statement calling for the investigation of the
meteorological and climate-driven  effects on particle concentrations and speciation. This  is
particularly relevant for the comment on page 9-44 regarding the depression of wintertime
particulate  nitrate for several years measured by the IMPROVE network. This feature may be
due to meteorological influences, but has not yet been investigated.
There are several mentions of "upwind sources" in the discussion on page 9-46. Presumably, this
refers to the San Francisco Bay Area, but it should be more clearly stated.
I found a couple of references that were missing or misdirected. All references should be
checked for accuracy. The reference to McDade (2006) is not in the list of references (see line 14
on page 9-44). The reference for the comment on lines 6-9 of page 9-48 (Pitchford, et al., 2007)
does not seem to be accurate. I could not find those statistics in that paper. I'm also skeptical of
the statement on line 11 of page 9-50 that over half the sulfate in remote areas of the Pacific
coastal sites is from outside the U.S. If this is a model result, it should be labeled as suspect.
The discussion of fine particles on page 9-84 should mention nanoparticle emissions as a  direct
source of fine particles. On page 9-87, "occult deposition" should be defined.
On page 9-88, the discussion of dry deposition should include mention of the surrogate surface
work and sampler developed by Dr. Cliff Davidson of Carnegie-Mellon University.
I believe there is some redundancy in section 9.5.1.2, and possibly elsewhere in the document. I
suspect this may be due to the number of authors required to produce such a comprehensive
document.  It may be possible to reduce its size by having a single person go through the entire
document for consistency and to remove redundancy. This task may be too time-consuming to
actually carry it out, though.
I will list a few detailed comments:
    •   On page 9-2 is the statement that nitrate important in most of California. It would be accurate to
       say  "much of California", but not "most".
    •   Equation 9-2 needs work. It looks like it was copied unsuccessfully from another document. For
       example, "Large" is printed as "Z arg e". Also, the text refers to bold text but there is none.
    •   There are grammatical errors throughout the chapter, but most of them are before section 9.3.5.
       For example, there are several uses of the word "dominate" that should be "dominant". There are
       cases of misplaced parentheses and some minor typographical spelling errors. The chapter needs
       to be reviewed carefully for grammatical accuracy.

-------
Comments from Mr. Ed Avol

General Comments:
The First Draft ISA is a remarkable overview of the breadth of recent PM research conducted
primarily in the United States over the past several years. EPA staff should collectively be
congratulated and acknowledged for distilling an amazing array of data from multiple sources
into one central location for consideration.  The authors and reviewers listed as participating in
the document's preparation are among the acknowledged experts in their respective fields, and
EPA should be congratulated for assembling such a talented team to review this vast amount of
material. There are, inevitably, differences of opinion in inferences, nuances, approaches, and
conclusions, but the major accomplishment of assembling this material should not go unnoticed.

There are at least two comments/concerns I have with the stated approach. I was disappointed in
the decision to minimize or dismiss extended consideration of ultra-fine particles or traffic-
related pollution exhaust.  The accumulating number of carefully-performed and peer-reviewed
research describing the observed associations between fresh combustion exhaust (primarily from
motor vehicles) and a range of health outcomes (respiratory, cardiovascular, birth-related, and
even developmental) suggest that there is a rapidly accumulating body of information on which
to objectively and seriously consider the need for a criteria standard. If this consideration does
not move forward within the context of the periodic review of the current PM standards, I am
concerned about what and where the EPA believes the appropriate venue to be.  Other countries,
most notably in the European Union, have moved forward in quantifying, monitoring, and
regulating ultra-fine PM, while the United States has been slower to respond.  In my opinion, this
decision should be re-considered and re-opened for discussion.

I was also disappointed to see the decision to effectively restrict document consideration of
health endpoints to primarily respiratory and cardiovascular results.  Although the discussion
"door" has arguably been left slightly ajar to allow some natal or pre-natal health outcomes to
possibly stand in the proverbial doorway, the decision to not endorse and include these and other
developmental outcomes as important and critical in our overall understanding of PM health
effects seems outdated and short-sighted.  Our responsibility to review the science and make
recommendations that will lead to the development of standards that protect the public health
extends to the full distribution of the population, and ought to include the widest breadth of
available credible scientific information in those deliberations.  Our understanding of the
mechanisms and pathways of exposure and effect have evolved beyond limiting our review to
primary target organs of immediately present receptors, and our consideration of viable data for
deliberations in standard-setting ought to, as well.

The document needs  some clarifying comments regarding health outcomes to be considered
based on the strength of relationships. The current implication that only the two uppermost
levels — causal and likely causal - are being considered beyond this initial point is apparently
mistaken, based on Staff comments, but the document leaves the reader with that impression.

The Agency's "official" perspective on Climate Change is inconsistently presented and
prioritized in the current document (mostly in Chapter 9). Given the importance of carbon
                                            4

-------
aerosol in both air pollution and global climate matters, the ISA should raise the visibility and
discussion of such issues.

Finally, no other pollutant so readily lends itself to consideration of multi-pollutant mixtures and
multi-pollutant exposures than PM.  The chemical variability, as well as the size and source
differences, make this an especially challenging exposure of interest, but the opportunity to view
Pm as a multi-pollutant entity and develop a means of dealing with it in the context of the review
process would be an invaluable step forward towards true consideration of multi-pollutant
mixtures, compared to the current, and much less-realistic, approach of considering individual
specific pollutants.

Specific Comments
Pl-14, lines 12-30 - Emphasis on clinical studies, given both the multi-pollutant nature of
ambient exposures and the multi-compositional (and size) nature of PM, makes elucidation in
clinical studies even a higher threshold to surmount than for other pollutant reviews.  Is there
some way to take this into consideration?

Chapter 3: Source to Human Exposure
P3-1, Introduction - There is a lot of discussion here (in the Introduction?) about soluble organic
aerosols (SOAs), without adequate justification as to why or what other other key constituents
might be present.  In my opinion, this should be moved or at least prefaced by some transitional
justification.

P3-1, lines 20-29, P3-2 lines 1-10 - move this to after SOA discussion in Section 3.3.

P3-2, lines20-22 - "Within a street canyon,..." seems to be tacked on to this paragraph and out
of place (no real connection to the paragraph, except only in the broadest possible sense).

P3-2, line 23 - "receptor modeling studies indicate that the  main source..." sounds odd - isn't
receptor modeling focused on receptors?

P3-2, line 27 - Is the observation that "vegetative burning and traffic-related emissions were less
consistently identified...", both which would presumably be in smaller-than-PM2.5 size ranges
close to the source, really a surprise? PM2.5 has a more regional spatial character.

P3-9, line 1 - It would be more accurate to begin this sentence "US national average.

P3-23, line 10 - So what is the conclusion one is to draw from all of this?

P3-26, Iine20 - There is LOTS of information presented on instrumentation and sampling; what
to do with all of this to make it accessible? A summary table (such as the one in the  Annex)
would be helpful here, or reference to the annex table...

P3-26, line 21 - Why  does this paragraph begin with this comment about "maximum
concentration sites"?  I would suggest that the paragraph begin with the sentence "... The AQS

-------
contains measurements..." in line 23, and the previous two sentences be reversed in order and
moved to P3-27, line 2.

P3-27, lines 8-9 - Isn't this description about neighborhood scale monitoring? How do air
pollution episodes relate to micro <100m) or middle (100-500m) scales?

P3-27, line 15 - "Urban scale - This scale applies for assessment of air quality at an urban scale"
is not much of an explanation...!

P3-27 - Reference is made to a number of different monitoring networks (NAMS, SLAMS,
CSN, IMPROVE, NCORE,...) without ever explaining what they are or how they overlap or
differ. How about a summary table showing the major networks, the frequency of sampling,
typical monitoring array, sampling requirements, etc?

P3-28, line 4 - replace "built" with "located"

P3-28, line 21 - ".. .local conditions..." comment should be clarified; it refers to atmospheric or
meteorology (temperature, humidity, pressure).

P3-33, Tables 3-5 to 3-10, include units in title ("...aged _ years..")

P3-38, line 12-13 - The comment is made here that there is a large amount of new information
about ultra-fines.. .but then the decision is made to not consider it in the context of the PM
review... seems a bit inconsistent.

P3-38, lines 27-28. While I don't disagree with the Watson et al comment about spatial scales of
interest, from a health standpoint, micro or middle scales (100s of meters) may be a critically
important consideration (for example, in the context of proximity to busy roads).

P3-44, Table 3-12 - where was the 828 ug/m3 1 hour average reading observed,  and what was its
explanation?

P3-85, Figure 3-41 - It seems  surprising that naphthalene, which typically is present in
concentrations a magnitude higher than other PAHs, is in the bottom half of these plots... ?

P3-90 - The layout of figures and text through this section is confusing to the reader, in that the
text under a given figure does  not describe or relate to that figure... can this be adjusted?

P3-92 - What is the purpose and utility of this "EPA region office" presentation  of data?  It is
arguably a bit deceptive, since Region 9 (for example) shows a nice downward or flat trend in
Figure 3-47 that is below the NAAQS line, but within that region [Southern
California/Riverside] large areas are out of compliance and will remain so for several years to
come (according the AQMP estimates by the local agency).

-------
P3-94, Figure 3-49 - Here, too, the plots by EPA region office is deceptive.  The trend lines look
to be well under the NAAQS, but Southern California is still out of compliance.  Are the most
appropriate metrics being plotted and presented?

P3-132 - The discussion on this and the previous pages about temporal, spatial, and physical
activity effects on exposure seems a little thin. There has been a LOT of work looking at
personal exposures, time-activity, and spatial determinations (diary-based, for the most part, until
recently).  None of it, as it relates to PM exposure, is summarized or referenced here...?

P3-134, "New Developments in Microenvironmental Exposure Monitoring Techniques -1
expected to see something about the NIH Exposure Biology Initiative samplers and innovations
here (lab-on-a-chip, miniaturized samplers, etc) but no apparent mention...?

P3-135, line 4 - replace "guiet"  with "quiet"

P3-135, lines 4-10 - "Personal clouds", variable locations, and other microenvironmental
considerations may provide legitimate reasons for differences between personal measurements
and fixed-site ambient measurements.  The paragraph  in the text makes it sound like the personal
measurements are less credible or somehow fall short of being "equal" to a fixed-site monitor,
when the true state of reality might very well be the reverse.

P3-138, lines 9-19 - Important points are raised in this section, but it should also be noted that
one might infer from this discussion that zip codes are the appropriate unit for community
assessment and assignment of exposure. This could be quite misleading in areas where the
population is such that a given zip code covers a wide  geographic, or variable topographic, or
variable trafficked/road-coverage/vehicle-loaded,  area. Accordingly, there should be some
caveats or considerations associated with the document text.

P3-138, lines 20 and 22 - replace "among" with "between"

P3-147, line 13 - Is "0.3%" a printing error?  This doesn't seem like much to be concerned
about...?
Miscellaneous Comments:
Chapter 6 is important, but could be dramatically trimmed in size by moving the "criteria-type"
litany of study reports to the annex (e.g., page 6-112 discussion about asthmatic children and the
page 6-116 discussion on asthmatic adults), and replacing these with a chapter-wide, more
consistent summary of recent studies.

Chapter 8 was difficult to integrate. The title (Public Health Impacts) seems misleading, since
what was presented were portions about concentration-response relationships and susceptibility
vs vulnerability. Table 8-1 is an evolutionary improvement, but the vulnerability portion of the
table is inconsistent and in need of revision.  Some of the listed factors are subsets of other
factors, while still others don't quite fit with the nomenclature as described.  Alternatively, this
sort of presentation and consideration is a real improvement and should be continued (but this

-------
specific presentation needs some work). Arguably, Chapter 8 should appear as the executive
summary, or as the concluding chapter, but its current location seems out-of-place.

-------
Comments from Dr. Joseph Brain

Chapter 1:1 presume that this document, the ISA-PM draft, sets the stage for recommendations
regarding the confirmation or changes in the current NAAQS Standard for PM.  I appreciate
Table 1-1, which outlines the evolution of this standard, from 1971-2006. But the document is
less clear as to the future of the PM standard. What options are being considered? Would
changes be likely in regard to the size of the aerosol, the level of the aerosol, or the averaging
time? As we consider the evidence, it would be helpful to know the options that are being
considered.

Section 1.2 delineates many of the issues which have been critical in the past. But some
additional paragraphs at the end of this section should highlight the most controversial aspects of
the latest EPA final decision (September 21, 2006), and both the timing of future decisions and
the possible scope. This would help CASAC and the public know what to look for. What are
the critical issues, e.g.  level, indicator size distribution, or averaging time? Are there other issues
that need to be considered, e.g. the relationship between outdoor and indoor air?

Is this a topic that should be addressed?  The majority of time for US citizens is  spent in indoor
environments. This may become an important driver of environmental considerations for human
health. Don't we need to do a better job in discussing the extent to which indoor air - although
not regulated by the EPA - is influenced by outdoor PM sources?

Chapter 3: The fields of nanotechnology and nanotoxicology are expanding rapidly.  In addition
to highly engineered novel particles, such as fullerenes and quantum dots, we are also
recognizing that common polydisperse aerosols, such as diesel particles or road dust, have a nano
component.  Data in Chapter 3 show that there are significant numbers of particles that are
smaller than 0.1 um (100 nm).  The presence of nanoparticles should be  explicitly
acknowledged. In this and in other chapters, we need to acknowledge the possibility that these
small particles have a significant role which is out of proportion to their  contribution to mass.
Even though their relative weight is small, their relative number and surface area may be large.
This is an area of active research, and this should be acknowledged. Some of this is included in
Section 3.4.1.3, but the fact that ultrafmes are comparable to nanoparticles should be
acknowledged, and the growing body of research on nanoparticles  should - in the future - be
incorporated.

I'm also concerned about Section 3.4.1.2, PM Speciation. There is a section on  speciation, as
well as on cations, but I believe that there should be a section entitled "Metals."  There is
increasing evidence that metal constituents of PM such as iron, manganese, vanadium, or
manganese are responsible for many of the PM effects. For example, certainly on a mass basis,
they are more toxic than elemental carbon.  Some data on the variety of metals seen and their
health effects should be noted in Chapter 3.  PM metal content is an important determinant of
toxicity.

Chapter 3 also is weighted down by a large  amount of figures showing spatial and temporal
variation of PM and PM components. Is this all necessary? It's difficult for the reader to learn
much from some of these data.  For example, see Figure 3-22 or Table 3-15.  See also Figure 3-

-------
29 and Table 3-17. Should some of these data be eliminated or put into an appendix? It's heavy-
going for much of this chapter, and it's difficult to identify the most important conclusions. This
point is underlined by the shear number of figures in Chapter 3.  There are 81 figures. That may
be excessive.
                                          10

-------
Comments from Dr. Wayne Cascio

Chapter 6

General comments: The authors, contributors, and reviewers of Chapter 6: Integrated Health
Effects of Short-Term PM Exposure of the ISA for Paniculate Matter have produced an
comprehensive compendium of environmental health data that in large part achieves its goal of
providing a thorough source of relevant human clinical and epidemiological data, and animal
toxicological data along with an integrated and balanced discussion of its public health
implications. This draft document appears to be unprecedented in scope and detail. In spite of
the broad range of content and complexity the descriptions of the cited studies have an
exceptional degree of accuracy and integration. The discussion is in general balanced, and the
evidence supports the conclusions. The style of the PMISA for is consistent and the document is
easy to read and the concepts are clearly communicated.
In general the Tables and Figures of Chapter 6 are used effectively to communicate the copious
information in a succinct manner. Annexes C, D and E are appropriate, adequate and effective in
supporting the goals of the PMISA.

The decision to include in Chapters 6 and 7 only toxicological inhalation and intratracheal
instillation studies when conducted at PM concentrations <2 mg/m3, and in vitro studies when
they contributed significantly to the understanding of health effects is logical because it is under
these conditions that such studies provide the most support of biological plausibility for the
health effects.
Specific Comments:

Figure 6-1. Excess risk estimates per 10|ig/m3 increase in PMi0, PM2.s, and PMi0 2.5 for studies
of CVD ED visits and hospitalizations.  Appropriate. The PMio 2.5 should read PMio-2.5- The
graphics of Figure  6-1 could be improved by placing the PMio, PM2 5, and PMi0-2.5 on the left
hand side of Figure 6-1 in the same way as it is done in Table 6-6.

Figure 6-2; 6-3; 6-4. The graphics could be improved by placing the PMio, PM2.5, and PMio-2.5
on the left hand side of Figure 6-2; 6-3; 6-4 in the same way as it is done in Table 6-6.

6-8, L. 3. ApoE-/- mice. This genetic model of atherosclerosis in the mouse is primarily a model
of peripheral vascular disease rather than coronary artery disease.  In the future other models
should be incorporated into the atherosclerosis research, arrhythmia and cardiac dysfunction.
Models that for example that modify the function of cells and proteins involved in the
atherosclerotic process should be incorporated. Mouse models available today that might provide
more relevant information for coronary heart disease include the ApoE/LDL receptor double
KO, the macrophage uPA-transgenic/apoE KO, SR-BI/apoE dKO, and the SR-BI/ApoE-Rei11711

6-8. 6.2.1. Heart Rate Variability. The measures of heart rate variability  (HRV) include time- and
frequency-domain  measures. The interpretation of HRV finding continues to be a challenge
                                           11

-------
because studies judged to be well designed and with appropriate methods continue to show
disparate findings. These unanticipated outcomes rather than explained by chance perhaps are
indicating that the HRV result is reflecting a fundamental response of an individual that is
determined in part by a number of factors including age and the presence of co-morbid
conditions. It is also possible that time- and frequency-domain measure of HRV are inadequate
to capture the true nature of PM's effect on heart rhythm. For example, no data contained in the
previous AQCD or the current ISA describes any studies utilizing non-linear measures of HRV.
Non-linear functions capture the structure and complexity of RR intervals and have the potential
to provide a more detail of randomness of the RR dynamics and possibly be more prognostic.

6-16, L. 4. With  respect to the paper by Gong et. al. 2003a,  "This effect was observed
immediately following the exposure and at 2-days post-exposure, ...", it should read ".. and at 1
day post-exposure, ...".

6-23, L. 13-25, The papers referenced regarding the assessment of arrhythmia by Dockery and
colleagues (Dockery et al. 2005a, and 2005b were the first studies to investigate the association
of air pollutants with arrhythmia utilizing an advanced internal cardiac defibrillator with the
capability to store electrograms thereby allowing a qualified cardiologist to distinguish
arrhythmia from artifact, and supraventricular from ventricular arrhythmia.  This is an important
point because such characterization was not present in the original paper by Peters et al. in 2000.

6-24. L. 11. "dame" should read "same".

6-24. L. 12. "frome" should read "frame".

6-27, L. 1-4.  The following  statement "Ectopic beats  are defined as extra cardiac depolarizations
and are the most common disturbance in heart rhythm. Ectopic beats are usually benign, and may
present with or without symptoms, such as palpitations or dizziness. When three or more occur
in succession, this is called a non-sustained ventricular tachycardia." should be rewritten because
of some inaccuracy. Consider the following, "Ectopic beats are defined  as heart beats that
originate at a location in the heart outside of the sinus node. They are the most common
disturbance in heart rhythm. Ectopic beats are usually benign, and may present with or without
symptoms, such  as palpitations or dizziness. Such beats can arise in the atria or ventricles. When
the origin is in the atria the beat is called an atrial or supraventricular ectopic beat. When such a
beat occurs earlier than expected it is referred to as a premature supraventricular or atrial
premature beat.  Likewise, when the origin is in the ventricle the beat is defined as a ventricular
ectopic beat, or when early a premature ventricular beat. When three or more occur ectopic beats
occur in succession, this is called a non-sustained run  of either supraventricular (atrial) or
ventricular origin. When the rate of the run is greater than 100 beats per minute it is defined as a
tachycardia."

6-29. L. 15. For completeness the following sentence, ".... likely owing to their rapid heart rates
(-600 and -350 bpm, respectively)." should be amended to, "... likely owing to their rapid heart
rates (-600 and -350 bpm, respectively) and repolarizing currents.".

6-29, L. 18-19. For accuracy the following phrase, ".. .ventricular depolarization elicits the QRS
                                           12

-------
complex, and the biopotential recovery of the ventricles is reflected by the T-wave. " should be
amended to, ".. .ventricular depolarization elicits the QRS complex, and the T-wave represents
the repolarization of the ventricles."

6-33. 6.2.3.2. Human Clinical Studies. L. 21-26. Mills et al. (2007) assume that increased ST
depression during exercise represents an increased magnitude of ischemia. This is a reasonable
conclusion particularly in view of canine studies indicating that CAPS decreases coronary artery
blood flow (Bartoli et al. 2008). Yet, it is not proven that the ST depression in man under
conditions of PM exposure is caused by greater ischemia. An alternative interpretation is that the
greater ST depression could be secondary to heterogeneity of electrophysiological effects of
particulate exposure on the myocardium that is enhance by the metabolic and ionic conditions
associated with ischemia or increased heart rate.

6-36. L. 4. "bronchia" should read "bronchial".

6-40. L. 8-16. Brauner et al. 2008 showed that eliminating particles with a HEP A filter in the
homes of healthy aged adults improved  microvascular function significantly.  The effect was
attributed to the removal of PM as gases were not impacted by the HEP A filter. These findings
are similar to those observed in the California Freeway Study funded by the California Air
Resource Board under the direction of Dr. William Hinds at UCLA School of Public Health
where the filtering of PM from the passenger cab of a van containing research subjects riding on
the 405 and 710 freeways in Los Angeles, California experienced fewer arrhythmias and a
relative decrease in pro-NT BNP. These findings (Cascio et.  al. 2009) will be presented at
American Thoracic Society in May 2009.  Like Brauner et al. 2008 these outcomes were based
on the elimination of PM rather than gases or the stress associated with being a passenger while
traveling on the freeway.

6-52. L 5-7.  For clarity the "QA interval" can be used interchangeably with the term "systolic
time interval". The term "QA interval"  appears throughout the text. For those less familiar with
the terminology using the "QA interval" might be confusing. A reference might be helpful as
well. For example, Cambridge D and Whiting MV. Evaluation of the QA interval as an index of
cardiac contractility in anaesthetized dogs: responses to changes in cardiac loading and heart
rate. Cardiovascular Res.  1986: 444-50, would provide the basis for this measurement.

6-49. L. 7-16. Wellenius et al. 2007 reported that in a group of subjects with LV dysfunction that
PM did not affect measurements of BNP, failing to prove the hypothesis that PM can worsen LV
function.  Yet, BNP the active peptide has a very short half-life and might not be the best
biomarker for such a study. NT-pro BNP, the inactive peptide also responsive to LV ventricular
pressure but with a substantially longer half-life might have been a better biomarker to assess
transient changes in LV function associated with PM and other air pollutants. Thus the absence
of a correlation between PM and BNP does not prove that PM does not have an impact on RV or
LV function in individuals with impaired cardiac mechanics.

6-59. Section 6.2.8. "Blood Coagulation" is more appropriately named "Hemostasis, Thrombosis
and Coagulation Factors".
                                           13

-------
6-97. L.  15-7. Just published in Environmental Health Perspectives are the findings of a CAPS
coarse particle study. The reference is Graff DW, Cascio WE, Pappold A, Zhu H, Huang Y-C T,
Devlin KB. Exposure to concentrated coarse air pollution particles causes mild cardiopulmonary
effects in healthy young adults. EHP doi: 10.1289/ehp.0900558. Available 23 March 2009.  In
this study 14 healthy young volunteers were exposed to coarse concentrated ambient particles
(CAPS) and filtered air. Coarse Particle concentration averaged 89.0 ug/m3  (range was 23.7 -
159.6). Volunteers were exposed to coarse CAPS and filtered air for two hours while undergoing
intermittent exercise in a single-blind, cross-over study. Pulmonary, cardiac, and hematological
endpoints were measured before exposure, immediately after exposure, and  again 24 hours after
exposure. Compared to filtered air exposure, coarse CAPS exposure produced mild pulmonary
inflammation as evidenced by a small increase in polymorphonuclear neutrophils (PMNs) in the
bronchoalveolar lavage (BAL) fluid 20 hours post-exposure. Changes in pulmonary function
were not observed. Blood tissue plasminogen activator (tPA) was decreased 20 hours after
exposure. SDNN, a measure of overall HRV was decreased 20 hours after exposure to CAPS.
Consequently it is concluded that coarse CAPS exposure induces a mild physiologic response in
healthy young volunteers approximately 20 hours after exposure. These changes are similar in
type and magnitude to changes reported previously for healthy volunteers exposed  to fine CAPS,
suggesting that both size fractions are comparable at inducing acute cardiopulmonary changes.

6-99. L. 10. "PM2." Should read "PM2.5".

6-100. L. 21 and 23. "H" should read either "HR" or "heart rate".

6-101, L. 29. "... organic and EC fraction.." should be written either as "...organic and elemental
carbon...", or "OC and EC".

6-106. L. 22. "H" should read "heart rate".

6-107, L. 23-24. "There is one study that used an indirect measure of cardiac contractility (QA-
interval) during ultrafme CAPs..." should be written as, "There is one study that used an indirect
measure of cardiac contractility, the systolic time interval, as assessed by the QA-interval during
ultrafme CAPs...".

6-125, L. 30-32., and 6-126, L.I. "Pietropaoli et al. (2004) observed a significant reduction in
MMEF and DLCO in healthy adults 21 hours after a 2-h exposure to ultrafme carbon particles
(50 ug/m3). This reduction in DLCO may reflect a PM-induced vasoconstrictive effect on the
pulmonary  vasculature."  This observation might be important to the observation in the California
Freeway Study (Cascio et al.  accepted for presentation at ATS May 2009) in which exposure to
traffic related particles produced a relative increase in arrhythmia and NT-proBNP  compared to
filtered air.  The release of NT-proBNP is consistent with an acute increase in pulmonary
pressures due to pulmonary vasoconstriction with attendant RV stress.

6-134, L. 13. "Differenct" should be written "Differences".

6-142, L. 25-26. Write out, "fme+ultrafine".
                                           14

-------
6-6-150. L. 10 . "patter" should read "pattern".

6-197, L. 22. "domaminergic" should be written as "dopaminergic".

6-198, 6.4.3. Summary and Causal Determination. It is agreed that the weight of the evidence is
inadequate to assign a causal relationship between PMio, PM2 5, PMio-2.5, ultrafine particles, or
specific PM components to CNS outcomes. Yet, this data is some of the most intriguing. These
early findings should stimulate more comprehensive studies to elucidate mechanism in the
animal models and man.

6-240, L. 16. "causes-specific!" should be written, "cause-specific".

6-243, L. 11. "HAs" should be written out as, "hospital admissions".


Chapter 7.

The conclusions related to causation are conservative and supported by the evidence.

7-2, L. 17-19.  The statement, "The advanced lesion that develops from this process can occlude
perfusion to distal tissue, causing ischemia, and erode,  degrade, or even rupture, revealing
coagulant initiators (tissue factor) that cause major clotting disorders and infarction or stroke."
Should be clarified. Consider the following, "The advanced lesion that develops from this
process can occlude perfusion to distal tissue, causing ischemia, and erode, degrade, or even
rupture, revealing initiators of thrombosis (tissue factor) that cause major disorders caused by
arterial thrombosis and thromboembolism such as coronary artery syndrome or stroke."

7-3, L. 8. The statement, "Agatston scores are frequently used to classify individuals into one of
five groups (zero; mild; moderate; severe; extensive) or according to age- and sex-specific
percentiles of the CAC distribution (Erbel et al., 2007)." Should be amended to read, ".. .the
CAC distribution (Erbel et al., 2007) and as such the greatest utility of the Agatson score is in
ascribing a long-term clinical risk for cardiac events. Yet, the Agatson score is not sensitive to
low levels of vascular calcium and as such this insensitivity might limit its utility in longitudinal
epidemiological studies."

7-3, L. 9-20. Carotid artery internal medial thickness is an established  measurement of large
vessel atherosclerosis.  Yet, recent clinical trials have raised the issue of its suitability as a
measure of progression of atherosclerosis.  As recent study puts this issue into perspective.
(Doggan S, Plantinga Y, Evans GW, Meijer R, Grobbee DE, Cots MS, Opal Investigators.
Ultrasound protocols to measure carotid intima-media thickness: a post-hoc analysis of the
OPAL study. Curr Med Res Opin 2009; 25(1): 109-22.). In their study they  showed that in
"healthy middle-aged  subjects mean common CIMT protocols that include measurements at both
near and far walls at multiple (> or = 2) angles provide highest reproducibility combined with
largest estimates of CIMT progression measured with high precision..." Such approaches are
generally not taken this most likely limits the sensitivity of the assay.
                                            15

-------
7-4, L. 3-14. As discussed previously the Agatston scale is expected to be insensitive to small
changes in vascular wall calcium.

7-84, L. 27. The font of "infant mortality" should be changed to match the text.

7-86, L. 4. The "n" should be written out as "numbers of subjects".

Chapter 8.

Comment: The C-R relationship focuses on mortality effects with short- and long-term exposure
to PM. The mortality endpoint is a hard endpoint but is also less sensitive indicator of health
effects attendant to PM. Hospitalizations for clinical conditions are likely to provide greater
sensitivity in determining the C-R relationship. From a public health perspective having a more
sensitive indicator will be important to fully  understanding the potential health impact of PM,
particularly at low concentrations. Yet, such data is limited and not  sufficient to establish firm
conclusions. More data is needed in this area.

The definitions of susceptibility and vulnerability are logical and functional. Such definitions can
be effectively used to categorize and summarize clinical studies with similar characteristics.

Pg. 8-6. L. 2. Spelling. "Medixare" should read "Medicare".

References

R-3. The full citation is now available: Andersen, Z. J., P. Wahlin, et al. (2008). "Size
distribution and total  number concentration of ultrafine and accumulation mode particles and
hospital admissions in children and the elderly in Copenhagen,  Denmark." Occup Environ Med
65(7): 458-66.

R-7. Full reference now available. Bell, M. L., K. Ebisu, et al. (2008). "Seasonal and regional
short-term effects of fine particles on hospital admissions in 202 US counties, 1999-2005." Am J
Epidemiol 168(11): 1301-10.


New References of Relevance not included in PMISA and their  Abstracts

Carder, M., R. McNamee, et al. (2008). "Interacting effects  of particulate pollution and
cold temperature on cardiorespiratory mortality in Scotland." Occup Environ Med 65(3):
197-204. To determine whether the effect of black smoke on cardiorespiratory mortality is
modified by cold temperatures. METHODS: Poisson regression models were used to investigate
the relationship between lagged black smoke concentration  and daily mortality,  and whether the
effect of black smoke on mortality was modified by cold temperature for three Scottish cities
from January 1981 to December 2001. MAIN RESULTS: For all-cause respiratory and non-
cardiorespiratory mortality,  there was a significant association between mortality and lagged
black smoke concentration.  Generally the maximum black smoke effect occurred at lag 0,
although these estimates were not statistically significant. A 10 mugm(-3) increase in the daily
                                           16

-------
mean black smoke concentration on any given day was associated with a 1.68% (95% CI 0.72 to
2.65) increase in all-cause mortality and a 0.43% (95% CI -0.97 to 1.86), 5.36% (95% CI 2.93 to
7.84) and 2.13% (95% CI 0.82 to 3.47) increase in cardiovascular, respiratory and non-
cardiorespiratory mortality, respectively, over the ensuing 30-day period. The effect of black
smoke on mortality did not vary significantly between seasons (cool and warm periods). For all-
cause, cardiovascular and non-cardiorespiratory mortality the inclusion of interaction terms did
not improve the models, although for all-cause and non-cardiorespiratory mortality there was a
suggestion for interaction between temperature and recent black smoke exposure. The results of
this study suggested a greater effect of black smoke on mortality at low temperatures.  Since
extremes of cold and particulate pollution may coexist, for example during temperature
inversion, these results may have important public health implications.

Franco Suglia, S., A. Gryparis, et al. (2008). "Association between traffic-related black
carbon exposure and lung function among urban women." Environ Health Perspect
116(10): 1333-7. BACKGROUND: Although a number of studies have documented the
relationship between lung function and traffic-related pollution among children, few have
focused on adult lung function or examined community-based populations. OBJECTIVE: We
examined the relationship between black carbon (BC), a surrogate of traffic-related particles, and
lung function among women in the Maternal-Infant Smoking Study of East Boston, an urban
cohort in Boston, Massachusetts. METHODS: We estimated local BC levels using a validated
spatiotemporal land-use regression model, derived using ambient and indoor monitor data. We
examined associations between percent predicted pulmonary function and predicted BC using
linear regression, adjusting for sociodemographics (individual and neighborhood levels),
smoking status, occupational exposure, type of cooking fuel, and a diagnosis of asthma or
chronic bronchitis. RESULTS: The sample of 272 women 18-42 years of age included 57% who
self-identified as Hispanic versus 43% white, and 18% who were current smokers. Mean +/- SD
predicted annual BC exposure level was 0.62 +/- 0.2 microg/m3. In adjusted analysis, BC (per
interquartile range increase) was associated with a 1.1% decrease [95% confidence interval (CI),
-2.5% to 0.3%] in forced expiratory volume in 1 sec, a 0.6% decrease (95% CI, -1.9% to 0.6%)
in forced vital capacity, and a 3.0% decrease (95% CI, -5.8% to -0.2%) in forced mid-expiratory
flow rate. We noted differential effects by smoking status in that former smokers were most
affected by BC exposure, whereas current smokers were not affected. CONCLUSION: In this
cohort, exposure to traffic-related BC, a component of particulate matter, independently
predicted decreased lung function in urban women, when adjusting for tobacco smoke, asthma
diagnosis, and socioeconomic status.

Goldberg, M. S., N. Giannetti, et al. (2008). "A panel study in congestive heart failure to
estimate the short-term effects from personal factors and environmental conditions on
oxygen saturation and pulse rate."  Occup Environ  Med 65(10): 659-66. This daily diary
panel study in Montreal, Quebec, was carried out to determine whether oxygen saturation and
pulse rate were associated with selected personal factors, weather conditions and air pollution.
Thirty-one subjects with CHF participated in  this study in 2002 and 2003. Over a 2-month
period, the investigators measured their oxygen saturation, pulse rate, weight and temperature
each morning and recorded these and other data in a daily diary. Air pollution and weather
conditions were obtained from fixed-site monitoring stations. The study made use of mixed
regression models, adjusting for within-subject serial correlation and temporal trends,  to
                                          17

-------
determine the association between oxygen saturation and pulse rate and personal and
environmental variables. Depending on the model, we accounted for the effects of a variety of
personal variables (e.g., body temperature, salt consumption) as well as nitrogen dioxide (NO2),
ozone, maximum temperature and change in barometric pressure at 8:00 from the previous day.
RESULTS: In multivariable analyses, the study found that oxygen saturation was reduced when
subjects reported that they were ill, consumed salt, or drank liquids on the previous day and had
higher body temperatures on the concurrent day (only the latter was statistically significant).
Relative humidity and decreased atmospheric pressure from the previous day were associated
with oxygen saturation. In univariate analyses, there was negative associations with
concentrations of fine particulates, ozone, and sulphur dioxide (SO2), but only SO2  was
significant after adjustment for the effects of weather. For pulse rate, no associations were found
for the personal variables and in univariate analyses the  study found positive associations with
NO(2), fine particulates (aerodynamic diameter of 2.5 microm or under, PM(2.5)), SO2, and
maximum temperature, although only the latter two were significant after adjustment for
environmental effects. The findings from the present investigation suggest that personal and
environmental conditions affect intermediate physiological parameters that may affect the health
of CHF patients.

Liao, D., E. A. Whitsel, et al. (2009). "Ambient particulate air pollution and ectopy—the
environmental epidemiology of arrhythmogenesis in  Women's Health Initiative Study,
1999-2004." J Toxicol Environ Health A 72(1): 30-8.  The relationships between ambient
PM(2.5) and PM(10) and arrhythmia and the effect modification by cigarette smoking were
investigated. Data from U.S. Environmental Protection Agency (EPA) air quality monitors and
an established national-scale, log-normal kriging method were used to spatially estimate daily
mean concentrations of PM at addresses of 57,422 individuals from 59 examination sites in 24
U.S. states in 1999-2004. The acute and subacute exposures were estimated as mean, geocoded
address-specific PM concentrations on the day of, 0-2 d  before, and averaged over 30 d before
the electrocardiogram (ECG) (Lag(O); Lag(l); Lag(2); Lag(l-30)). At the time of standard 12-
lead resting ECG, the mean age (SD) of participants was 67.5 (6.9) yr (84% non-Hispanic White;
6% current smoker; 15% with coronary heart disease; 5% with ectopy). After the identification
of significant effect modifiers, two-stage random-effects models were used to calculate center-
pooled odds ratios and 95%  confidence intervals (OR, 95% CI) of arrhythmia per 10 mug/m(3)
increase in PM concentrations. Among current smokers, Lag(O) and Lag(l) PM concentrations
were significantly associated ventricular ectopy (VE)-the OR (95% CI) for VE among current
smokers was 2 (1.32-3.3) and 1.32 (1.07-1.65) at Lag(l) PM(2.5) and PM(10), respectively. The
interactions between current smoking and acute exposures (Lag(O); Lag(l); Lag(2))  were
significant in relationship to VE. Acute exposures were  not significantly associated with
supraventricular ectopy (SVE), or with VE among nonsmokers.  Subacute (Lag(l-30)) exposures
were not significantly associated with arrhythmia. Acute PM(2.5) and PM(10) exposure is
directly associated with the odds of VE among smokers, suggesting that they are more
vulnerable to the arrhythmogenic effects of PM.

Ljungman, P. L., N. Berglind, et al. (2008). "Rapid effects of air pollution on ventricular
arrhythmias." Eur Heart J 29(23): 2894-901. Air pollution has been associated with ventricular
arrhythmias in patients with implantable cardioverter defibrillators  (ICDs) for exposure periods
of 24-48 h. Only two studies have investigated exposure periods <24 h. We aimed to explore
                                           18

-------
such effects during the 2 and 24 preceding hours as well as in relation to distance from the place
of the event to the air pollution monitor. METHODS AND RESULTS: We used a case-crossover
design to investigate the effects of particulate matter <10 microm in diameter (PM10) and
nitrogen dioxide (NO2) in 211 patients with ICD devices in Gothenburg and Stockholm,
Sweden. Events interpreted as ventricular arrhythmias were downloaded from the ICDs, and air
pollution data were collected from urban background monitors. We found an association
between 2 h moving averages of PM10 and ventricular arrhythmia [odds ratio (OR) 1.31, 95%
confidence interval (CI) 1.00-1.72], whereas the OR for 24 h moving averages was 1.24 (95% CI
0.87-1.76). Corresponding ORs for events occurring closest to the air pollution monitor were
1.76 (95% CI 1.18-2.61) and 1.74 (95% CI 1.07-2.84), respectively. Events occurring in
Gothenburg showed stronger associations than in Stockholm.  CONCLUSION: Moderate
increases in air pollution appear to be associated with ventricular arrhythmias in ICD patients
already after 2 h, although future studies including larger numbers of events are required to
confirm these findings. Representative geographical exposure classification seems important in
studies of these effects.

Lucking, A. J., M. Lundback, et al. (2008). "Diesel exhaust inhalation increases thrombus
formation in man." Eur Heart J 29(24): 3043-51.  Although the mechanism is unclear,
exposure to traffic-derived air pollution is a trigger for acute myocardial infarction (MI). The aim
of this study is to investigate the effect of diesel exhaust inhalation on platelet activation and
thrombus formation in men. METHODS AND RESULTS: In a double-blind randomized
crossover study, 20 healthy volunteers were exposed to dilute diesel exhaust (350 microg/m(3))
and filtered air. Thrombus formation, coagulation,  platelet activation, and inflammatory markers
were measured at 2 and 6 h following exposure. Thrombus formation was measured using the
Badimon ex vivo perfusion chamber. Platelet activation was assessed by flow cytometry.
Compared with filtered air, diesel exhaust inhalation increased thrombus formation under low-
and high-shear conditions by 24% [change in thrombus area 2229 microm(2), 95% confidence
interval (CI) 1143-3315 microm(2), P = 0.0002] and 19% (change in thrombus area 2451
microm(2), 95% CI 1190-3712 microm(2), P = 0.0005), respectively. This increased
thrombogenicity was seen at 2 and 6 h, using two different diesel engines and fuels. Diesel
exhaust also increased platelet-neutrophil and platelet-monocyte aggregates by 52% (absolute
change 6%, 95% CI 2-10%, P = 0.01) and 30% (absolute change 3%, 95% CI 0.2-7%, P = 0.03),
respectively, at 2 h following exposure compared with filtered air. CONCLUSION: Inhalation of
diesel exhaust increases ex vivo thrombus formation and causes in vivo platelet activation in
man. These findings provide a potential mechanism linking exposure to combustion-derived air
pollution with the triggering of acute MI.

Pereira Filho, M. A., L. A. Pereira, et al. (2008). "Effect of air pollution on diabetes and
cardiovascular diseases in  Sao Paulo, Brazil." Braz J Med Biol Res 41(6): 526-32. Type 2
diabetes increases the risk of cardiovascular mortality and  these patients, even without previous
myocardial infarction, run the risk of fatal coronary heart disease similar to non-diabetic patients
surviving myocardial infarction. There is evidence showing that particulate matter air pollution is
associated with increases in  cardiopulmonary morbidity and mortality. The present study was
carried out to evaluate the effect of diabetes mellitus on the association of air pollution with
cardiovascular emergency room visits in a tertiary referral  hospital in the city of Sao Paulo.
Using a time-series approach, and adopting generalized linear Poisson regression models, we
                                           19

-------
assessed the effect of daily variations in PM10, CO, NO2, SO2, and O3 on the daily number of
emergency room visits for cardiovascular diseases in diabetic and non-diabetic patients from
2001 to 2003. A semi-parametric smoother (natural spline) was adopted to control long-term
trends, linear term seasonal usage and weather variables. In this period, 45,000 cardiovascular
emergency room visits were registered. The observed increase in interquartile range within the 2-
day moving average of 8.0 microg/m(3) SO2 was associated with 7.0% (95%CI: 4.0-11.0) and
20.0% (95%CI: 5.0-44.0) increases in cardiovascular disease emergency room visits by non-
diabetic and diabetic groups, respectively. These data indicate that air pollution causes an
increase of cardiovascular emergency room visits, and that diabetic patients are extremely
susceptible to the adverse effects of air pollution on their health conditions.

Pope, C. A., 3rd, D. G. Renlund, et al. (2008). "Relation of heart failure hospitalization to
exposure to fine particulate air pollution." Am J Cardiol 102(9):  1230-4. Cardiopulmonary
disease has been associated with particulate matter (PM) air pollution. There is evidence that
exposure to elevated PM concentrations increases risk of acute ischemic heart disease events,
alters cardiac autonomic function, and increases risk of arrhythmias. It is plausible, therefore,
that PM exposure may exacerbate heart failure (HF). A case-crossover study design was used to
explore associations between fine PM (PM(2.5):  particles with an aerodynamic diameter < or
=2.5 microm) and 2,628 HF hospitalizations. Patients lived on Utah's Wasatch Front and were
drawn from those hospitalized at Intermountain Healthcare facilities with a primary diagnosis of
HF. A 14-day lagged cumulative moving average of 10 microg/m(3) PM(2.5) was associated
with a 13.1% (95% confidence interval 1.3 to 26.2) increase in HF admissions. The strongest
PM(2.5)-HF associations were for elderly patients who had previously been admitted for HF and
who required only a short period of hospitalization. HF hospitalizations are associated with
lagged cumulative exposure to PM(2.5) of approximately 2 weeks. In conclusion, particulate air
pollution may play a role in precipitating acute cardiac decompensation in otherwise well-
managed patients with HF, perhaps through effects of PM on myocardial ischemia, cardiac
autonomic function, and/or arrhythmic effects.

Rundell, K. W. and R. Caviston (2008). "Ultrafine and fine particulate matter inhalation
decreases exercise performance in  healthy subjects." J Strength Cond Res 22(1): 2-5. The
purpose of this study was to investigate effects of PM1 (particulate matter with aerodynamic
diameter 0.02-2 microm) inhalation on exercise performance in healthy subjects. Inhalation of
internal combustion-derived PM is associated with adverse effects to the pulmonary and muscle
microcirculation. No data are available concerning air  pollution and exercise performance.
Fifteen healthy college-aged males performed 4 maximal effort 6-min cycle ergometer trials
while breathing low or high PM1 to achieve maximal work accumulation (kJ). Low PM1
inhalation trials 1 and  2 were separated by 3 days; then after a 7 day washout, trials 3 and 4
(separated by 3  days) were done while breathing high PM1 generated from a gasoline engine;
CO was kept below 10 ppm.  Lung function was done after trial 1 to verify nonasthmatic status.
Lung function was normal before and after low PM1 exercise. PM1 number counts were not
different between high PM1 trials (336,730 +/- 149,206 and 396,200 +/- 82,564 for trial 3 and 4,
respectively) and were different from low PM1 trial number counts (2,260 +/- 500) (P < 0.0001).
Mean heart rate was not different between trials (189 +/- 6.0, 188 +/- 7.6, 188 +/- 7.6, 187 +/-
7.4, for low and high PM1 trials;  respectively). Work accumulated was not different between low
PM1 trials (96.1 +/- 9.38 versus  96.6 +/-  10.83 kJ) and the first high PM1 trial  (trial 3, 96.8 +/-
                                           20

-------
10.65 kJ). Work accumulated in the second high PM1 trial 4, 91.3 +/- 10.04 kJ) was less than in
low PM1 trials 1 and 2, and high PM1 trial 3 (P = 0.004, P = 0.003, P = 0.0008; respectively).
Acute inhalation of high (PM1) typical of many urban environments  could impair exercise
performance.

Sint, T., J. F. Donohue, et al. (2008). "Ambient air pollution particles and the acute
exacerbation of chronic obstructive pulmonary disease." Inhal Toxicol 20(1): 25-9.
Investigation has repeatedly demonstrated an association between exposure to ambient air
pollution particles and numerous indices of human morbidity and mortality. Individuals with
chronic obstructive pulmonary disease (COPD) are among those with an increased sensitivity to
air pollution particles. Current and ex-smokers account for 80 to 85% of all those with COPD.
The human breathing in an urban site with a significant level of particulate matter (PM) may be
exposed to 720 microg daily. A single cigarette introduces 15,000 to  40,000 microg particle into
the respiratory tract of the  smoker. It is subsequently confounding why such a relatively small
mass of airborne PM should have any biological effect in the patient  with COPD, as these
individuals are repeatedly exposed to particles (with a similar size and composition) at perhaps a
thousandfold the mass of ambient PM. Regarding this increased sensitivity of COPD patients to
air pollution particles, there are several possible explanations for this seeming contradiction,
including correlations of PM levels with other components of air pollution, an accumulation of
multiple independent risk factors in a patient, changes in individual activity patterns, disparities
in dosimetry between healthy subjects and COPD patients, and some unique characteristic of an
ambient air pollution PM. Regardless of the underlying mechanism for the increased sensitivity
of COPD patients, exposures of these individuals to elevated levels of PM should be
discouraged. To provide a  greater awareness of PM levels, the U.S. Environmental Protection
Agency now includes levels of air pollution particles in an air quality index.

Rhoden, C. R., E. Ghelfi, et al. (2008). "Pulmonary inflammation by ambient air particles
is mediated by superoxide anion."  Inhal Toxicol 20(1): 11-5. Lung inflammation is a key
response to increased levels of paniculate air pollution (PM); however, the cellular mechanisms
leading to this response remain poorly understood. We have previously shown that oxidants are
critical mediators of the inflammatory response elicited by inhalation of ambient air particles.
Here we tested the possible role of a specific oxidant, superoxide anion, by using the membrane-
permeable analog of superoxide dismutase, Mn(III) tetrakis(4-benzoic acid)porphyrin chloride
(MnTBAP). Adult Sprague-Dawley rats were instilled with either urban air particles (UAP) or
saline. MnTBAP-treated rats received 10 mg/kg (ip) MnTBAP 2 h prior to exposure to UAP.
Recruitment of inflammatory cells into bronchoalveolar lavage was evaluated 4 h after
instillation. Rats exposed to UAP showed significant increases in the total cell number (8.9 +/-
0.6 x 10(6); sham: 5.1 +/- 0.6 x 10(6), p < .02), the numbers of polymorphonuclear leukocytes
(26 +/- 4%; sham: 6 +/- 1%, p < .0001), protein levels (1.2 +/- 0.5 mg/ml, sham: 0.4 +/- 0.1
mg/ml, p < .001), and a trend of increase in myeloperoxidase levels (5 +/- 1; sham: 2 +/- 1
mU/ml) in bronchoalveolar lavage (BAL). Pretreatment with MnTBAP at a dose that prevented
UAP-induced increases in  oxidants effectively prevented increase in  BAL cells (2.7 +/-  0.6 x
10(6), p < .0001 vs. UAP), PMN influx  into the lungs (4 +/- 3%, p <  .0001 vs. UAP), and
increase in myeloperoxidase (2 +/- 1 mU/ml) and  protein levels in BAL (0.1 +/- 0.1 mg/ml).
These data indicate that superoxide anion is a critical mediator of the inflammatory response
elicited by PM deposition in the lung.
                                           21

-------
Yatera, K., J. Hsieh, et al. (2008). "Particulate matter air pollution exposure promotes
recruitment of monocytes into atherosclerotic plaques." Am J Physiol Heart Circ Physiol
294(2): H944-53. Epidemiologic studies have shown an association between exposure to
ambient paniculate air pollution <10 microm in diameter (PM(10)) and increased cardiovascular
morbidity and mortality. We previously showed that PM(10) exposure causes progression of
atherosclerosis in coronary arteries. We postulate that the recruitment of monocytes from the
circulation into atherosclerotic lesions is a key step in this PM(10)-induced acceleration of
atherosclerosis. The study objective was to quantify the recruitment of circulating monocytes
into vessel walls  and the progression of atherosclerotic plaques induced by exposure to PM(10).
Female Watanabe heritable hyperlipidemic rabbits, which naturally develop systemic
atherosclerosis, were exposed to PM(10) (EHC-93) or vehicle by intratracheal instillation twice a
week for 4 wk. Monocytes, labeled with 5-bromo-2'-deoxyuridine (BrdU) in donors, were
transfused to recipient rabbits as whole blood, and the recruitment of BrdU-labeled cells into
vessel walls and plaques in recipients was measured by quantitative histological methodology.
Exposure to PM(10) caused progression of atherosclerotic lesions in thoracic and abdominal
aorta. It also decreased circulating monocyte counts, decreased circulating monocytes expressing
high levels of CDS 1 (platelet endothelial cell adhesion molecule-1) and CD49d (very late
antigen-4 alpha-chain), and increased expression of CD54 (ICAM-1) and CD106 (VCAM-1) in
plaques. Exposure to PM(10) increased the number of BrdU-labeled monocytes adherent to
endothelium over plaques and increased the migration of BrdU-labeled monocytes into plaques
and smooth muscle underneath plaques. We conclude that exposure to ambient air pollution
particles promotes the recruitment of circulating monocytes into atherosclerotic plaques and
speculate that this is a critically important step in the PM(10)-induced progression of
atherosclerosis.

Baccarelli, A., P. A. Cassano, et al. (2008). "Cardiac autonomic dysfunction: effects from
particulate air pollution and protection by dietary methyl nutrients and metabolic
polymorphisms." Circulation 117(14): 1802-9.  Particulate air pollution is associated with
cardiovascular mortality and morbidity. To help identify mechanisms of action and
protective/susceptibility factors, we evaluated whether the effect of particulate matter <2.5  mum
in aerodynamic diameter (PM(2.5)) on heart rate variability was modified by dietary intakes of
methyl nutrients (folate, vitamins B(6) and B(12), methionine) and related gene polymorphisms
(C677T methylenetetrahydrofolate reductase [MTHFR] and C1420T cytoplasmic serine
hydroxymethyltransferase [cSHMT]).  METHODS AND RESULTS: Heart rate variability and
dietary data were obtained between 2000 and 2005 from 549 elderly men from the Normative
Aging Study. In carriers of [CT/TT] MTHFR genotypes, the SD of normal-to-normal intervals
was 17.1% (95% CI, 6.5 to 26.4; P=0.002) lower than in CC MTHFR subjects. In the same
[CT/TT] MTHFR subjects, each 10-mug/m(3) increase in PM(2.5) in the 48 hours before the
examination was associated with a further 8.8% (95% CI, 0.2 to 16.7; P=0.047) decrease in the
SDNN. In [CC] cSHMT carriers, PM(2.5) was associated with an 11.8% (95% CI,  1.8  to 20.8;
P=0.02) decrease in SDNN. No PM(2.5)-SSDN association was found in subjects with either
[CC] MTHFR or [CT/TT] cSHMT genotypes. The negative effects of PM(2.5) were abrogated in
subjects with higher intakes (above median levels) of B(6), B(12), or methionine. PM(2.5) was
negatively associated with heart rate variability in subjects with lower intakes, but no PM(2.5)
effect was found  in the higher intake groups. CONCLUSIONS: Genetic and nutritional
                                          22

-------
variations in the methionine cycle affect heart rate variability either independently or by
modifying the effects of PM2.5.

Gallus, S., E. Negri, et al. (2008). "European studies on long-term exposure to ambient
particulate matter and lung cancer." Eur J Cancer Prev 17(3): 191-4. European
epidemiological studies on ambient air pollution and cancer published before December 2006 are
reviewed, with focus on five analytic studies providing data on the association between various
measures of particulate matter (PM) and lung cancer. A case-control study of 755 men who died
from lung cancer in Trieste, Italy, reported that, compared with less than 0.18 g/m/day of
deposition of particulate, the relative risk (RR) was  1.1 [95% confidence interval (CI): 0.8-1.5]
for 0.18-0.30 and 1.4 (95% CI: 1.1-1.8) for more than 0.30 g/m/day. In the Netherlands Cohort
Study on Diet and Cancer with 60 deaths from lung cancer, the RR was 1.06 (95% CI: 0.43-2.63)
for an increase of 10 mug/m in black smoke. In the French Pollution Atmospherique et
Affections Respiratoires Chroniques study cohort based on 178 deaths from lung cancer, the RR
associated with an increase in exposure to 10 mug/m of total suspended particulate was 0.97
(95% CI: 0.94-1.01). A nested case-control study within the European Prospective Investigation
on Cancer and Nutrition included 113 nonsmokers or exsmokers diagnosed with lung cancer and
312 controls. The RRs were 0.91 (95% CI: 0.70-1.18) for an increase in PM with diameter 
-------
PM(2.5) was taken into account. We did, however, demonstrate a significant increase in the risk
of death from respiratory causes in association with an increase in ozone concentration.

Pope, C. A., 3rd, M. Ezzati, et al. (2009). "Fine-particulate air pollution and life expectancy
in the United States." N Engl J Med 360(4): 376-86. Exposure to fme-particulate air pollution
has been associated with increased morbidity and mortality, suggesting that sustained reductions
in pollution exposure should result in improved life expectancy. This study directly evaluated the
changes in life expectancy associated with differential changes in fine particulate air pollution
that occurred in the United States during the 1980s and 1990s. METHODS: We compiled data
on life expectancy, socioeconomic status, and demographic characteristics for 211 county units
in the 51 U.S. metropolitan areas with matching data on fme-particulate air pollution for the late
1970s and early 1980s and the late 1990s and early 2000s. Regression models were used to
estimate the association between reductions in pollution and changes in life expectancy, with
adjustment for changes in socioeconomic and demographic variables and in proxy indicators for
the  prevalence of cigarette smoking. RESULTS: A decrease of 10 microg per cubic meter in the
concentration of fine particulate matter was associated with an estimated increase in mean (+/-
SE) life expectancy of 0.61+/-0.20 year (P=0.004). The estimated effect of reduced exposure to
pollution on life expectancy was not highly sensitive to adjustment for changes in
socioeconomic, demographic, or proxy variables for the prevalence of smoking or to the
restriction of observations to relatively large counties. Reductions in air pollution accounted for
as much as 15% of the overall increase in life expectancy in the study areas. CONCLUSIONS: A
reduction in exposure to ambient fme-particulate air pollution contributed to significant and
measurable improvements in life expectancy in the United  States.

Kamdar, O., W. Le, et al. (2008). "Air pollution induces enhanced mitochondrial oxidative
stress in cystic fibrosis airway epithelium." FEES Lett 582(25-26): 3601-6. We studied the
effects of airborne particulate matters (PM) on cystic fibrosis (CF) epithelium. We noted that PM
enhanced human CF bronchial epithelial apoptosis, activated caspase-9 and PARP-1;  and
reduced mitochondrial membrane potential. Mitochondrial  inhibitors (4,4-
diisothiocyanatostilbene-2,2'disulfonic acid, rotenone and thenoyltrifluoroacetone) blocked PM-
induced generation of reactive oxygen species and apoptosis. PM upregulated pro-apoptotic Bad,
Bax, p53 and p21; and enhanced mitochondrial localization of Bax. The anti-apoptotic Bel-2,
Bcl-xl, Mcl-1 and Xiap remained unchanged; however, overexpression of Bcl-xl blocked PM-
induced apoptosis.  Accordingly, we provide the evidence that PM enhances oxidative stress and
mitochondrial signaling mediated apoptosis via the modulation of Bel family proteins in CF.

Nadadur, S. S., N. Haykal-Coates, et al. (2009). "Endothelial effects of emission source
particles: acute toxic response gene expression profiles." Toxicol In Vitro 23(1): 67-77. Air
pollution epidemiology has established a strong association between exposure to ambient
particulate matter (PM) and cardiovascular outcomes. Experimental studies in both humans and
laboratory animals  support varied biological mechanisms including endothelial dysfunction as
potentially a central step to the elicitation of cardiovascular events. We therefore hypothesized
that relevant early molecular alterations on endothelial cells should be assessable in vitro upon
acute exposure to PM components previously shown to be involved in health outcomes. Using a
model emission PM, residual oil fly ash and one of its predominant constituents (vanadium-V),
we  focused on the development of gene expression profiles to fingerprint that particle and its
                                           24

-------
constituents to explore potential biomarkers for PM-induced endothelial dysfunction. Here we
present differential gene expression and transcription factor activation profiles in human vascular
endothelial cells exposed to a non-cytotoxic dose of fly ash or V following semi-global gene
expression profiling of approximately 8000 genes. Both fly ash and it's prime constituent, V,
induced alterations in genes involved in passive and active transport of solutes across the
membrane; voltage-dependent ion pumps; induction of extracellular matrix proteins and
adhesion molecules; and activation of numerous kinases involved in signal transduction
pathways. These preliminary data suggest that cardiovascular effects associated with exposure to
PM may be mediated by perturbations in endothelial cell permeability, membrane integrity; and
ultimately endothelial dysfunction.

Sun, Q., P. Yue, et al. (2008). "Air pollution exposure potentiates hypertension through
reactive oxygen species-mediated activation of Rho/ROCK." Arterioscler Thromb Vase
Biol 28(10): 1760-6. OBJECTIVE: Fine particulate matter <2.5 microm (PM(2.5)) has been
implicated in vasoconstriction and potentiation of hypertension in humans. We investigated the
effects of short-term exposure to PM(2.5) in the angiotensin II (All) infusion model. METHODS
AND RESULTS:  Sprague-Dawley rats were exposed to PM(2.5) or filtered air (FA) for 10
weeks. At week 9, minipumps containing All were implanted and the responses studied over a
week. Mean concentration of PM(2.5) inside the chamber was 19.1+1-1'.4 microg/m(3). After All
infusion, mean arterial pressure was significantly higher in PM(2.5)-AII versus FA-All group.
Aortic vasoconstriction to phenylephrine was potentiated with exaggerated relaxation to the Rho-
kinase (ROCK) inhibitor Y-27632 and increase in ROCK-1 mRNA levels in the PM(2.5)-AII
group. Superoxide (O(2).(-)) production in aorta was increased in the PM(2.5)-AII compared to
the FA group, inhibitable by apocynin and L-NAME with coordinate upregulation of NAD(P)H
oxidase  subunits p22(phox) and p47(phox) and depletion of tetrahydrobiopterin. In vitro
exposure to ultrafine particles (UFP) and PM(2.5) was associated with an increase in ROCK
activity, phosphorylation of myosin light chain, and myosin phosphatase target subunit
(MYPT1). Pretreatment with the nonspecific antioxidant N-acetylcysteine and the Rho kinase
inhibitors (Fasudil and Y-27632) prevented MLC and MYPT-1 phosphorylation by UFP
suggesting a O(2)(.-)-mediated mechanism for PM(2.5) and UFP effects. CONCLUSIONS:
Short-term air pollution exaggerates hypertension through O(2)(.-)-mediated upregulation of the
Rho/ROCK pathway.

Min, K. B., J. Y. Min, et al. (2008). "The relationship between air pollutants and heart-rate
variability among community residents in Korea." Inhal Toxicol 20(4): 435-44. Air
pollution, both particulate and gaseous, is known to cause adverse health effects and is associated
with increased cardiovascular mortality and morbidity. With a growing recognition in the
importance of the  autonomic nervous system in air pollution, we examined the effects of air
pollutants, namely, particulate matter (PM10), sulfur dioxide (SO2), and nitric dioxide (NO2), on
cardiac autonomic function by measuring heart-rate variability (HRV) among community
residents. This study was conducted at Taein Island, located off the southern coast of South
Korea; 1349 subjects (596 males and 753 females) were included in this analysis.  Subjects
responded to the interview about general characteristics and an HRV examination was
conducted. Exposure data were collected from the Environmental Management Corporation
during the same period of HRV measurement. Linear regression analyses were carried out to
evaluate the association over 72 h, and the parameters of HRV indices were presented as the
                                          25

-------
percentage change. The exposures to PM(10), SO(2), and NO2 were associated with reduced
HRV indices, and significant decreases in the standard deviation of the normal to normal interval
(SDNN) and low frequency (LF) domain effect, and the effect was largely continued until 12 h.
Our results suggest that air pollutants stimulate the autonomic nervous system and provoke an
imbalance in cardiac autonomic control. Thus, these subclinical effects may lead to pathological
consequences, particularly in high-risk patients and susceptible subjects.

Stafoggia, M., J. Schwartz, et al. (2008). "Does temperature modify the association between
air pollution and mortality? A multicity case-crossover analysis in Italy." Am J Epidemiol
167(12): 1476-85.  Adverse health effects of particulate matter <10 microm in aerodynamic
diameter (PM(10)) and high temperatures are well known, but the extent of their interaction on
mortality is less clear.  This paper describes effect modification of temperature in the PM(10)-
mortality association and tests the hypothesis that higher PM(10) effects in summer are due to
enhanced exposure to particles. All deaths of residents of nine Italian cities between 1997 and
2004 were selected. The case-crossover approach was adopted to estimate the effect of PM(10)
on mortality by season and temperature level. Three strata of temperature corresponding to low,
medium, and high "ventilation" were identified, and the interaction between PM(10) and
temperature within each stratum was examined. Season and temperature levels strongly  modified
the PM(10)-mortality association: for a 10-microg/m(3) variation in PM(10), a 2.54% increase in
risk of death in summer (95% confidence interval: 1.31, 3.78) compared with 0.20% (95%
confidence interval: -0.08, 0.49) in winter. Analysis  of the interaction between PM(10) and
temperature within temperature strata resulted in positive but, in most cases, nonstatistically
significant coefficients. The authors found much higher PM(10) effects on mortality during
warmer days. The hypothesis that such an effect is attributable to enhanced exposure to particles
in summer could not be rejected.

Tamagawa, E., N. Bai, et al. (2008). "Particulate matter exposure induces persistent lung
inflammation and endothelial dysfunction." Am J Physiol Lung Cell Mol Physiol 295(1):
L79-85. Epidemiologic and animal studies have shown that exposure to particulate matter air
pollution (PM) is a risk factor for the development of atherosclerosis. Whether PM-induced lung
and systemic inflammation is involved in this process is not clear. We hypothesized that PM
exposure causes lung and systemic inflammation, which in turn leads to vascular endothelial
dysfunction, a key step in the initiation and progression of atherosclerosis. New Zealand White
rabbits were exposed for 5 days (acute, total dose 8 mg) and 4 wk (chronic, total dose 16 mg) to
either PM smaller than 10 mum (PM(10)) or saline intratracheally. Lung inflammation was
quantified by morphometry; systemic inflammation was assessed by white blood cell and platelet
counts and serum interleukin (IL)-6, nitric oxide, and endothelin levels. Endothelial dysfunction
was assessed by vascular response to acetylcholine (ACh) and sodium nitroprusside (SNP).
PM(10) exposure increased lung macrophages (P<0.02), macrophages containing particles
(P<0.001), and activated macrophages (P<0.006). PM(10) increased serum IL-6 levels in the first
2 wk of exposure (P<0.05) but not in weeks 3 or 4. PM(10) exposure reduced ACh-related
relaxation of the carotid artery with both acute and chronic exposure, with no effect on SNP-
induced vasodilatation. Serum IL-6 levels correlated with macrophages containing particles
(P=0.043) and ACh-induced vasodilatation (P=0.014 at week 1, P=0.021 at week 2). Exposure to
PM(10) caused lung and systemic inflammation that were both associated with vascular
endothelial dysfunction.  This suggests that PM-induced lung and systemic inflammatory
                                           26

-------
responses contribute to the adverse vascular events associated with exposure to air pollution.

Yatera, K., J. Hsieh, et al. (2008). "Particulate matter air pollution exposure promotes
recruitment of monocytes into atherosclerotic plaques." Am J Physiol Heart Circ Physiol
294(2): H944-53. Epidemiologic studies have shown an association between exposure to
ambient particulate air pollution <10 microm in diameter (PM(10)) and increased cardiovascular
morbidity and mortality. We previously showed that PM(10) exposure causes progression of
atherosclerosis in coronary arteries. We postulate that the recruitment of monocytes from the
circulation into atherosclerotic lesions is a key step in this PM(10)-induced acceleration of
atherosclerosis. The study objective was to quantify the recruitment of circulating monocytes
into vessel walls and the progression of atherosclerotic plaques induced by exposure to PM(10).
Female Watanabe heritable hyperlipidemic rabbits, which naturally develop systemic
atherosclerosis, were exposed to PM(10) (EHC-93) or vehicle by intratracheal instillation twice a
week for 4 wk. Monocytes, labeled with 5-bromo-2'-deoxyuridine (BrdU) in donors, were
transfused to recipient rabbits as whole blood, and the recruitment of BrdU-labeled cells into
vessel walls and plaques in recipients was measured by quantitative histological methodology.
Exposure to PM(10) caused progression of atherosclerotic lesions in thoracic and abdominal
aorta. It also decreased circulating monocyte counts, decreased circulating monocytes expressing
high levels of CDS 1 (platelet endothelial cell adhesion molecule-1) and CD49d (very late
antigen-4 alpha-chain), and increased expression of CD54 (ICAM-1) and CD106 (VCAM-1) in
plaques. Exposure to PM(10) increased  the number of BrdU-labeled monocytes adherent to
endothelium over plaques and increased the migration of BrdU-labeled monocytes into plaques
and smooth muscle underneath plaques. We conclude that exposure to ambient air pollution
particles promotes the recruitment of circulating monocytes into atherosclerotic plaques and
speculate that this is a critically important step in the PM(10)-induced progression of
atherosclerosis.

Totlandsdal, A. L, M. Refsnes, et al. (2008). "Particle-induced cytokine responses in cardiac
cell cultures—the effect of particles versus soluble mediators released by particle-exposed
lung cells." Toxicol  Sci 106(1): 233-41. Increased levels of parti culate matter have been
associated with adverse effects in the respiratory as well as the cardiovascular system.  The
biological mechanisms behind these associations are still unresolved. Among potential
mechanisms,  particulate matter-associated cardiac effects may be initiated by inhaled small-sized
particles, particle components and/or mediators related to inflammation that translocate into the
pulmonary  circulation. In the present study cytokine responses (interleukin [ILJ-6, IL-lbeta, and
tumor necrosis factor [TNF]-alpha) of primary rat cardiomyocytes and cardiofibroblasts in
mono- and  cocultures induced by direct exposure to particles, were compared with cytokine
responses induced by mediators released by particle-exposed primary rat epithelial lung cells
(conditioned media). Cells were exposed to a model ultrafme particle (ultrafme carbon black,
Printex 90) and in selected experiments  to an urban air particle sample (SRM 1648, St Louis,
MO). In lung cell cultures both particle types induced release of IL-6 and IL-lbeta, whereas
TNF-alpha was only  detected upon exposure to St Louis particles. The release of IL-6 by cardiac
cells was strongly enhanced upon exposure to conditioned media, and markedly exceeded the
response to direct particle exposure. IL-1, but not TNF-alpha, seemed necessary, but not
sufficient, for this enhanced IL-6 release. The role of IL-1 was demonstrated by use of an IL-1
receptor antagonist that partially reduced the effect of the conditioned media, and by a
                                           27

-------
stimulating effect on the cardiac cell release of IL-6 by exogenous addition of IL-1 alpha and IL-
Ibeta. These in vitro findings lend support to the hypothesis that particle-induced cardiac
inflammation and disease may involve lung-derived mediators.

Totlandsdal, A. L, T. Skomedal, et al. (2008). "Pro-inflammatory potential of ultrafine
particles in mono- and co-cultures of primary cardiac cells." Toxicology 247(1): 23-32.
Inhalation of particulate air pollution has been associated with increased risks for cardiovascular
mortality and morbidity, but the underlying mechanisms are still under discussion. One possible
pathway may be that inhaled particles cross the air-blood barrier and interact directly with
cardiac tissue. The aim of the present study was to examine the pro-inflammatory potential of
particles in cardiac cells. Mono- and co-cultures of primary adult male Wistar (Han) rat
cardiomyocytes (CMs) and cardiofibroblasts (CFs) were exposed to increasing concentrations of
ultrafine (<100nm) carbon black particles (Printex 90). Expression and release of cytokines (IL-
6, IL-lbeta and TNF-alpha) were measured by using quantitative real-time PCR and ELISA,
respectively. Cytotoxicity was estimated by measuring cellular release of lactate dehydrogenase
(LDH). A particle concentration-dependent increase in IL-6 release was observed in both CM
mono- and co-cultures (EC(50) approximately 57microg/ml).  Furthermore, IL-6 levels detected
in both control and particle-exposed co-cultures were synergistically increased compared to
mono-cultures (10-19-fold, dependent on the exposure). Experiments with contact and non-
contact co-cultures indicate that direct cellular contact is of key  importance for the enhanced
release of IL-6 in co-cultures. An apparent particle-induced release of IL-lbeta was only detected
in co-cultures. The release of TNF-alpha was low and did not seem notably influenced by
particle exposure. Treatment with an IL-1 receptor antagonist apparently eliminated the particle-
induced release of IL-6. In conclusion, ultrafine particles  have a pro-inflammatory potential in
primary cardiac cells. Furthermore, IL-1 seems critical in triggering particle-induced release of
IL-6. These pro-inflammatory responses may be elicited when particles are translocated into the
pulmonary circulation upon inhalation or administered intravascularly during medical
procedures.

Lee, S. J., S. Hajat, et al. (2008). "A time-series analysis of any short-term effects of
meteorological and air pollution factors on preterm births in London, UK." Environ Res
106(2):  185-94. Although much is known about the incidence and burden of preterm birth, its
biological mechanisms are not well understood. While several studies have suggested that high
levels of air pollution or exposure to particular climatic factors may be associated with an
increased risk of preterm birth, other studies do not support such an association.  To determine
whether exposure to various environmental factors place  a large London-based population at
higher risk for preterm birth, we analyzed 482,568 births  that  occurred between 1988 and 2000
from the St. Mary's Maternity Information System database. Using an ecological study design,
any short-term associations between preterm birth and various environmental factors were
investigated using time-series regression techniques. Environmental exposures included air
pollution (ambient ozone and PM(10)) and climatic factors (temperature, rainfall, sunshine,
relative humidity, barometric pressure, and largest drop in barometric pressure).  In addition to
exposure on the day of birth, cumulative exposure up to 1 week before birth was investigated.
The risk of preterm birth did not increase with exposure to the levels of ambient air pollution or
meteorological factors experienced by this population.  Cumulative exposure  from 0 to 6 days
before birth also did not show any significant effect on the risk of preterm birth.  This large study,
                                           28

-------
covering 13 years, suggests that there is no association between preterm births and recent
exposure to ambient air pollution or recent changes in the weather.

Rankin, J., T. Chadwick, et al. (2009). "Maternal exposure to ambient air pollutants and
risk of congenital anomalies." Environ Res 109(2): 181-7.
Studies have suggested an association between maternal exposure to ambient air pollution and
risk of congenital anomaly. The aim of this study is to investigate the association between
exposure to black smoke (BS; particulate matter with aerodynamic diameter <4 microg/m(3))
and sulphur dioxide (SO(2)) during the first trimester of pregnancy and risk of congenital
anomalies. We used a case-control study design among deliveries to mothers resident in the UK
Northern health region during 1985-1990. Case data were ascertained from the population-based
Northern Congenital Abnormality Survey and control data from national data on all births. Data
on BS and SO(2) from ambient air monitoring stations were used to average the total pollutant
exposure during the first trimester of pregnancy over the daily readings from all monitors within
10 km of the mother's residence. Logistic regression models estimated the association via odds
ratios. A significant but weak positive association was found between nervous system anomalies
and BS (OR=1.10 per increase of 1000 microg/m(3) total BS; 95% CI:  1.03, 1.18), but not with
other anomaly subtypes. For SO(2), a significant negative association was found with congenital
heart disease combined and patent ductus arteriosus: OR significantly <1  for all quartiles relative
to the first quartile. The relationship between SO(2) levels and other anomaly subtypes was less
clear cut: there were either no significant associations or a suggestion of a U-shaped relationship
(OR significantly <1 for moderate compared to lowest levels, but not with high SO(2) levels).
Overall, maternal exposure to BS and SO(2) in the Northern region had limited impact on
congenital anomaly risk. Studies with detailed exposure assessment are needed to further
investigate this relationship.

Son, J. Y., Y. S. Cho, et al. (2008). "Effects of air pollution on postneonatal infant mortality
among firstborn infants in Seoul, Korea: case-crossover and time-series analyses." Arch
Environ Occup Health 63(3): 108-13. Infants are known to be susceptible to the adverse health
effects of ambient air pollution. The authors examined the relationship between air pollution and
postneonatal mortality from all causes among firstborn infants in Seoul, Korea, during 1999-
2003, using both case-crossover and time-series analyses. Using a bidirectional control-sampling
approach, the authors compared the effects of various types of air pollution on postneonatal
mortality. The relative risk of postneonatal mortality from all causes was  1.000 (95% confidence
interval  [CI] = 0.998-1.002) for particulate matter with a diameter <10 mum, 1.002 (95% CI =
0.994-1.009) for nitrogen dioxide, 1.015 (95% CI = 0.973-1.058) for sulfur dioxide, 1.029 (95%
CI = 0.833-1.271) for carbon monoxide, and 0.984 (95% CI = 0.977-0.992) for ozone for each 1-
unit increase of air pollution level in the 1:6 control selection scheme. The authors observed a
positive association between air pollution and infant daily mortality except for the studied
particulate matter and ozone, although it was not statistically significant. They obtained similar
results in the time-series analysis. The risk of postneonatal infant death from all causes was
positively associated with all studied air pollutants except ozone. The authors also confirmed that
the bidirectional method with many controls will give a more efficient estimator than will a
method with fewer controls.

Epton, M. J., R. D. Dawson, et al. (2008). "The effect of ambient air pollution on
                                           29

-------
respiratory health of school children: a panel study." Environ Health 7: 16. Adverse
respiratory effects of particulate air pollution have been identified by epidemiological studies.
We aimed to examine the health effects of ambient parti culate air pollution from wood burning
on school-age students in Christchurch, New Zealand, and to explore the utility of urine and
exhaled breath condensate biomarkers of exposure in this population. METHODS: A panel study
of 93 male students (26 with asthma) living in the boarding house of a metropolitan school was
undertaken in the winter  of 2004. Indoor and outdoor pollution data was continuously monitored.
Longitudinal assessment of lung function (FEV1 and peak flow) and symptoms were undertaken,
with event studies of high pollution on biomarkers of exposure (urinary 1-hydroxypyrene) and
effect (exhaled breath condensate (EEC) pH and hydrogen peroxide concentration). RESULTS:
Peak levels of air pollution were associated with small but statistically significant effects on lung
function in the asthmatic students, but not healthy students. No significant effect of pollution
could be seen either on airway inflammation  and oxidative stress either in healthy students or
students with asthma. Minor increases in respiratory symptoms were associated with high
pollution exposure. Urinary 1-hydroxypyrene levels were raised in association with pollution
events by comparison with low pollution control days. CONCLUSION: There is no significant
effect of ambient wood-smoke parti culate air pollution on lung function of healthy school-aged
students, but a small effect on respiratory symptoms. Asthmatic students show small effects of
peak pollution levels on lung function. Urinary 1-hydroxypyrene shows potential as a biomarker
of exposure to wood smoke in this population; however measurement of EEC pH and hydrogen
peroxide appears not to be useful for assessment of population health effects of air pollution.
Some of the data presented in this paper has previously been published in Kingham and co-
workers Atmospheric Environment, 2006 Jan; 40: 338-347 (details of pollution exposure), and
Cavanagh and co-workers Sci Total Environ. 2007 Mar l;374(l):51-9 (urine hydroxypyrene
data).

Ma, L., M. Shima, et al. (2008). "Effects of airborne particulate matter on respiratory
morbidity in asthmatic  children." J Epidemiol 18(3): 97-110.  The effects of airborne
particulate matter (PM) are a major human health concern. In this panel study, we evaluated the
acute effects of exposure to PM on peak expiratory flow (PEF) and wheezing in children.
METHODS: Daily PEF and wheezing were examined in  19 asthmatic children who were
hospitalized in a suburban city in Japan for approximately 5 months. The concentrations of PM
less than 2.5 mum in diameter (PM(2.5)) were monitored at a monitoring station proximal to the
hospital. Moreover, PM(2.5) concentrations inside and outside the hospital were measured using
the dust monitor with a laser diode (PM(2.5(LD))). The changes in PEF and wheezing associated
with PM concentration were analyzed. RESULTS: The changes in PEF in the morning and
evening were significantly associated with increases in the average concentration of indoor
PM(2.5(LD)) 24 h prior to measurement (-2.86 L/min [95%CI: -4.12, -1.61] and -3.59 L/min
[95%CI: -4.99, -2.20] respectively, for 10-mug/m(3) increases). The change in PEF was also
significantly associated with outdoor PM(2.5(LD)) concentrations, but the changes were smaller
than those observed for indoor PM(2.5(LD)). Changes in PEF and concentration of stationary-
site PM(2.5) were not associated. The prevalence of wheezing in the morning and evening were
also significantly associated with indoor PM(2.5(LD)) concentrations (odds ratios = 1.014
[95%CI: 1.006, 1.023] and 1.025 [95%CI: 1.013, 1.038] respectively, for 10-microg/m(3)
increases). Wheezing in the evening was significantly associated  with outdoor PM(2.5(LD))
concentration. The effects of indoor and outdoor PM(2.5(LD)) remained significant even after
                                          30

-------
adjusting for ambient nitrogen dioxide concentrations. CONCLUSION: Indoor and outdoor
PM(2.5(LD)) concentrations were associated with PEF and wheezing among asthmatic children.
Indoor PM(2.5(LD)) had a more marked effect than outdoor PM(2.5(LD)) or stationary-site
PM(2.5).

O'Connor, G. T., L. Neas, et al. (2008). "Acute respiratory health effects of air pollution on
children with asthma in US inner cities." J Allergy Clin Immunol 121(5): 1133-1139 el.
Children with asthma in inner-city communities may be particularly vulnerable to adverse effects
of air pollution because of their airways disease and exposure to relatively high levels of motor
vehicle emissions. OBJECTIVE: To investigate the association between fluctuations in outdoor
air pollution and asthma morbidity among inner-city children with asthma. METHODS: We
analyzed data from 861 children with persistent asthma in 7 US urban communities who
performed 2-week periods of twice-daily pulmonary function testing every 6 months for 2 years.
Asthma symptom data were collected every 2 months. Daily  pollution measurements were
obtained from the Aerometric Information Retrieval  System. The relationship of lung function
and symptoms to fluctuations in pollutant concentrations was examined by using mixed models.
RESULTS: Almost all pollutant concentrations measured were below the National Ambient Air
Quality Standards. In single-pollutant models, higher 5-day average concentrations of NO2,
sulfur dioxide, and particles smaller than 2.5  microm were associated with significantly lower
pulmonary function. Higher pollutant levels were independently associated with reduced lung
function in a 3-pollutant model. Higher concentrations of NO2 and particles smaller than 2.5
microm were associated with asthma-related  missed school days, and higher NO2 concentrations
were associated with asthma symptoms. CONCLUSION: Among inner-city children with
asthma, short-term increases in air pollutant concentrations below the National Ambient Air
Quality Standards were associated with adverse respiratory health effects. The associations with
NO2 suggest that motor vehicle emissions may be causing excess morbidity in this population.

Odajima, H., S. Yamazaki, et al. (2008). "Decline in peak expiratory flow according to
hourly short-term concentration of particulate matter in asthmatic children." Inhal
Toxicol 20(14): 1263-72. The aim was to investigate the effects of the 3-h mean concentration of
suspended particulate matter (SPM) on peak expiratory flow (PEF) in asthmatic children.
Subjects were 4- to 11-year-old boys and girls with asthma in Fukuoka, Japan. Daily
measurements of PEF were made between April 2002 and March 2003. PEF was measured
morning and evening. To assess the association between 3-h mean concentrations of SPM and
morning PEF and evening PEF separately, we used generalized estimating equations. All models
took into consideration seasonal effects: modified effects were examined using a two-level
indicator variable for the warmer months (April through September) and the colder months
(October through March). Among the 70 asthmatic children,  44 were boys. In warmer months,
declines in morning PEF for 10 microg/m(3)  differences in 3-h concentration of SPM measured
at 2 a.m. to 5 a.m. of the same day and 11 p.m. to 2 a.m. and  8 p.m. to 11 p.m. of the previous
day were -0.78 L/min (95% CI: -1.40, -0.16), -0.61 L/min (-1.18, -0.05) and -0.73 L/min (-1.32, -
0.15), respectively. Results were robust even after adjustment for other air pollutants. We also
found that declines in evening PEF were weakly associated with increases in 3-h concentrations
of SPM in warmer months, but the associations were not statistically significant. In colder
months we did not find any robust association between SPM and morning/evening PEF. In
conclusion, an increased 3-h concentration of SPM is associated with declines in PEF in warmer
                                          31

-------
months.

Tecer, L. H., O. Alagha, et al. (2008). "Particulate matter (PM(2.5), PM(10-2.5), and
PM(10)) and children's hospital admissions for asthma and respiratory diseases: a
bidirectional case-crossover study." J Toxicol Environ Health A 71(8): 512-20.
Epidemiological studies reported adverse effects of air pollution on the prevalence of respiratory
diseases in children. The purpose of this study was to examine the association between air
pollution and admissions for asthma and other respiratory diseases among children who were
younger than 15 yr of age. The study used data on respiratory hospital admissions and air
pollutant concentrations, including thoracic particulate matter (PM(10)), fine (PM(2.5)), and
coarse (PM(10-2.5)) particulate matter in Zonguldak, Turkey. A bidirectional case-crossover
design was used to calculate odds ratios for the admissions adjusted for daily meteorological
parameters. Significant increases were observed for hospital admissions in children for asthma,
allergic rhinitis (AR), and upper (UPRD) and lower (LWRD) respiratory diseases. All fraction of
PM in children showed significant positive associations with asthma admissions. The highest
association noted was 18% rise in asthma admissions correlated with a 10-microg/m(3) increase
in PM(10-2.5) on the same day of admissions. The adjusted odds ratios for exposure to PM(2.5)
with an increment of 10 microg/m(3) were 1.15 and 1.21  for asthma and allergic rhinitis with
asthma, respectively. PM(10) exerted significant effects on hospital admissions for all outcomes,
including asthma,  AR, UPRD, and LWRD. Our study suggested a greater effect of fine and
coarse PM on asthma hospital  admissions compared with PM(10) in children.

Van Roosbroeck, S., R. Li, et al. (2008). "Traffic-related outdoor air pollution and
respiratory  symptoms in children:  the impact of adjustment for exposure measurement
error." Epidemiology 19(3):  409-16. Outdoor concentrations of soot and nitrogen dioxide
(NO2) outside of schools have been associated with children's respiratory and eye symptoms.
We assessed how adjustments  for measurement error affect these associations. METHODS:
Concentrations of air pollutants outside children's schools were validated by personal
measurements of exposure to traffic-related air pollution. We estimated prevalence ratios of 4
health outcomes (current wheeze, conjunctivitis, phlegm, and elevated total serum
immunoglobulin E) using school outdoor measurements,  and then adjusted for measurement
error using the personal exposure data and applying a regression calibration method. The
analysis adjusting for measurement error was carried out using a main study/external validation
design. RESULTS: Adjusting for measurement error produced effect estimates related to soot
and NO2 that were 2 to 3 times higher than in the original study.  The adjusted prevalence ratio
for current phlegm was 5.3 (95% confidence interval = 1.2-23) for a 9.3 microg/m3 increase in
soot, and 3.8 (1.0-14), for a 17.6 microg/m3 increase in NO2, compared with the original results
of 2.2 (1.3-3.9) and 1.8 (1.1-2.8), respectively.  Corrections were of similar magnitude for the
prevalence of current wheeze,  current conjunctivitis, and total elevated total immunoglobulin E.
CONCLUSIONS: The estimated effects of outdoor air pollution on respiratory and other health
effects in children may be substantially attenuated when based on exposure measurements
outside schools instead of personal exposure.

Ballester, F., S. Medina, et al. (2008). "Reducing ambient levels of fine particulates could
substantially improve health: a mortality impact assessment for 26 European cities." J
Epidemiol Community Health 62(2): 98-105. Recently  new European policies on ambient air
                                           32

-------
quality—namely, the adoption of new standards for fine particulate matter (PM(2.5)), have
generated a broad debate about choosing the air quality standards that can best protect public
health. The Apheis network estimated the number of potential premature deaths from all causes
that could be prevented by reducing PM(2.5) annual levels to 25 microg/m3, 20 microg/m3, 15
microg/m3 and 10 microg/m3 in 26 European cities. The various PM(2.5) concentrations were
chosen as different reductions based on the limit values proposed by the new European
Directive, the European Parliament, the US Environmental Protection Agency and the World
Health Organization, respectively. The Apheis network provided the health and exposure data
used in this study. The concentration-response function (CRF) was derived from the paper by
Pope et al (2002). If no direct PM(2.5  )measurements were available, then the PM(10)
measurements were converted to PM(2.5 )using a local or an assumed European conversion
factor. We performed a sensitivity analysis using assumptions for two key factors—namely, CRF
and the conversion factor for PM(2.5). Specifically, using the "at least" approach, in the 26
Apheis cities with more than 40 million inhabitants, reducing annual mean levels of PM(2.5) to
15 microg/m3  could lead to a reduction in the total burden of mortality among people aged 30
years and over that would be four times greater than the reduction in mortality that could be
achieved by reducing PM(2.5) levels to 25 microg/m3 (1.6% vs 0.4% reduction) and two times
greater than  a reduction to 20 microg/m3. The percentage reduction could grow by more than
seven times if PM(2.5) levels were reduced to 10 microg/m3 (3.0% vs 0.4%). This study shows
that more stringent standards need to be adopted in Europe to protect public health, as proposed
by the scientific community and the World Health Organization.

Lee, S. J., S. Hajat, et al. (2008). "A  time-series analysis of any short-term effects of
meteorological and air pollution factors on  preterm births in London, UK." Environ Res
106(2): 185-94.) In this case 482,568 births were analyzed in a London based study. The births
occurred between 1988 and 2000 and  data was obtained from the St. Mary's Maternity
Information System database. Short-term associations between preterm birth and various
environmental factors were  assess with time-series regression techniques. Environmental
exposures included air pollution (ambient ozone and PM10)  and climatic factors (temperature,
rainfall, sunshine, relative humidity, barometric pressure, and largest drop in barometric
pressure). In addition to exposure on the day of birth, cumulative exposure up to 1 week before
birth was investigated. The risk of preterm birth did not increase with exposure to the levels of
ambient air pollution or meteorological factors experienced by this population. Cumulative
exposure from 0 to 6 days before birth also did not show any significant effect on the risk of
preterm birth. This large study, covering 13 years, suggests that there is no association between
preterm births and recent exposure to ambient air pollution or recent changes in the weather.

Rankin,  J., T. Chadwick, et al. (2009). "Maternal exposure to ambient air pollutants and
risk of congenital anomalies." Environ Res 109(2): 181-7.
Studies have suggested an association between maternal exposure to ambient air pollution and
risk of congenital  anomaly.  The aim of this study is to investigate the association between
exposure to black smoke (BS; particulate matter with aerodynamic diameter <4 microg/m(3))
and sulphur dioxide (SO(2)) during the first trimester of pregnancy and risk of congenital
anomalies. We used a case-control  study design among deliveries to mothers resident in the UK
Northern health region during 1985-1990. Case data were ascertained from the population-based
Northern Congenital Abnormality Survey and control data from national data on all births. Data
                                           33

-------
on BS and SO(2) from ambient air monitoring stations were used to average the total pollutant
exposure during the first trimester of pregnancy over the daily readings from all monitors within
10 km of the mother's residence. Logistic regression models estimated the association via odds
ratios. A significant but weak positive association was found between nervous system anomalies
and BS (OR=1.10 per increase of 1000 microg/m(3) total BS; 95% CI: 1.03, 1.18), but not with
other anomaly subtypes. For SO(2), a significant negative association was found with congenital
heart disease combined and patent ductus arteriosus: OR significantly <1 for all quartiles relative
to the first quartile. The relationship between SO(2) levels and other anomaly subtypes was less
clear cut: there were either no significant associations or a suggestion of a U-shaped relationship
(OR significantly <1 for moderate compared to lowest levels, but not with high SO(2) levels).
Overall, maternal exposure to BS and SO(2) in the Northern region had limited impact on
congenital anomaly risk. Studies with detailed exposure assessment are needed to further
investigate this relationship.

Son, J. Y., Y.  S. Cho, et al. (2008). "Effects of air pollution on post-neonatal infant
mortality among firstborn infants in Seoul, Korea: case-crossover and time-series
analyses." Arch Environ Occup Health 63(3): 108-13.) The relationship between air pollution
and post-neonatal mortality from all causes among firstborn infants in Seoul, Korea, during
1999-2003 was studied.  Both case-crossover and time-series analyses with a bidirectional
control-sampling approach were used to compare the effects of air pollutants on post-neonatal
mortality. The relative risk of post-neonatal mortality from all causes was 1.000 (95%
confidence interval [CI] = 0.998-1.002) for PM10, 1.002 (95% CI = 0.994-1.009) forNO2, 1.015
(95% CI = 0.973-1.058) for SO2, 1.029 (95% CI = 0.833-1.271) for CO, and 0.984 (95% CI =
0.977-0.992) for O3 for each 1-unit increase of air pollution level in the 1:6 control selection
scheme. The authors observed a positive association between air pollution and infant daily
mortality except for the studied particulate matter and ozone, although it was not statistically
significant. They obtained similar results in the time-series analysis. The risk of post-neonatal
infant death from all causes was positively associated with all studied air pollutants except
ozone. Therefore, while a positive association was observed between PM10  and post-neonatal
mortality the findings were not statistically significant.
                                           34

-------
Comments from Dr. Ellis Cowling

My specific assignment for review of the ISA for PM was Chapter 9 - Ecosystem and
Welfare Effects.

The set of 10 Charge Questions provided in John Vandenberg's letter of January 30, 2009 to
Holly Stallworth included the following  questions with regard to Chapter 9:

       How useful and complete is the scientific evidence presented in Chapter 9 regarding the
       effects of Atmospheric PM on the environment, including (a) effects on visibility, (b)
       effects on individual organisms, (c) direct and indirect effects on ecosystems, (d) effects
       on materials, and (e) effects on climate? To what extent do the discussions and
       integration of evidence correctly  represent and clearly communicate the state of the
       science?

General Comments on the Presentation of Scientific Evidence Regarding Effects of PM on
       Visibility, Individual Organisms, Ecosystems, Materials, and Climate in the ISA

It is excellent that each of the "Summary and Conclusion" statements on pages 9-1, 9-5, 9-6,
and 9-7 begins with a simple declarative sentence written in bold type indicating that "The
evidence is sufficient to infer a causal relationship between ambient PM and" [all five
significant welfare effects of ambient PM covered in Chapter 9] visibility, individual
organisms, ecosystems, materials, and climate. These statements set the stage for a through-
going discussion about the body of scientific evidence that supports these very firm statements of
causality.  That is very good!

Also useful (but frequently lacking in precision of communication because of poor definitions of
terms and the absence of a list of acronyms for the whole ISA) are the series of detailed
presentations of scientific evidence regarding the effects of PM on all five of these same
significant welfare effects including:
  •    visibility on pages 9-7 through 9-86;
             mechanisms of atmospheric deposition on pages 9-86 through 9-106 especially
             with regard to effects of airborne PM on
          o   individual organisms on pages 9-106 through 9-119,
          o   ecosystems on pages 9-119 through 9-127,
  •    materials on pages 9-127 through 9-132;
             and finally effects of PM  on various aspects of
  •    climate on pages 9-132 through 9-156.
Fortunately, some of these detailed presentations fall more closely within the areas of special
competence of some of my fellow PM CAS AC Panel Members who also were assigned to
review Chapter 9. For example, Bill Malm, regarding PM effects on visibility, and also David
Grantz, Rich Poirot, and Lowell  Ashbaugh regarding PM effects on materials and climate.
Thus, I am looking forward to listening very attentively  to their Individual Comments on some of
these special topics in contrast to the discussion about both mechanisms of atmospheric
deposition and effects of PM constituents on individual organisms and ecosystems that are
more close to my own areas of special competence.

-------
Detailed Comments on Some Limitations in the Clarity and Completeness of Chapter 9.

First with regard to Clarity: Most of my difficulty in assessing the quality of the information
provided in Chapter 9 is a result of a significant lack of precision in the terminology and lack of
definitions of the acronyms used in this chapter.

On pages 9-1 through 9-7 for example, it is not clear what is meant by the acronym EC or the
terms "crustal material," "urban excess," "urban excess carbonaceous PM," Midwest nitrate
bulge," "four light extinction components," "hydrated sulfate particles," etc.

Clear definitions of terms and especially the acronyms used in any ISA document are essential to
understanding the science that is being assessed. Many more definitions of terms and a complete
list of all acronyms used in all parts should be included in the Second External Review Draft of
this document.

Much confusion in many different parts of Chapter 9 (and in the other chapters of this ISA) are
caused by sloppiness in the use (and many different meanings) of the word "level"). The word
"level" is sometimes used to mean "air concentration," other times to mean "amount," other
times to mean "consistency," other times to mean "altitude" (distance above sea level)," other
times to mean "unchanged," other times to mean "degrees" or "extent" or "types" of visibility
impairment, species of organisms, or concepts of physiology or ecology, and still other times to
mean "tolerance" or "acceptability" (for example in attitudes of citizens about haziness in the
atmosphere). My advice is to use any  or all of these other terms when they avoid the ambiguity
that so often accompanies use of the word "level."

An illustration of a biological concept that also is presented very inadequately is  found in the
following paragraph near the bottom of page 9-5 within the Summary of Effects on Individual
Organisms and Ecosystems in the Summary and Conclusions section of Chapter 9:

      "Many  of the most important effects [of PM] occur in the soil.  The soil environment is
      one of the most dynamic sites of biological diversity in nature.  It is inhabited by
      microbial communities of bacteria, fungi, and actinomycetes. These organisms are
      essential participants in  the nutrient cycles that make nutrients available for plant uptake.
      Changes in the soil environment that influence the role of bacteria and fungi in nutrient
      cycling determine  plant and ultimately ecosystem response."

These five sentences (and especially the last sentence in this paragraph) give readers the
unintended impression that scientific studies of "microbial communities" in soil have been
shown to be affected by airborne PM [and, by implication, that the many and diverse "non-
microbial" organisms in soil — such as rodents, insects, earth worms, nematodes, viruses, and
roots of higher plants themselves - all  have been studied but have NOT been shown to be
affected by airborne PM].

Surprisingly, these very same ideas are presented very much more adequately in  the detailed
discussions on pages 9-106 through 9-119 regarding Effects on Individual Organisms and on
pages 9-119 through 9-127 regarding Effects on Ecosystems.
                                           36

-------
Second with regard to Completeness:  During the recent efforts to streamline the NAAQS
review processes, the following general policy question was recommended as a guide during the
NAAQS review processes:

       What scientific evidence and/or scientific insights have been developed since the last
       review to indicate if the current public-health based and/or the current public-welfare
       based NAAQS need to be revised or if alternative levels, indicators, statistical forms, or
       averaging times of these standards are needed to protect public health with an
       adequate margin of safety and to protect public welfare?"

Since the Clean Air Act defines "public welfare" to include the effects of air pollution on
"visibility, ... and personal comfort and wellbeing." and the present draft ISA for particulate
matter concludes that "The evidence is sufficient to infer a causal relationship between
ambient PM and visibility impairment," the most directly relevant policy questions to be
resolved by the current NAAQS-review process for PM are:

   1)  "Does the current public-welfare based NAAQS need to be revised?" and, if so
   2)  "What alternative levels, indicators, statistical forms, or averaging times of the
       secondary standard for PM should be considered in order to protect public welfare?"

The only parts of the present First External Review Draft of the ISA for PM that deal directly
with the current secondary standard for PM are:
   1)  The very confusing paragraph on lines 8 through 16 on page 1-9 of Chapter 1, and
   2)  The section of Chapter 9 titled "Urban Visibility Valuation and Preference" on pages
       9-74 through 9-83 where a preliminary attempt is made to:
             "separate out the aesthetic and wellbeing components associated with the
             visibility condition produced by a given level of air pollution when assessing the
             need for additional regulation to protect [against] the public welfare effect[s] of
             visibility under the Secondary NAAQS."  This was attempted in the present ISA
             by analyzing results from the Urban Visibility Preference Studies completed in
             recent years in Denver, Colorado, Phoenix, Arizona, British Columbia, Canada,
             and Washington, DC (see pages 9-76 through 9-81 in the ISA).

Thus, I recommend that the Second External Review Draft of the ISA for PM and the First Draft
Risk and Exposure Assessment for PM include sections that deal explicitly with the two
questions printed in italic type in the above paragraph.
My specific assignment for review of the February 2009 document titled: ''''Particulate
Matter National Ambient Air Quality Standards: Scope and Methods Plan for Urban Visibility
Impact Assessment" was Chapter 3 - Quantitative Visual Air Quality Assessment.

Having carefully reviewed all three chapters in this document, I was generally satisfied with the
approach outlined in this Scope and Methods Plan but recommend that earnest consideration also
                                           37

-------
be given to including sections in the Second External Review Draft of the ISA for PM to be
completed in July 2009 and the First Review Draft of the Urban Visibility Impact Assessment
(UVA) to be completed in September 2009 that deal explicitly with the following two specific
policy questions:

   1)  "Does the current public-welfare based NAAQS need to be revised?" and, if so

   2)  "What alternative levels, indicators, statistical forms, and averaging times of the
       secondary standard for PM should be considered in order to protect public welfare?"
                                          38

-------
Comments from Dr. James Crapo

The EPA Staff are to be congratulated on producing a comprehensive and well designed
overview of issues related to ambient paniculate matter including the substantial amount of
research on PM published since the last CASAC review of this topic. In general, the integrative
tables and figures are clear and presented in a form that allows consideration of the scope of the
data with respect to potential adverse health effects for PM.

Chapter 2 - Integrative Health Effects
This chapter provides an integrative summary and assesses the likelihood of causal relationships
for specific health effects with respect to PM exposure. I like the general approach the EPA has
used, including their  five-level hierarchy for causal determination. This is a rigorous approach to
the central issue  that  has been requested on multiple previous ISA documents. The EPA staff has
critically reviewed the relevant literature on this topic and has made concrete recommendations
regarding their conclusions with regard to causality. I applaud the approach of starting with a list
of the key health effects and a firm statement with regard to conclusions of causality and
following this by a brief description of the key findings supporting the conclusions. Hopefully,
this will become a model for future ISA documents.

In general, I agree with the causality conclusions with regard to the key health effects and I agree
with the selection of the health effects for emphasis in this chapter. The focus of this chapter is
on short- and long-term  exposures to PMi0 and PM2 5. This implies that a firm decision has been
made not to consider regulatory recommendations with regard to PMio-2.5.1 believe we will
ultimately need to evaluate epidemiologic and animal studies data with regard to PMi0-2.5 in
order to understand the contributory effects of coarse particulates. Data on PMio-2.5 is
incorporated in other chapters on Health Effects but not brought forward to the summary and
conclusions in Chapter 2. From a regulatory perspective today this may be appropriate;  however,
I would recommend that at least a summary section on PMio-2.5 be included in Chapter 2.

I also liked section 2.2.3 on constituents or sources linked to health outcomes. There are
insufficient data  to derive concrete conclusions; however, there are suggestions that specific
constituents or sources will eventually be linked with specific health effects. This is likely to be
an area that will  expand in future reviews of PM.

Chapter 5 - Possible Pathways/Modes of Action
This chapter provides a reasonable review of the possible pathways  or modes of action of
particulates in contributing to adverse health effects. The chapter is short, focused and provides a
number of alternative frameworks to consider the mechanisms of actions of particulates. I found
this chapter useful. Part  of its strength is its brevity and its liberal use of illustrations. I would
recommend keeping this approach for this section of the ISA.

Chapter 7 - Integrated Health Effects of Long-Term PM Exposure
Chapter 7 is thorough and well written. It provides an excellent overview of the long-term health
effects of PM exposure.  An important strength of this chapter is the  creation of Figures that
clearly illustrate  the combined primary findings with each of the topics (i.e., cardiovascular and
systemic risks, respiratory risks, reproductive and developmental risks, cancer risks and overall
                                           39

-------
mortality. This is well done in Figure 7.8. Figure 7.1 is also a very useful summary figure. I
would recommend that summary figures be created for each of the sections - even where the
data suggests no significant effect. The written summaries in each section are appropriate;
however, I think there should be tables or figures that summarize and compare the primary data
for each section.

Chapter 7 would be strengthened by shortening it, achieving greater focus on the new findings,
and a focus on the summary tables as suggested above.
                                           40

-------
Comments from Dr. Douglas Crawford-Brown
I was not assigned to a specific charge question, so I chose those that I felt are most in line with
my area of expertise. A general comment is that the document is longer than I might have hoped
to see. The other ISAs (for other NAAQS) produced recently have set a trend of significantly
reducing the amount of material that would have been covered under the older Criteria
Document system. This PM ISA reverses that trend. I realize there is a LOT of PM material
available in the literature, but I still would have liked to see a little more distillation of the
literature down to the key studies and points before production of the ISA.

1.      The framework for causal determination and judging the overall weight of evidence is
presented in Chapter 1. Is this framework appropriately applied for this PM ISA? How might
the application of the framework be improved for PM effects?

There is now emerging a consistent framework being applied across the various ISAs, so it is
becoming easier to see the reasoning behind the causal conclusions drawn in this and other ISAs.
There are quibbles I might have with the particular framework chosen, but it is as good as any
other I have seen and at least serves to draw the analyst's attention to the key questions to be
addressed. It appropriately separates a judgment of causality from one of there being a C-R
function. The categories of strength of causality are sufficient and clearly defined, so I found it
easy to understand why the analyst chose to place a given study and effect into a given category.
Overall, I am satisfied with this system as a sufficient tool, and recommend no further tweaking,
which allows everyone to now settle on examining the particular conclusions being drawn within
the framework.

The only point that continues to concern me is that the judgments of causality still tend to be
categorical: PM is or is not causally connected to effect X. I think a better system is one in which
the degree of support for such a claim changes as the level of exposure changes. It would be
useful to see how the support changes as one approaches low levels being considered for
regulatory targets. There is some hint of this throughout the application of the framework, but it
is never made explicit, especially in Chapter 1.

2.      Chapter 2 presents the integrative summary and conclusions from the health effects
evidence at the beginning of the ISA with the evidence characterized in detail in subsequent
chapters. (Environmental  and public welfare effects evidence is evaluated and summarized in
Chapter 9.) Is this a useful and effective summary presentation? How does the Panel view the
inclusion in Chapter 2 of only health categories with causal determinations of (a) likely to be a
causal relationship or (b) a causal relationship?

I found Chapter 2 very useful, and it went a long way towards alleviating the general concern I
mentioned above in regards to the length of the document. I feel it is appropriate to summarize
only the findings from the studies that fall into these two causal categories, although mention can
be made of studies/effects that did not meet these criteria (without providing summaries of
these). I found the discussion of the implications of exposure and dosimetry for the conduct and
                                           41

-------
interpretation of epidemiologic studies useful, which was a real strength of this ISA, especially
since this is one pollutant for which a lot can be said on these points.

The discussion in the chapter concerning composition of particles was adequate but not
especially insightful. The summary table was good, but I would have expected a bit more of an
assessment of, and justification for, a decision to either consider or not consider composition in
any final risk assessments that might be conducted based on the ISA. I must admit that I read the
REA first, and so was primed for this issue in turning to the ISA (the wrong order in which to
read these), but still I expected more in the ISA on this point.

The chapter correctly summarizes the susceptibility and vulnerability issues raised in subsequent
chapters, and so serve this issue well. I was less happy with the discussion of C-R curves, feeling
it needed more of a mathematical treatment in this chapter. I always expect a discussion of such
curves - whether  in summary or in the complete chapter being summarized - to include data
plotted to visually indicate where there is a clear positive slope and where this disappears into
then noise of studies, and some indication of the shape of the curve at all levels of exposure.
Chapter 2 does not go very far down this road and so is less effective as a summary than I would
have wanted.

4.      The dosimetry of PM is discussed in Chapter 4.  The primary focus is on factors that
might lead to differences in deposition and clearance between individuals, species, and as a
function of the physicochemical properties of particles.  Is the review of basic dosimetric
principles presented in sufficient detail?  Are the new particle translocation data adequately and
accurately described?  Recognizing an overall goal of producing a clear and concise chapter, are
there topics that should added or receive additional discussion? Similarly, are there topics that
should be  shortened or removed?  To what extent does the Panel find Annex B appropriate,
adequate and effective in supporting the ISA?

I was comfortable with Chapter 4. It provided an overview of mechanisms of deposition and
clearance at just enough detail to give the reader an understanding of the issues. It was perhaps
left to the reader to see the implications for study interpretation, or for the drawing of
conclusions on causal connections in particular studies, but the information is there to help in this
regard. I found the level of technical detail also appropriate for a topic that can quickly become
too mathematical  for many readers.

The one area in which I would suggest a bit of improvement is in the inhomogeneity of
deposition and clearance, especially in regards to carinal ridges. This is mentioned by never
really explored.

I kept returning to Annex B for detail, and so I suppose this indicates the Annex served its
purpose. I found it contain a lot of useful information, even if it was a bit difficult to sort through
it to find exactly the point being raised in the chapter itself.
                                           42

-------
8.      What are the views of the Panel on the definitions of susceptibility and vulnerability in
Chapter 8? Are the characteristics included within the broad susceptibility and vulnerability
categories appropriate and consistent with the definitions used?

I am comfortable with these definitions. This is a topic that always gets a lively debate, and
people tend to have some pretty entrenched positions on the issue (with good arguments to
support their positions). But in the end, what matters is that the terms are well defined both
verbally and operationally, and then applied consistently. I feel the authors have satisfied these
criteria in the ISA and so the characteristics considered serve the intended purpose as well as any
other set of definitions might.

10.    This first external review draft PM ISA is of substantial length and reflects the copious
amount of research recently conducted on PM. EPA has attempted to succinctly present and
integrate the policy-relevant scientific evidence for the review of the PM NAAQS.  Does the
Panel have opinions on how the document can be shortened without eliminating important and
necessary content?

This is back to my original comment. The document was too long for me to read in its entirety. I
instead read about 500 pages of it in enough detail to determine whether the information was
sufficient. This leaves me with the uncomfortable position of worrying about whether the other
500 or so pages might contain information or analyses that would influence the interpretations
and judgments I had formed in the sections I did read in detail.

Chapter 3 strikes me as a chapter that could usefully be pruned by half. It reads a bit like a data
dump, with every bit of information to hand put into the document just in case someone might
find it of interest. In most of the cases of page after page of data, I had little sense that these data
were telling a story that would be needed later in conducting an exposure assessment or
interpreting epidemiologic studies. This doesn't mean the data will not prove useful at some
point - only that the authors didn't focus my attention onto specific reasons WHY the data might
be important.

By way of contrast, there is about the same amount of data available in the literature on particle
deposition and clearance, but Chapter 4 is a third the size of Chapter 3. Chapter 4 also does a
better job of focusing the reader's attention onto the significance of the data being presented.

I saw little in Chapter 5 that could be deleted without loss of significant content, and so would
not expect cuts to be made here.

Chapter 6 is the longest by far. I found myself focusing almost entirely on summary tables and
figures, rather than the exhaustive page-after-page listings of characteristics of specific studies.
The chapter could easily be shortened by half without loss of any content that the reader will care
about. As it is written currently, it suffers from all of the problems of the older criteria
documents. This shortening would be accomplished by provision of even better summary tables
and figures, perhaps with a Comments column to explain any special features of a study the
author wants to highlight.
                                           43

-------
In fact, Chapter 6 is even longer than the 200+ pages contained within it, because much of
Chapter 7 is a similar presentation of study results. The purpose of Chapter 7 is to provide an
integrative approach, but I could find little of that in the chapter. It became instead yet another
place to serve as a repository of exhaustive study descriptions. There should be little need to
describe studies by this chapter, and certainly no need to present even more data. The chapter
should instead focus on how the materials of earlier chapters, including the dosimetry
considerations, can be integrated into a consistent picture of which effects are causal, and at what
levels of exposure, and under what conditions. There are places where Chapter 7 starts to
approach this goal, but then the reader is plunged back again into listings of studies. This chapter
is the one most in need of not shortening per se, but a better focus on the purpose of the chapter.
                                            44

-------
Comments from Dr. David Grantz
The document as a whole makes a real contribution and is well done, quite readable, and
relatively complete. The following represent comments or suggestions for revision of the ISA,
not in priority order.

   1.  It would be helpful to provide a list of acronyms. Further, it would be useful to describe
       briefly air monitoring programs such as AQS (page 2-2, line 11) to present the
       characteristics that make them particularly suitable for the specific analysis under
       discussion.

   2.  The approach of Chapter 9, which deals with all welfare effects considered together,
       differs considerably from that of Chapter 2 (and subsequent supporting chapters 3-8),
       which deals with discrete health endpoints. There seems to have been little effort to apply
       the criteria of causality laid out earlier (e.g. Table 1-2), even for visibility which is the
       strongest welfare component in terms of available data. Indeed visibility is a nearly
       discrete endpoint which could be separated into its own chapter. There is a sense that
       other components of welfare effects have been determined prior to preparation of this
       ISA to be non-"policy relevant", and therefore have received little critical attention.

       Inconsistencies between the treatment of health and welfare effects are numerous. It is
       interesting that (page 9-113, line 23-26) the most interesting plants are said to be the
       tolerant members of the population, whereas in the health sections (and in the legislation)
       it is the  most sensitive and most vulnerable members of the population that are to receive
       greatest attention. A committed  search for the most sensitive receptors for specific
       endpoints (such as the sentinel species considered on page 9-116) might help to better
       focus the discussion of welfare effects and aid the search for quantifiable C-R
       relationships that might guide development of a separately determined secondary
       standard for PM.

       A detailed chart of causality begins each subsection in Section 2.3, and would be useful if
       added for each subsection of Chapter 9. The three non-visibility findings of "inferred
       causality"~between PM and effects on organisms and  ecosystems, on materials, and on
       climate  are much too general, analogous to a finding of causality between PM and human
       health, rather than the various endpoints  laid out in Chapters 2-8. Further,  it is unclear
       whether the phrase (Section 9.2.2) of "sufficient to infer a causal relationship" has
       exactly the same meaning as the phrase "sufficient evidence to conclude that a causal
       relationship exists" that appears in Chapter 2. This should be clarified.

       A consequence of this general approach to welfare effects is that data that would have
       been considered to contribute to a finding of "likely causal" or "contributing to biological
       plausibility" in the health context, receive little more than an uncritical listing in the
       welfare context. For example (page 9-108, line 1-12), the  analysis of PM by Sheesley et
       al. (2004) might be used to infer likely causality or biological plausibility, but is instead
       summarized as "does not provide the data needed to assess risk". Similarly, data such as
                                            45

-------
   that presented by Oliveira and Pampulha (2006; page 9-118, line 28-30) and other studies
   might be amenable to semi-quantitative analysis, yielding estimates of soil
   concentrations, inferred rates of deposition, and potentially even atmospheric
   concentrations of PM. The challenges laid out for ecosystem data (page 9-120, line 26
   through page 9-121, line 2) are not so different from the early  stages of health effects
   investigations. These and other possible opportunities could be pursued throughout the
   welfare section, as a means to begin development of the C-R relationships required to
   support a secondary PM standard. A more critical reading of the available literature for
   quantitative relationships may lead closer to C-R relationships than seems possible from
   the current presentation. The discussion in Section 9.2.2, for example, states generalities
   but never discusses the direction or magnitude of PM-induced changes. The text does not
   effectively support the finding of causality that begins the section.

3.  The material in Chapter 9 could be reorganized to emphasize discrete endpoints, and
   possible causal relationships related to each. Visibility could be placed in its own, better
   focused chapter. Material presented in the current chapter 9 on characterization of
   particles could be relocated to precede the health sections, and the material on deposition
   to extensive landscapes  could be placed earlier where it might inform discussion of
   resuspension and personal PM clouds in the health context. This would isolate the
   material on effects on organisms and effects on ecosystems or communities, which could
   be considered together or separately, but in the context of discrete endpoints. This would
   focus the discussion and remove the need for  abrupt transitions, such as that into
   discussion of deposition in Section 9.4 (page 9-83).

   The material on toxic metals should be assembled as needed for causality determinations
   within each endpoint. In the current draft, discussion  of the nature and role of heavy
   metals is diffused throughout the chapter, but never brought to a focus  or conclusion. As
   above, this would avoid the  abrupt introduction of metals in Table 9-3 (page 9-89) and in
   the text at top  of page 9-90,  where it appears out of place.

   The discussion of organic pollutants is somewhat weak. These materials represent direct
   threats to organisms and ecosystems, and substantial indirect threats to food chains
   including human food supply. As such they may become central to the determination of a
   secondary PM standard. This discussion (page 9-114, line 27 to 9-116, line 10) is really a
   discussion of semivolatile compounds. These have unique transport properties,
   characterized by repeated distillation on surfaces as a function of diel trends in
   temperature. This is not well captured by the text. The discussion of fugacity, which is
   defined at a constant temperature (page 9-116, line 5-7), does  not capture the well known
   leap-frog behavior of SVOCs across the surface of the earth. The discussion of POPs
   (page 9-103, line 30-31), is confusing with regards to the air plant response route (really a
   contact route), and uptake by above ground plant tissue, (also  a contact route) in contrast
   to the distinction between the two (top of page 9-104).

   The discussion of bioaccumulation of organics (page 9-124 to 9-125) can be made quite a
   bit more quantitative, even if causality cannot be demonstrated at this stage.
                                        46

-------
4.  There is some uncertainty about mercury residence times, and about the oxidation state of
   Hg (page 9-101-102, lines 31-6) under discussion at each point. It would be useful to
   state explicitly what combining forms Hg(II) is found in, and (e.g. page 9-103, top)
   whether gas or particle, and what oxidation state, is under discussion at each point.

5.  The use of visibility as a surrogate for welfare effects may have unintended
   consequences. This is considered more fully in comments regarding the visibility plan. In
   order to keep regulatory environments similar for day and night hours, a movement
   toward modeled or virtual visibility is to be encouraged. This avoids the application of
   dark viewing techniques  (e.g. page 9-12, line 11), and allows explicit use of developed
   relationships between visibility and parameters such as RH, to be applied at all hours. In
   the limit, this could avoid use of dark periods to dump particulates that exceed the
   secondary standard.

   Visibility as a standard metric is confounded by regionally differing policy relevant
   backgrounds, and regionally differing popular perceptions of acceptable visibility. In the
   ISA this is shown for residents of Chilliwack and Abbotsford (Page 9-80). Buy-in of
   regulated communities may be compromised by regionally differing standards for PM
   mass, to achieve differing standards of visibility.

6.  The treatment of urban visibility valuation (Section 9.3.5 is written rather loosely, and
   never reaches a focus or conclusion. The acronym, VAQ, is used in several incompatible
   senses. The definition (page 9-74, line 24) improbably equates impairment with quality.
   VAQ is considered something to be seen (page 9-75, line 12), and then more
   appropriately as a quality that can be better or worse. The modifier, "low", could be
   inserted prior to "VAQ" (line 25-26), since it is not likely that high VAQ would diminish
   a sense of well being.

7.  The discussion of dry deposition (particularly page 9-86, lines 20-26 and page 9-90, lines
   17-28) is somewhat vague and incomplete. The roles of sedimentation and resuspension
   are not considered adequately. The discussion is not entirely consistent with Figure 9-52.
   The consideration of eddy covariance techniques (page 9-88, line 34) might reconsider
   whether coarse particles are transported efficiently in high frequency eddies.

8.  Page 9-47, line 8, in wintertime it may be that oxidizing potential to convert NOx to nitric
   acid is limiting. While this appears to be the implication of this sentence, this is an
   important point that requires further elaboration with respect to mechanisms and
   consequences for air quality.

9.  The discussion regarding exchange of PM-borne materials across vegetated surfaces
   (page 9-95, line  1-3) is quite important and generally well done here. The concept of
   measured dry deposition exceeding bulk deposition requires further elaboration,
   particularly as to measurement technique for the two parameters. In the discussion of
   exchange of mercury (page 9-97, line 26-27) the two forms of deposition that are
   mentioned are both dry deposition, in contrast to the text.
                                        47

-------
10. The discussion of individual species near the Harjavalta smelter (page 9-126, line 21-24)
   is of interest, but should be placed in the context of the section heading—Ecosystem
   Function. The isolated consideration  of a few insect and spider species, without reference
   to their roles in the ecosystem, is analogous to a statement that influenza is good for
   buzzards. Misses the point a bit.

11. The discussion of direct impacts of coarse particles (page 9-108  to 9-109) is incomplete.
   Many consequences are mediated by disruption of leaf cuticle and waxes, from pathogen
   defense to control of water loss. Surface pH is important,  particularly for colonization,
   but also for wax degradation. These need to be elaborated upon. It is not clear, and at
   least requires better referencing, that  salt is generally taken up through abrasions in the
   shoot rather than through the roots, and that this leads to plants leaning away from the
   ocean.

12. Minor comments:
   Page 9-2,line 5, there appears a discontinuity between  discussion of pristine skies and
       polluted ones.
   Page 9-5, line 13, it is not correctly parallel to refer to PM deposition as a pollutant (PM
       is the pollutant).
   Page 9-6, line 2, N should be Ni.
   Page 9-7, line 11, BC is introduced here for possibly the first time. Is this the same as EC,
       introduced earlier? Either way there needs to be a definition given.
   Figure 9-3 (page 9-14), it is unclear what 1.5 stands for in the internal labels for each line.
   Figure 9-31 (page 9-49) it is unclear what the three sizes of pie charts in the legends, and
       the numbers 20, 60, and 100 represent.
   Page 9-22, line 10-14, better to provide a mathematical equation. The form and intent of
       this verbally described transformation is unclear.
   Page 9-48, line 8, should be "east" end of gorge.
   Figures 9-38 to 9-43 (Pages 9-58 to 9-64) are quite difficult to interpret. The legend and
       caption can both be made more user friendly.
   Page 9-105, line 14, it is unclear whether petrogenic refers to petroleum or geology.
   Page 9-113, line 12-22, a better distinction should be made between tire dust and road
       dust.
   Page 9-127, line 27-28, use of a bioluminescence assay is a type of bioassay that directly
       measures the toxicity of the mixture of metals under consideration. This should be
       presented in the context of such direct effect measurements.
   Page 9-136, line 24, the ratio of diffuse to total solar radiation is lowest at noon.
   Page 9-139, line 4-5, it is not clear what factor of 2 refers to.
   Page 9-140, line 26-28, the reduction in radiative forcing  over the 80 year recover period
   requires further explanation here.
                                        48

-------
Comments from Dr. Joseph Helble
1.  Particle morphology may affect particle transport properties, depending upon particle size.  It
   will also affect particle surface area regardless of size.  Because surface area and particle
   transport properties may both affect exposure, discussion of current understanding of
   morphology or surface area distributions would be beneficial.  This is acknowledged in the
   ISA in section 3.2 (line 23, page 3-3) where shape along with size and composition is listed
   as a variable characteristic of atmospheric PM.

2.  This point is reinforced in Section 3.4.13. page 3-24, where it is noted that "methods are
   being developed to measure the surface area of ultrafine particles (line 23)."  No additional
   discussion of surface area is provided. It  arises again on p. 3-37  line 9, where "particle
   shape" is listed as one possible cause of differences noted among samples at  supersites.

3.  A listing of acronyms and abbreviations is needed at the start of the ISA.  Some are not
   defined at all, others are not defined until  they have been used many times in earlier sections.
   A listing is included in the Scope and Methods documents and is also warranted in the ISA.

4.  Fly ash from combustion of coal, oil, and wood can also produce particles in the PM1.0 size
   range and is therefore also a potential source of Fine particles (Table 3-1, page 3-6). It is in
   part a source of the metals that are listed in the table.

5.  Similarly, primary PM ("minerals") can be emitted as fly ash from fossil fuel and wood
   combustion, a source category listed in Table 3-1 but not in Table 3-2. The tables are largely
   consistent, but review for full agreement would be helpful.

6.  A brief narrative description of the sources contributing to the "miscellaneous" categories of
   PM 10 and PM 2.5 sources (Figure 3-2) should be included.

7.  Section 3.3.2.2 page 3-12 line 18 the sentence beginning Robinson et al. (2007a) needs
   clarification.  Presumably the statement indicates that direct measurements of combustion
   source emissions do not accurately capture all of the semi-volatile organic content, leading to
   an underestimation of the contribution of combustion sources to  semi-volatile organics
   observed in PM.

8.  P. 3-37 line 20 discussed an artifact when QBT filtering was used.  Earlier in this section,
   page 3-21, only QBQ and TBQ are mentioned as commonly used methods. If QBT is not a
   typo, it should also be discussed on p.  3-21 when the other filter  methods are described.

9.  Comments on the higher maxima noted in Table 3-11 PM 10 distributions from AQ data
   would be helpful.

10. Page 3-104 notes that "compositional profiles used in receptor models can be extensive."
   And variable? Worth addressing explicitly if the literature suggests differences would affect
   conclusions drawn from CMB models.
                                           49

-------
11. Page 3-146 line 2 the sentence beginning "Particle chemistry" is confusing.  It implies that
   chemical composition affects particle deposition behavior.  While there is certainly a link
   between composition and size it is not a determinative relationship, and clearly it is size (and
   structure) that primarily affect deposition patterns.

12. The discussion relating exposure assessment and socioeconomic status makes points
   regarding neighborhood scale monitor distribution that do not appear to be consistent with
   the data in the subsequent tables. Specifically, lines 20-23 page 3-161 argue that
   "neighborhood scale SES issues are not shown to be well represented." The inference drawn
   from the text as written is that monitors are not sufficiently well distributed, at the
   neighborhood level, in neighborhoods of lower SES status.  A comparison of data in Tables
   3-30 through 3-32 indicates, however, that a higher percentage of the less-educated
   population lives within 1 km of a monitor than is true for other educational attainment levels.
   It may be reasonable to argue that a higher concentration of monitors is needed in these
   neighborhoods, but the data as presented do not appear to suggest disparity in their siting.
13. Typographical Errors

       a.  Page 3-8 line 9 the word "exhaust" should be inserted after "vehicle"
       b.  Page 3-27 line 23 "from" should be inserted between "Except" and "some"
       c.  Page 3-66 the figures present correlation as fractional values, the text describes them
          in percentage terms. The discussion should be made consistent with the figures.
       d.  Page 3-72 Figure 3-30 caption indicates "Pittsburgh," but the legend indicates
          "Philadelphia.  This needs to be corrected.
       e.  Page 3-85 line 10 insert "are" between "than" and "ultrafine"
       f.  Page 3-103 line 24 insert appropriate table numbers
       g.  Page 3-114 line 12 delete "described"
       h.  Page 3-115 line 112 sentence beginning "In this section" unclear - comma needed
          after "average" - ?
       i.  Page 3-115 line 28 "have" should be changed to "has"
       j.  Page 3-116 line 5 a colon needed after "components"
       k.  Figures 3-61 through 3-63 need to be made larger and sharper
       1.  Page 3-141 line 14 space needed between "be" and "evaluated"
       m. Page 3-143 line 28 "suggests" should be "suggest"
       n.  Page 3-145 line 2 "utilities" should be "utility"
       o.  Page 3-148 Line 9 the "88" in concentrations needs to be deleted
       p.  Page 3-149 line 1 Figure 3-71  should be changed to Figure 3-74
       q.  Page 3-150 line 10 insert "and" before "indoor"
       r.  Page 3-152 line 2 insert) after the first Zn

-------
Comments from Dr. Rogene Henderson
Chapter 1-Framework
I thought the explanation of the EPA framework for causal determination was well done. It will
provide the basis for future NAAQS documents for all criteria pollutants. In the future this
standard approach might be placed in an appendix, but it is appropriate to include it in the PM
ISA text for discussion.

Chapter 2—Integrative Health Effects Overview
I found this to be a useful chapter. I thought the chapter needed a few more words about the
health effects of short and long term exposures to ultrafme PM and PMi0-2.5  particles, not just the
one sentence at the bottom of page 2-11 saying we have inadequate evidence to infer causal
relationships. The chapter might also be expanded to include the summary sections from
Chapters 6 and 7 (see suggestion below) and serve as the health effects section of the ISA, with
large parts of chapters 6 and 7 as appendices. That is not to say that Chapters 6 and 7 are not
important. Those chapters contain the detailed information upon which Chapter 2 is based. But
if one valued  a shorter ISA, one  could use an expanded Chapter 2 with appendices for the health
effects. For my own benefit, I made a table that included each of the four particle sizes for rows
and each of the major endpoints as columns, with the resultant cells showing the finding in terms
of causality. I did this for both the short term and the long-term exposures. Such a simple
summary table would be appropriate in Chapter 2.

Chapters 6 and 7:
These are  highly valuable summaries of the health effects  of short and long term exposures to
PM conducted since the last review.  I found the sections labeled "Overall Summary and Causal
Determination" found at the end of each health effects section especially useful.  I wonder if you
might combine these  sections with what is in Chapter 2, edit it to remove redundancies, put the
details of the new studies described in Chapters 6 and 7 in Appendices, and use the expanded
Chapter 2  as the Health Effects chapter.
                                           51

-------
Comments from Dr. Phil Hopke
This Chapter is generally well written.  However, there are a number of small additions and
corrections that would improve its overall content and presentation.
Xia and Hopke (2006) observed seasonal variation for the two major diastereoisomers coming
from isoprene oxidation with the concentrations occurring during the summer and the lowest
concentrations occurring during the winter.  The variation of the carbon contribution of 2-
methyltetrols to OC was found to follow the same pattern as the concentration variation of 2-
methyltetrols. During summer, the period of high photochemical activity, the maximum carbon
contribution of 2-methyltetrols to OC was 2.8%.
It should be noted that the byproducts of terpene oxidation forming SO A are inherently oxidative
in nature.  Docherty et al. (2005) found that 47% to 85% of SOA mass is composed of peroxide
compounds. Organic peroxides represent an important class of reactive oxygen species (ROS)
that have high oxidizing potential and could cause oxidative stress in cells on which such
particles deposit.
A major question is whether to persist in using a filter based method as the Federal Reference
Method where we know that the current FRM for PM2 5 has good precision, totally undefined
accuracy and when operating perfectly provides data with 66% of the days unmeasured.  I
suggest that such a measurement protocol is no longer acceptable.  It is critical to have
essentially complete data capture, higher time resolved data, and measurements that better reflect
the actual  concentration of particulate matter that exists in the atmosphere without the unknown
and unknowable losses of semivolatile materials that come from a 24 "equilibration" period. The
document  is currently silent on these fundamental flaws of the FRM approach while putting in
detail on studies of how people try and replicate the inaccurate measurements known to be
provided by the FRM.  It is time for a change in paradigm and move to continuous monitors that
can remove some of the particle-bound water without as drastically disturbing the other
semivolatile components.
There should be some description of the Wagner and Leith passive monitors for PMio and PMio-
2.5.  Since the short-term standard for PMio has been eliminated, it would be possible to provide
long term  monitoring for PMio or PMio-2.5 for many more locations using passive samplers and
thus, they  are worth discussing.
I find Figures 3-4 and 3-5 very hard to follow. They are described in a single sentence on page
3-29 and the point being made is left quite vague. Either clarify why the figures are there or
remove them.
There is no discussion of trends in composition.  We have data now from 2000 to 2008 at a
number of locations across the US. Is the composition changing relative to the overall mass
concentration?
There is no mention of the ultrafine measurements in Rochester (Jeong et al., 2004, 2006)
looking at the seasonal variability in the diurnal patterns of ultrafines and distinguishing between
vehicular,  plume, and regional nucleation events.
Although not widely used, Unmix should also be mentioned on Page 3-105.  It is mentioned on
3-107, but the methods should be introduced together.  Also it should be noted that PMF and
Unmix have been applied to particle size distributions to obtain source identification of ultrafine
particles.
                                           52

-------
An issue that is not presented is whether intercontinental dust events are really contributors to the
PRB or just constitute unusual events. For example, there was a large Saharan dust storm in
early July 2002 that contributed 25 |ig/m3 on July 1 and 24 ng/m3 on July 2, 2002 (Lee et al.,
JGR 2006).  It is logical to assign these well defined specific events to the PRB?
There have been useful new approaches to using satellite data to estimate fine particle exposure
that are worth noting.
Estimating fine particulate matter component concentrations and size distributions using
satellite-retrieved fractional aerosol optical depth: Part  1 - Method development
Liu Y, Koutrakis P, Kahn R: JOURNAL OF THE AIR & WASTE MANAGEMENT
ASSOCIATION Volume: 57  Issue: 11  Pages: 1351-1359 Published: NOV 2007
Estimating fine particulate matter component concentrations and size distributions using
satellite-retrieved fractional aerosol optical depth: Part 2 - A case study, Liu Y, Koutrakis P,
Kahn R, et al. JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION Volume:
57  Issue: 11  Pages: 1360-1369 Published: NOV 2007

It should be noted that one of the effects of ambient ozone when it infiltrates into indoor air is
that it can produce new particle formation by reaction with the terpenes released by a variety of
cleaning and personal care products.  Weschler has published extensively on this subject.

-------
Comments on Chapter 4
There has been some useful recent work on the penetration of fibrous particles (high aspect ratio)
through head and TB airways.
Record 1 of 6
Wang, Z; Hopke, PK; Ahmadi, G; Cheng, YS; Baron, PA. 2008. Fibrous particle deposition in
human nasal passage: The influence of particle length, flow rate, and geometry of nasal airway.
JOURNAL OF AEROSOL SCIENCE 39 (12): 1040-1054..
Author Full Name(s): Wang, Zuocheng; Hopke, Philip K.; Ahmadi, Goodarz; Cheng, Yung-
Sung; Baron, Paul A.
DOI: 10.1016/j.jaerosci.2008.07.008

Record 2 of 6
Zhou, Y; Su, WC; Cheng, YS. 2008. Fiber Deposition in the Tracheobronchial Region:
Deposition Equations. INHALATION TOXICOLOGY20  (13): 1191-1198..
Author Full Name(s): Zhou, Yue; Su, Wei-Chung; Cheng, Yung Sung
DOI: 10.1080/08958370802233082

Record 3 of 6
Zhou, Y; Su, WC; Cheng, YS. 2007. Fiber deposition in the tracheobronchial region:
Experimental measurements. INHALATION TOXICOLOGY 19 (13): 1071-1078..
Author Full Name(s): Zhou, Yue; Su, Wei-Chung; Cheng, Yung Sung
DOI: 10.1080/08958370701626634

Record 4 of 6
Su, WC; Cheng, YS. 2006. Deposition of fiber in a human airway replica. JOURNAL OF
AEROSOL SCIENCE 37 (11): 1429-1441..
Author Full Name(s): Su, Wei-Chung; Cheng, Yung Sung
DOI: 10.1016/j.jaerosci.2006.01.015

Record 5 of 6
Su, WC; Cheng, YS. 2006. Fiber deposition pattern in two human respiratory tract replicas.
INHALATION TOXICOLOGY18 (10): 749-760..
Author Full Name(s): Su, Wei-Chung; Cheng, Yung Sung
DOI: 10.1080/08958370600748513

Record 6 of 6
Su, WC; Cheng, YS. 2005. Deposition of fiber in the human nasal airway. AEROSOL SCIENCE
AND TECHNOLOGY 39  (9): 888-901..

DOI: 10.1080/02786820500295685
                                       54

-------
Comments from Dr. Donna Kenski
General impressions (Chapters 1-3 and 9): Great job summarizing and organizing an enormous
amount of data. Obviously it's a very large document, but in light of the amount of new material
available, I don't think the length is excessive. Only Chapter 9 seemed to overshoot the mark and
need to be trimmed down some (see suggestions below).  But overall, the document read well,
used appropriate graphics to demonstrate and summarize data, and it was easy to follow the logic
and organization.
Chapter 1:
1-20, 23-25: This wording makes me a little uncomfortable; just what is meant by 'an objective
appraisal ... which includes weighing alternative views on controversial issues'? Must every
comment be considered equally?  Surely all comments are considered in light of the supporting
science?
1-25, Sec. 1.6: Well, it's hardly concise, but given the extensive new information available, the
length is  probably unavoidable.  Wish there was an audio version we could download to our
Ipods.
1-25, 13: edpolicy -> policy
Chapter 2:  I think this summary section really belongs at the end of the document, with perhaps
a much briefer (3-4 pg) executive summary in the beginning of the report, like what was
developed for the NOx/Sox 2ary ISA. This chapter is just too long and detailed to serve as an
executive summary, although it is an admirable wrap-up of the whole report.  It would just work
better at the end.
Chapter 3:  This chapter was really well done, and I think it had just the right level of detail and
focus.  My comments are mostly minor.  Section 3.1 needs more work - it reads like it was cut-
and-pasted from other sections without any clear organization of its own.  Once you get past the
first 2 paragraphs, there's no 'flow', and it meanders between PM, PM2.5, PMcoarse,, and PM10
in a somewhat confusing way. At the end of the 3rd paragraph (p. 3-2, lines 7-10) the text
abruptly switches from discussing SOA to evaluating FRM monitors; it is not clear what the
connection is between these subjects, if any. Paragraph 4 (p. 3-2, lines 11-12) states that levels
and spatial distributions of PM2.5 have remained relatively unchanged, but later sections go on
to more accurately state that concentrations of PM2.5 have declined in most places  around the
country by about 10% since 1999.
I think the balance of data presented in Chapter 3 and supplementary data presented in Annex A
is just about right. However, the references to data in Annex A would be more helpful if they
were more specific (i.e., refer to table A. 11, not just Appendix A). And a table of contents for
Appendix A would be helpful as well, just to make it easier to find the relevant information.
Section 3.4.1.5 on Emerging methods was really too brief to be useful. Beef up the description
on satellite methods here or in the Annex,.
On page 3.37, the short section on comparison of monitors at supersites seems out of place.
Artifacts etc. are better discussed in Sec. 3.4.1.2.
Section 3.5 on ambient concentrations is really well done and nicely organized.  Tables 3-11 and
3-12 are great, but please repeat the headings when a table has to be split between 2 pages as
T3-11 is (p. 3-40, 3-41).
The paragraph on p. 3-53  could be expanded to more comprehensively discuss what the FRM
measures vs. what the CSN monitors measure.  These issues impact subsequent interpretations of
                                          55

-------
the exposure and health data as well so they deserve to be thoroughly vetted here.  Also, please
include a mention (probably obvious to all by now, but still...) that the crustal component can
have significant industrial or anthropogenic origins, despite the name.
Figures 3-34 through 3-36 are particularly nice, with their corollary figures 3-25 through 3-27.  I
appreciate the care that went into illustrating these basic relationships between PM2.5 and PM10
with appropriate data analyses.
3-7, 14: of the microorganisms listed, 3 of the 4 are typically less than 2.5 um. Why are they
presumed to be in the coarse mode? Are bacteria  and viruses necessarily associated only with
larger particles? I don't think this is the case .  How much do these contribute to mass in the fine
and coarse fractions?
3-8, 9: motor vehicle exhaust
3-10, 9: replace the word nature with characterization
3-11, Fig 3-2 caption: add 'anthropogenic' to this caption to clarify that natural sources are not
included.
 3-23, 25: the phrase 'increased detection limits'  is ambiguous,  'lowered detection limits'
would be preferable.
3-25, 17: reword: 'electrical mobility diameter is often used to
3-28, 4: replace the word built with intended
3-28, 13: very -> every
3-32: Tables 3-3 and 3-4 are identical, despite the table titles.  It looks like the PM10 data was
mistakenly put in the PM2.5 table.
3-37, 7: What table is this referring to?
3-77, 9: change 'visually' to noticeably
3-41, 2: change 'observations' to sites. Most of the observations are 1 hr, but most of the sites
are 24-hr.
3-54, 11: the nitrate fraction in Chicago and Detroit is larger than in Denver.
3-55, 4: nitrate is elevated from fall through spring, rather than spring through fall.
3-57, 8: the bacteria again - still unclear why they are expected in the coarse fraction.
3-66, 1-4: Should also mention that monitor placement plays a very large role in these
correlations - different states/cities have different priorities for monitoring (some focused more
on particular sources or industries, some more  on  background or population oriented monitoring)
so some of the variability will surely be due simply to monitor  siting  decisions.
3-72: This figure is Philadelphia, but it should be Pittsburgh.
3-77, 20: replace 'sensible' with expected
3-81, 7: delete 'the' at end of this line
3-82, 10: Population density seems what is meant here, rather than neighborhood density
3-85, 9: Despite the declaration here that 'it is clear...', it really isn't. Perhaps this is because
the figure is supposed to be replaced?
3-50, 8-10:  does the increase in PM number emissions necessarily translate to equivalent
increases in  mass emissions?
3-100, 3-101:  Figures 3-54 and 3-55 are difficult  to read. Box plots would be less cluttered.
3-103, 24: At last a reference to specific tables in Annex A, but without the appropriate
numbers.
3-104: The  discussion of profiles for receptor models is a bit long-winded and could be
shortened. The point to make is that  no source profile will perfectly represent emissions from a
                                            56

-------
source category. This inherent variability in profiles is just a fact of life we need to live with, not
usually a significant impediment to extracting useful information from a model.
3-105, 12: PMF is not only for time series data, but whatever data it is applied to must contain
many observations.
3-106, 9-11: It's not strictly true that all mass can't be apportioned. CMB works fine as long as
all the major sources are included. And there are ways to incorporate secondary sources as well.
3-107, 3-5: This sentence is murky; should be clarified.
3-110, 1: prime should be primary or principal
3-111, 1: categories as presented above?  Where, exactly?
3-112, 9: delete 'transport from'
3-113: this discussion of emission inventories (-lines 29-34) should also mention the
importance or advantages of using link-based networks for vehicle emissions (here or on the next
page under subregional scale)
3-114: Its not clear whether this discussion of subregional scale models is referring to stand-
alone models, or the incorporation of subregional models into CTMs. Might mention nested
grids, Plume-in-grid, etc.
3-115, 24: mechanisms
3-116, 5: needs a colon after components
3-116, 23: MODIS got mangled - Moderate Resolution Imaging Spectroradiometer
3-119, 6: shouldn't aromatic be anthropogenic?
3-119, 9: particularized -> particulate
3-120, 4-5: reword : predicted also by CMAQ (otherwise it is ambiguous). Also, change
pervasively to perversely or notoriously, and change CASTNET nodes to CASTNET sites.
3-122, 3: The figure numbers are wrong here - should be 3-64 through 3-66 and 3-67 through 3-
69.
3-122, 8: I don't understand the conclusion that the Voyageurs site was not affected by wild
fires,  since measured concentrations increased. Please explain?
3-128: I wish these tables could be converted to graphics and a few summary statements
provided, for example, 'PRB averages 5-10% of annual average mass at eastern sites and 15-
30% at western sites'.. .It's difficult to extract regional differences and similarities from the
tables.
3-132: Table 3-27 is confusing; it's not clear what E and A and C are, and whether r is the same
asR
3-134, 22: is DE diesel exhaust?  Spell out?
3-134, 28: have-> has
3-135: The discussion here of errors associated with monitor height was nice, especially in light
of prior CAS AC discussions on this issue. Good to have some quantification.
3-148, 9: concentrations
3-169, line 2: Chicago and Detroit both have larger % nitrate than Denver.  Might also note that
the 'crustal' can include contributions from industrial and construction activities (eg., in Detroit),
so isn't just 'soil', which implies some benign, nonanthropogenic origin.
3-171, sec. 3.8.1.5:  These NEI inventory estimates don't really agree with the source
apportionment information derived from the ambient data.  The SA models and studies show
little or no impact from primary emissions of EGUs, but very large contributions from secondary
emissions.  These paragraphs seem to overemphasize the role of primary pm25 and  underplay
                                           57

-------
the role of secondary formation.  Could be more balanced, perhaps by giving a quantification of
primary vs. secondary. Sec. 3.8.1.6 gives a more realistic picture.

Chapter 9:
While I like starting a document with an executive summary, I think beginning this chapter with
the summary and conclusions is pushing the concept too far. Would prefer the chapter summary
and conclustions at the end of the chapter rather than the beginning.  ( for one thing, it is then
easier to determine whether the conclusions are appropriate based on the rest of the chapter's
contents).
Nighttime visibility is obviously  something that deserves more attention from EPA, perhaps
through additional research efforts and financial support.
The chapter in general overemphasizes visibility in rural areas, and especially western rural
areas. I would have expected, given the plans for an urban visibility impact assessment, that the
authors might have chosen to add emphasis to urban visibility instead, especially given the
wealth of particle data now available for urban areas that can be easily analyzed  for light
extinction and visibility impacts. Likewise, because visibility impacts are driven primarily by
PM2.5, and PM2.5 is highest in the eastern US, one would expect the chapter to  reflect a more
nationally balanced perspective.
Some of the data presented seemed oddly out of date, or at least could have been easily updated,
on par with the more current data in Chapter 3. Rather than discussing 10-year trends in
IMPROVE data for 1995-2004 and 1988-1999 (Sec. 9.3.4.4), why don't we have trends for
1988-2007? The differences in trends on best and worst haze days seems to merit much more
discussion - why are best conditions apparently improving, but worst days (at most western
sites) are not? The explanation on p. 9-44 is too brief and the reference to trends varying
depending on the length of the period selected is odd. Shouldn't trends become more stable as
time increases? If they don't, because of meteorological or other factors (emissions changes), it
would be  nice to see an example  of that, supported by some evidence of those influences.
Figures 9-8 through 9-11 should be redrawn with consistent color scales within the group of
maps being compared. The point of these is to make quantitative comparisons between seasons,
and the lack of consistent scale makes this vastly more difficult.  One might correctly infer from
Fig. 9-9 that ammonium nitrate concentrations are lower in the third quarter, but the importance
of how much lower they really are is completely lost.
Likewise, figures 9-12 and 9-13 should have a consistent color scale. Or more appropriately if
you really want to show the urban-rural difference, calculate it and plot that, rather than
expecting the reader to do  it on the fly from these 2 plots.  Section 9.3.4.3 in general did a poor
job of presenting urban rural differences.  Figure 9-24 is the only one worth keeping.
Section 9.3.4.5 should be shortened. Here again, there is way too much detail on western sites -
most of this could be shortened drastically. The most useful of it was the urban-rural contrast
discussed in the Schichtel  study.  The results from the weighted emissions tool don't really
belong in  the ISA, especially not the projections to 2018.  They don't add significantly more to
the discussion than the coverage of emissions in Chapter  3. The information on  sensitivity to
changes in precursors perhaps should be discussed in a separate section, or shortened/condensed
to some statements (or even better, a map or two) that show which parts or the country are
ammonia  limited and which are nitrate limited. Even that might be better said in Chapter 3 since
it is not strictly visibility information.
                                           58

-------
Section 9.3.5:  Don't forget the value that people place on views from and over water, and views
of distant skylines (even in the flat Midwest, we value our skyline, and the views of Chicago and
Milwaukee from Lake Michigan are fantastic). It's not just about mountains.
Chap 9 in general - lots of typos, subject-verb agreement problems, word use issues, inconsistent
capitalization,  especially of geographic areas.  Needs careful proofing. Dominate ->dominant,
predominate ->predominant
Could we please discuss nitrate in the Midwest without referring to it as a 'bulge'? Nitrate in the
Midwest isn't inherently any different from nitrate in California - they both arise as a result of
abundant ammonia and NOx emissions.
Specific comments, given by Page, Line Number:
9-4,  17: were -> where
9-46, 6: there is no significant amount of nitrate in the Midwest in summer.  And dominate
should be dominant, here and almost everywhere else it has been used (well, after about page .
9-69, 22: where-> were
9-75, 29: effecting -> affecting
9-83, 29: gas-phase
9-84, lines 9-11: is this statement really supported by the data shown in Chapter 8? It seems
highly geographically variable, with many eastern sites having more mass in the fine fraction
than coarse, and overall about equal mass in each fraction.
9-157, 4: shouldn't this be W/m2/kg/yr? (missing first / )
                                           59

-------
Comments from Dr. Mort Lippmann
       The draft PM ISA provides a reasonably thorough and balanced presentation of a
voluminous and complex literature on ambient air PM and its health and environmental effects.
The summations and judgments made therein are, for the most part, reasonable and well focused
on the needs of the Agency for its mandated task of periodic reconsideration of the suite of PM
NAAQS.

       The most serious failing is the decision to organize the discussion as though PMi0 was a
separate pollutant from PM2.5 and PMio-2.s. It should be obvious that PMio is a mongrel mixture
of the other two, and was only retained because of the limited  database for the exposures to, and
effects of, PMio-2.5 and the need to retain some control of the PMio-2.s concentrations in the name
of public health protection. At this point, the only reasonable way to remedy this problem in the
next draft is to describe what is known about the makeup of PMio in the exposure  atmosphere in
any description of an association  of measured PMio with a health-related effect. In some eastern
US cities, where PM2.5 is 90% of PMio, there will be little reason to implicate coarse particles in
any association. By contrast, in some western US cities, where coarse PM accounts for 50% or
more of PMio, effects associated with PMio are more likely to  be due to coarse PM. The reader
can then decide whether the effect associated with PMio is likely due to the fine or course
components.

       Other unfortunate decisions in the organization of the toxicology presentations were to
include: 1) descriptions of the effects produced by exposures to diluted engine exhausts and other
lab-generated mixtures in with the discussions of the effects of ambient air PM; and 2)
descriptions of the effects produced by IT exposures to high doses with those of inhalation
exposures at more realistic exposures. These more artificial exposures should be described, but
in separate sections, with some discussion of what useful information they can provide.

       My comments specific to  each chapter of the draft ISA follow.

Chapter 1:

Overall: Very well done.

Specific Comments:
p. 1 -14, line 23      change "collected" to "retained".
p. 1-25 , line  13      what is "edpolicy"?

Chapter 2:

Overall: Needs clarification and specific corrections.

Specific Comments:
p. 2-5, lines 13,22   what is "diel"?
                                           60

-------
p. 2-11, line 21,22    Change to "(i.e., cardiovascular morbidity and mortality, respiratory
morbidity and mortality)". It is exceedingly strange to omit cardiovascular mortality, especially
for chronic exposure, where it is the biggest health impact of them all!
pp. 2-12 - 2-14: Section 2.3.1. As noted above in my first Overall Comment, it is very
unfortunate and confusing to talk about effects of PMio without at least discussing what is known
about the size distribution of the particular PMio mixture in each case. It would also be helpful to
include any information on PM composition whenever it is known.
P. 2-20, lines 1-9     This brief summary of long-term exposure and excess mortality should be
expanded to note that almost all of the excess is cardiovascular (Pope et al. 2004).
P. 2-22, Table 2-1.    The Oil Combustion section should include the epidemiological studies
cited in Lippmann et al. (2006) and Hedley et al. (2002).

Chapter 3:

Overall: Provides a good summary review of important background material.

Specific Comment:
p. 3-8, line 9  insert "emissions" after "vehicle".

Chapter 4:

Overall: Provides a good summary review of important background material.

Specific Comments:
p. 4-5, line 15       change "particulates" to  "particles".
p. 4-7, line
p. 4-27, line 26       change "filter" to "filtered".
p. 4-36, line 15       fix reference citation.

Chapter 5:

Overall: Provides a good summary review of important background material.

Specific Comments: None

Chapter 6:

Overall Comment: Provides a reasonably good summary review of important material, but needs
some cleanup for careless errors.

Specific Comments:
p. 6-3, line 13       change "affect" to "effect".
p. 6-3, line 31       change "is" to "are".
p. 6-4, line 23       change "linear" to "log-linear".
p. 6-5, line 2         change "the" to " an enriched".
p. 6-5, line 23       change "cyclone" to "centrifuge".
                                           61

-------
p. 6-5, line 29        insert "and modified by" before "Maciejczyk".
p. 6-6, line 16        insert "particles" after "exhaust".
p. 6-6, line 24        change "particulates" to "particles".
p. 6-6, line 25        insert influence of the" before "co-pollutants".
p. 6-19, line 7        change "Tuxedo, NY" to "Manhattan".
p. 6-21, line 11       This discussion of short-term changes in HR and HRV needs to be
expanded to include the short-term changes reported by Chen and Hwang et al. (2005),
p. 6-197, line 30     the paper by Sama et al. (Sama, P., Long, T.C., Hester, S., Tajuba, J.,
       Parker, J., Chen, L-C., Veronesi, B. (2007). The cellular and genomic response of an
       immortalized microglia cell line (BV2) to concentrated ambient particulate matter. Inhal.
       Toxicol. 19:1079-1087).
p. 6-247, Table 6-16  the Lippmann et al. reference at the beginning should be "2006", not
       "2000".

Chapter 7:

Overall: Provides a less than adequate summary review of important material. Did the authors
       run out of time to get it right?

Specific Comments:
p. 7-1, Heading for Section 7.2      change to "Cardiovascular and Systemic Morbidity
       Effects"
p. 7-2, line 24        define "CAC".
p. 7-3, line 9         define "CIMT".
p. 7-3, line 21        define "ABI".
p. 7-3, line 30        define "AAC".
p. 7-4, lines 3-14     Why start off this discussion with a paper that shows essentially no
       response?
p. 7-5, lines 23-29    Why give such prominence to an IT study?
p. 7-6, line 1         the Sun et al. (2005) paper cited at the start of this section, and the Sun et
       al.  (2008) paper are important reports on aspects of NYU's second 6-month subchronic
       inhalation exposure study. The results of the first such study were summarized by
       Lippmann et al. (2005a), and referred to, in part, as 'an other project' on lines 24 and 25
       of this page. The first study should be cited first, and reference to the Sun et al. (2005)
       paper should follow.
p. 7-7, lines 23-24    "despite 85% of the total particle number concentration for PM2.5 being
       comprised of ultrafme PM" is a truly nonsense statement. What were the authors trying to
       say here?
p. 7-7, line 30        How can anybody take seriously a study in which the "control" exposure
       to PMio was 50% of that in the "exposed"?
p. 7-8, lines 7-20     How can anybody take seriously a study in which such a high IT dose was
       administered?
p. 7-9, line 23        The Sun et al. (2008) and Reed et al.  (2004; 2006; 2008) papers were not
       from the "same group". The Sun et al.  paper was based on a study performed at NYU,
       while the Reed et al. papers described work done at Lovelace (LRRI).
p. 7-9, line 24        "HWS" was not defined, [hardwood smoke?]
                                           62

-------
p. 7-10, line 9       "TAT" was not defined.
p. 7-12, lines 4-8     I am mystified that the effect for the preceding 60 days of exposure were
       reported rather than those for the preceding 30 days of exposure. The most interesting
       finding was that the effect increased with time up to 30 days, and then decreased!
p. 7-17, lines 14,15   Isn't a -45% decrease a 45% increase? Please clarify!
p. 7-17, line 21       Isn't this saying that PMi0 exposure was protective? Please clarify!
p. 7-23, line 6-8      The citation to Peters et al. (1999) is no longer informative, and should be
       deleted. This cross-sectional preliminary  analysis does not, in any way, contradict the
       subsequent prospective cohort analyses of the same population by Gauderman et al.
       (2000, 2002, 2004).
p. 7-24, lines 1-3     The Avol et al.  (2001) paper is not adequately presented. Not only were
       the kids moving to cleaner PM areas having greater lung development, but the kids
       moving to dirtier PM areas were having reduced lung development.
p. 7-39, line 15       The Gunnison and Chen (2005) paper was not a preliminary study, but
       was based on the exact same exposures cited in line 11.
p. 7-39, line 17-19   The Maciejczyk et al. (2005) study is a simultaneous (parallel) study of in
       vitro exposures on the same days as the subchronic inhalation exposures, but is not a
       "second" parallel study.
p. 7-104, Fig.7-8     Why weren't the Pope et al. (2002) lung cancer data depicted here?
p. 7-117, lines 4,5    Delete "likely to be". The  authors should have the courage to accept the
       overwhelming epidemiological evidence  of excess annual mortality associated with
       annual average PM25, which is amply supported by the results of a series of subchronic
       PM2.5 exposure studies in animals showing progressive changes in atherosclerotic plaque
       development and progressive deterioration of cardiac function.
p. 7-117, line 11      Change "short-term" to "short- and long-term".

Chapter 8

Overview: Provides a less than adequate summary review of extremely important material. Did
       the authors run out of time to get it right?

Specific Comments
p. 8-2, lines 23-26    The authors list some valid reasons for heterogeneity in estimates between
       cities, but leave out the two most likely to account for the differences, i.e., differences in
       particle size distributions and in PM compositions.
p. 8-7, line  16       Delete "f.
p. 8-8, line 26       The AHSMOG papers should be acknowledged here, with a notation that
       cigarette smoking is not a confounder for this population.
p. 8-17, Section 8.2.4 The clearest cases for the influence of SES are in the annual  mortality
       cohort studies. WHY WEREN'T THEY CITED HERE?
p. 8-19, lines 5,6     The Miller et al. (2007b) paper does not fit the statement, and warrants
       discussion here.

As EPA staff prepares to revise the ISA for PM, I'd like them to consider and, as appropriate, cite some
additional peer reviewed papers that are highly relevant to PM NAAQS consideration. These are:
                                           63

-------
Chen, L.C. and Lippmann, M. Effects of metals within ambient air particulate matter on human
health. Inhal. Toxicol. 21:1-31 (2009).
Peltier, R.E., Hsu, S.i., Lall, R., and Lippmann, M. Residual Oil Combustion: A major source of
airborne nickel in New York City. J. Expos. Sci. Environ. Epidemiol. (in press).
Peltier, R.E., and Lippmann, M. Residual Oil Combustion: 2. Distribution of airborne nickel and
vanadium within New York City. J. Expos. Sci. Environ. Epidemiol. (in press).
Lippmann, M. Semi-Continuous Speciation Analyses for Ambient Air Particulate Matter: An
Urgent Need for Health Effects Studies. J. Expos. Sci. Environ. Epidemiol. (in  press).
                                       64

-------
Comments from Dr. William Malm

Comments on the Visibility Section of Integrated Science Assessment for Particulate Matter,
First External Review Draft

       The purpose of the draft ISA is to identify, evaluate, and summarize scientific
information on the health and welfare effects associated with PM. The ISA is intended to
"accurately reflect the latest scientific knowledge useful in indicating the kind and extent of
identifiable effects on public health which may be expected from the presence of [a] pollutant in
ambient air"

Charge to the CASAC PM Panel

       We ask the Panel to focus on the following questions in their review:

   •   How useful and complete is the scientific evidence presented and summarized in Chapter
       9 regarding the effects of atmospheric PM on the environment, including (a) effects on
       visibility, (b) effects on individual organisms, (c) direct and indirect effects on
       ecosystems, (d) effects on materials, and (e) effects on climate? To what extent do the
       discussions and integration of evidence correctly represent and clearly communicate the
       state of the  science?

The focus of this review is on Chapter 9, regarding the effects of atmospheric PM on visibility.
To what extent do the discussions and integration of evidence correctly represent and clearly
communicate the state of the science?

In general, the section on visibility is well written and by and large reflects the state of the
science. However, I have a few comments and corrections, which are summarized below.

Section 9.2.1 (Summary of effects on visibility):

Page 9.2:  In the second paragraph it is stated that the visibility impairment from each of the
major species can be estimated.  I would use the word "approximated."  Moreover, the reader is
left with the feeling that a certain amount of visibility impairment "is" attributable to each
chemical species, which, even if perfect measurements were made with zero uncertainty, is not
true. The attribution of chemical species to extinction is an ill-defined problem and cannot in
principle be done.  However, with certain approximations and assumptions (internal versus
external mixtures or assumptions concerning mixing  properties of internally mixed aerosols)
meaningful apportionments of chemistry to extinction can and are made. There should be some
discussion of these issues to demonstrate that EPA has a clear and in-depth understanding of all
the issuable scientific questions.

Bottom of page 9.2:  What is the Midwest nitrate bulge?

Page 9.3:  Why do OC and EC have the highest extinction  efficiencies - not because of size or
index of refraction but largely because of density differences. Some discussion should be
                                           65

-------
included.  Biomass burning is referred to as wild and prescribed fire; there is no mention of
agricultural fires.

What is urban excess carbonaceous material?

Comment on how much of the contemporary carbon may be associated with SO A.

Page 9.4:  Some additional discussion and emphasis on "sense of well being" studies as they
relate to urban visibility.  Topics are mentioned on page 9.75 but implications of these studies
should be highlighted in the summary section and expanded upon in section 9.3.5.

Page 9.8:  Again, agricultural burning is left out of the discussion. It is a clear anthropogenic
source of carbonaceous material.

Page 9.10: Nice discussion of night sky visibility.  The sentence, "...addition of alight into the
sight path, the brightness of the night sky...," seems to me to be a bit unclear.  What is meant by
adding a light to a sight path? And why would this be important?

The extinction of starlight is a result of both scattering and absorption.

Page 9.13: Mie theory calculations do not address "particle  shape".  Usually, spherical aerosols
are assumed; however, more complicated shapes such as cylinders have been used.  However,
arbitrary shapes are not assumed because calculations are nearly impossible to carry out.

Page 9.15: Replace the reference of Sisler et al. (1996) with Malm et al. (1994).

Page 9.16: Particle size is the most influential property of aerosol with respect to their dry light
extinction efficiency.  This generally a true statement; however, if one has particles within the
accumulation mode, the biggest difference between scattering by, say, sulfate and organics is not
generally size but particle density. Ammonium nitrate has a density of 2.3, POM a  density of
1.4; that is about a 60% difference in mass scattering efficiency, just due to density!

Page 9.17: Add Malm et al.  (2003) to the reference list.

Page 9.19: This figure and associated implications should be incorporated in the monitoring
discussion in Chapter 2.

Page 9.20: At the top of page should be 3.9±1.5 m2/g.

Page 9.21: Here is the place where a discussion of internal versus external mixtures should be
included.  The improve algorithm is only an approximation of what is usually an internally
mixed aerosol. Furthermore, the main difference between inorganic and organic dry efficiencies
is due to assumptions of particle density.
                                           66

-------
Pages 9.25-9.30: Great discussion of spatial trends.  However, the scales are all different, which
leads to very misleading interpretations. The scales for each species, at a minimum, between
quarters for each species should be the same.

Pages 9.30-9.40: Same issue for comparisons between STN and IMPROVE.

Page 9.35: Differences between STN and IMPROVE samplers should be discussed. STN uses 5
different samplers (Met One, URG, two different R&P instruments,  and Anderson), all with
different filter sizes and face velocities.  These differences result in significantly different
monitored OC values for the same ambient OC levels because of volatilization issues associated
with SVOCs. Furthermore, STN values, as reported by EPA, are not corrected for positive
artifacts effects.  How were these issues addressed in the presentations of POM data in Figures
9.19 and 9.20 and in any other figures where STN data are discussed?

Page 9.39: In Figure 9.24 where urban excess is discussed (comparison of IMPROVE to STN),
the above issues are critical. Urban excess is interpreted as having its cause associated with
emissions, whereas some of the difference may be due to sampling artifacts (blank issues, in
particular).

Page 9.41: Trends  in haze are very dependent on POM, especially in the West. The episodic
nature of wildfire makes it very hard to infer trends of haze.  Some discussion of the role of POM
in long-term trends especially as it relates to fire should be included here. It is especially critical
in the northern Rocky Mountains and the southeastern United States where there are significant
emissions  associated with prescribed fire.

Page 9.47: Throughout all sections on attribution it would be helpful to have some estimate of
uncertainties. I would guess that most of this discussion  is semi quantitative at best (bottom of
page); certainly, there isn't any difference between 45% and 50%, as is implied in the write up.

Page 9.49: Some cumbersome wording:  "However, the  Gorge's ..."  Consider rewriting. What
is the implication of this sentence?  So what?

Figure 9.32:  On this graph and all others like it, it would be helpful  to have labels on each of the
pies  identifying the pertinent receptor site.

Page 9.51: "track various the organic..." Something is missing.

Page 9.52: It is recognized that current regional models underestimate SOA by significant
amounts (see Robinson et al., 2007) for a variety of reasons. One is that source profiles
(emission estimates) of SVOCs are severely underestimated, and two, the chemistry in the
chemical transport  models is very much lacking. Some discussion of these potentially severe
model limitations should be presented before modeling results are discussed.

Page 9.57, second paragraph: It should be clearly stated  that the estimate discussed here for
SOA is clearly a lower bound.
                                           67

-------
There is considerable discussion on the WRAP weighted emissions potential.  This technique
needs to be clearly discussed as to what exactly the calculation is and how it should be
interpreted.  What winds were used in the trajectory calculations? From what heights were the
trajectories started? All the assumptions can have significant effects on the results.  The reader
should be made explicitly aware of the assumptions.

Page 9.71:  The fact that sulfate explains 90%  of the variance in scattering when sulfate only
contributes 64% of the extinction means what? I don't think it means anything more than there
are collinearities between sulfate and other species.  What is the point of Figure 9.50? The
legend for Figure 9.50 is wrong or the graph legend is wrong. I think the legend should read
"... particulate nitrate and organic mass..." instead of "... particulate nitrate..."

Page 9.72, bottom of page:  "... sulfate concentrations than to changes in nitric acid
concentrations..."  Not sure what the implications of this statement are. Nitric acid is a
precursor to particle formation while sulfate is a secondary aerosol. Would it be more
meaningful to talk about SO2 and nitric acid where SO2 is the precursor to sulfate aerosol.

Page 9.75:  I think the whole discussion on the bottom of page 9.75 should be expanded to
present some conclusions of all the studied referenced here, maybe one more paragraph.  I think
these results are absolutely key to assessing the relevancy of a secondary visibility standard.

Pages 9.78 and 9.79: There are some conversions between variable issues here. For instance,
the point is made that all three urban studies have about 20 dv as being acceptable visibility.  Yet
it is stated that 0.76/km is the median extinction value for Denver's acceptable visibility.  Well,
0.76/km is something like 40 dv, not 20. Same issue with the Phoenix study - certainly 45 km
and 3.5 km do not correspond to 87 and 12 km visual ranges. The implication that about 20 dv
impairment is acceptable at all three locations  may not be true.

Page 9.83:  $134 - $360/what? - day? year?

REFERENCES

Malm, W. C., Sisler, J. F., Huffman, D., Eldred, R. A., and Cahill, T.  A.  1994. Spatial &
seasonal trends in particle concentration & optical extinction in the United States. Journal of
Geophysical Research, 99, 1347-1370.

Malm, W. C., Day, D. E., Kreidenweis, S. M., Collett, J. L., and Lee,  T.  5-10-2003.  Humidity
dependent optical properties of fine particles during the Big Bend Regional Aerosol and
Visibility Observational study (BRAVO). Journal of Geophysical Research-Atmospheres, 108,
art. no. 4279.

Robinson, A. L., Donahue, N. M., Shrivastava, M. K., Weitkamp, E. A., Sage, A. M., Grieshop,
A. P., Lane, T. E., Pierce, J. R., and Pandis, S.  N.  2007.  Rethinking organic  aerosols:
Semivolatile emissions and photochemical aging. Science, 315, 1259-1262.
                                           68

-------
Comments from Dr. Kent Pinkerton
Charge Question #5:

Chapter 5 is intended to support the evaluation of health effects evidence for both short-term and
long-term exposures to PM.  Some potential modes of action may underlie a number of health
outcomes and may contribute to health effects of both short- and long-term exposures. Thus, the
potential modes of action are described briefly in Chapter 5, and some specific study findings are
discussed in more detail in the relevant sections of Chapters 6 or 7.  What are views of the Panel
on this approach and on the characterization of potential modes of action for PM-related effects
in Chapter 5?

Overall impression:

The Integrated Science Assessment for Particulate Matter: First External Review Document is an
excellent document. Chapter 5 provides to the reader a highly relevant summary of the plausible
mechanisms by which inhaled particulate matter could lead to adverse health effects.  The
chapter is well-written and is presented in a highly succinct and clear fashion. I like the
approach used by the authors to present further details in chapters 6 and 7 which summarize past
information available from the 2004 document and cover the latest literature for each topic under
Chapter 5 in a clear and very complete fashion.

Additional Comments:

Possible pathways/modes of action, chapter 5 focuses on a myriad of pulmonary effects via a
thorough coverage of critical cellular and biological parameters to encompass reactive oxygen
species, cell signaling pathways, inflammation, epithelial barrier function, antioxidant defenses
and adaptive responses. Measures of pulmonary function, allergic disorders, impaired lung
defense mechanisms, resolution of inflammation with progression of disease and pulmonary
DNA damage are also included in plausible mechanisms that are outlined in a highly succinct
fashion.  Systemic  effects of PM are summarized to encompass endothelial dysfunction, altered
vasoreactivity and  activation of coagulation, as well as acute phase responses.  The role of the
autonomic nervous system by pulmonary reflexes is discussed. Translocation of ultrafine PM
and soluble PM components is presented, along with  new inhalation studies which contribute to
modes of action.

Chapter 6 presents the integrated health effects of short-term exposure.

Chapter 7 presents the integrated health effects of long-term exposure.   Recent epidemiologic
studies confirm and provide further support between long-term exposure and increased risk of
mortality in areas with mean concentrations from 14 to 29 |ig/m3.
Specific Comments:
                                           69

-------
Page 5-1. The potential pathways or modes of action described below do not appear to be
specific to a particular size class of PM. This statement is perhaps too strong and needs
to be used with caution based on the growing evidence the ultrafme PM effects. Yes, the mode
of action could eventually be the same, but the degree to which these effects are manifested
could be strikingly different.

Sections in the chapter are clearly and accurately prepared. In most instances, excellent
references are provided. However, some sections lack references and/or assume similar
pathways of action based on the presence of a particular mediator that may not have been
documented with PM-based studies.

Page 5-4. PAH-induced changes suggested only documentation in mice. This has been
documented in other species as well.

Each section provides a clear summary of pulmonary effects. However, it is important to
establish whether these effects are similar across species and under what conditions. This could
be true for paths  of activation, species sensitivity and/or tolerance, gender-based differences in
response.

Page 5-7.  Pulmonary function - has AHR be defined as  airway hyper-reactivity?

Chapter diagrams are helpful. Some could be further clarified as to meaning or association. In
addition, summary or outcome of study would be clearly helpful.

Page 5-9.  DNA  damage. The final sentence of this section seems to be more speculative than
confirmed by direct studies. DNA repair mechanisms may also influence genotoxic, mutagenic
or carcinogenic potential of DNA damage.  Is this statement true for particle-induced DNA
damage?

Page 15-13. Have acute phase responses been shown to be PM-induced?  The authors should
cite or reference  such studies. Cytokines such as il-6 stimulate the liver to produce acute phase
proteins including C-reactive protein and antiproteases. (van Eden et al. 2001).

Page 5-15.  Section 5.5 Disease of the cardiovascular and other organ systems.  This section
could benefit with further discussion of those plausible mechanisms that could be attributed to
particles. It is important to make clear whether particles lead to cardiovascular disease or are
simply able to perturb existing cardiovascular disease.

Page 5-15.  CAPs - defined, BC - defined?

Page 5-15 to 5-17.  New studies. The summary of new findings is extremely good. It might be
nice to reference these new findings as well, even though it is understood that these are presented
in detail in chapters 6 and 7.

Page 5-17.  Gaps in knowledge.  A nice addition to the chapter. Other gaps not stated  could also
be listed:
                                           70

-------
   a)  measured effects due to timing of exposure and timing postexposure
   b)  the spatial distribution of retained particles and its impact
   c)  deposition, uptake and clearance of ultrafine particles.

Chapters 6 and 7:

Page 6-3: Nice presentation provided for the importance of lag times in the consideration of
health impacts.

Pages 6-4 to 6-6: The history of concentrator systems used in experimental studies is critical in
placing into perspective how we can use these findings from both animal and human studies to
evaluate PM health effects.  The description of advantages and limitations of concentrator
systems is very helpful.

The use of tables throughout the chapter is very helpful.  However, for Table 6-1 it would be
beneficial to add an additional column to list a summary or outcome for each study as done in
Table 6-2.

Page 6-19 to 6-21: The toxicology studies reported are very reasonable, but are newer studies
also available?

Page 6-29, line 15: Is there a reference for heart beat frequency in mice and rats?

Page 6-35 to 6-36: The section on vasomotor function clearly points out the importance of the
endothelium and endothelial dysfunction.

Page 6-46 to 6-49: The sections on epidemiologic studies and right ventricular pressure would
benefit to have simple summary statements at the end  of each section.  An example of such a
summary statement is found on Page 6-66.

Page 6-95: Cardiac Contractility - Toxicological studies: Reference needed.

Page 6-96: Systemic Inflammation - Toxicological studies: Reference needed.

Page 6-100: Section on Toxicological studies is somewhat confusing and lacks references.

Page 6-103: Section on Toxicological studies lacks references.

Page 6-108: Blood coagulation - Toxicological studies: Reference needed.

Page 6-156 to 6-159: Allergic sensitization and methylation studies are fascinating.

Excellent summaries are provided throughout chapters 6 and 7.  Although the data is very
complete and typically detailed, the organization of each chapter is easy to follow, especially in
being able to  seek out specific information on any given topic.
                                           71

-------
One random consideration:  I really like the strong, well-documented respiratory components of
the Integrated Science Assessment document. However, I wonder if the autonomic control
sections might be further expanded via the consideration of central neural control of the
autonomic system through further discussion of the mid-brain circuit connections?
                                           72

-------
Comments from Dr. Robert Phalen

Chapter 1.  Section 1.5 EPA Framework for Causal Determination.

This section presents "a consistent and transparent basis to evaluate the causal nature of air
pollution-induced health or environmental effects" (p. 1-11).  However, there are some
weaknesses. My main concern is that causality might be accepted when it is not true.

    1.     In section 1.5.2 the statement "An association is prima facie evidence for causation:
          alone, however..." does not recognize that the strength of an association is also a
          factor that should be considered.
    2.     In section 1.5.3, page 1-16, lines 4 to 20 discusses the difficulty in dealing with an
          extraneous factor  and in finding "the likely causal pollutant" in mixtures of
          pollutants.  This issue may be a profound one for PM2 5, given that evidence for risks
          for mortality upon long term exposure are found in "eastern and central regions, but
          not in the western United States" (Zeger et al., Env. Health Persp. 116 (12): 1614-
          1619, 2008).  Thus, the likely causal pollutant is not PM2.5. Should section  1.5.3
          include a requirement that the indicator be valid at all locations before accepting
          causality?
    3.     Based on comment #2, should Table 1-2 be modified?  Specifically with respect to
          "Consistency of the observed  association", should consistency among various
          geographical regions be added as an aspect for judging causality?
    4.     Also in Table 1-2 under "strength of the observed association" the entry concludes
          with the statement that "while large effects support causality, modest effects therefore
          do not preclude it".  It might be good to comment on small effects also, or at least to
          define how much  of an effect  is considered to be "large" and "modest". Are any
          strengths too small to be used to support causality?


Chapter 3.  Source to Human Exposure

The chapter adequately updates information related to human exposures to PM.  There are
several undefined acronyms.

    1.  pg. 3-2, lines 23 &  31: Two main sources of PM2.5 in the East are initially presented, but
       the rest of the paragraph doesn't smoothly follow.  Perhaps an introductory sentence
       could explain what this paragraph is about.
    2.  Table 3.1: Under the column "Accumulation", some new sources could be added, e.g.
       wind-blown dust and  biological processes.
    3.  Figure 3.2: "EGUs" should be defined in the caption or on the figure.
    4.  Section 3.6: Policy Relevant Background might also include excessive levels of
       resuspended dust on windy days.  For example very strong dry wind periods in  Southern
       California are probably capable of producing PM2.5 exceedences.  If any data on such
       conditions are available they should be mentioned. This natural problem could lead to
       excessive regulation of anthropogenic sources of PM2.5, which can have serious adverse
       economic consequences.
                                           73

-------
   5.  pg. 3-172, lines 5 to 7:  Fugitive dust also contributes to PM2.5 in the West. This should
       be added to the paragraph.
Chapter 4.  Dosimetry

In general the chapter presents an excellent review of the current understanding of inhaled
particle dosimetry. The suggestions for improvement are minor.

    1.     pg. 4-5, lines 16 & 17: Add "and individual genetic variation", as it is a significant
          factor in airway morphology.
    2.     Figures 4-3 and 4-4: In the caption, state whether or not an Inhalability correction was
          used.
    3.     Figure 4-5 could be eliminated. The non-natural breathing diminishes its utility, and
          section 4.2.4.2 treats gender more thoroughly (eg see pg. 4-16, lines 8 to 13. Also,
          see the admonition on pg. 4-17, lines 26-28.
    4.     pg. 4-11, line 21: Replace "lower respiratory tract", which is not clearly defined, with
          "tracheobronchial and possibly alveolar regions".  This clarification is helpful in view
          of the predicted and measured increased TB efficiency in small children, which can
          sometimes decrease alveolar doses.
    5.     Section 4.2.5: This summary could be eliminated.  The preceding sections are
          sufficient, and the summary has some  over simplified conclusions.  There are several
          problems in the Summary, e.g. Inhalability in humans is 77% for a dae, not "near
          100%" as stated, the carinal enhancements look wrong, etc.
    6.     Sections 4.3.1,  4.3.1.1 and 4.3.1.2: Two major important factors modifying particle
          clearance  should be added; 1) respiratory tract infections and 2) cough. These factors
          are strong and widely occur in the population at any given time.
    7.     pg. 4-23, lines 4 and 5: The last sentence should be dropped. It conflicts with other
          studies, and the previously-mentioned dog data were for 2um particles, while the rat
          data were for 15-20nm and 80nm particles given to intubated ventilated rats.
    8.     Section 4.3.4.3: The effect of respiratory tract infection on particle clearance is very
          important and should be added to this  section.
    9.     Section 4.3.5:  This summary could be dropped. It is not an adequate summary and it
          is a bit misleading regarding differences between rats and humans.
    10.    Section 4.4, pgs 4-32 to 4-36:  The terminology is poor.  The section begins with
          "Soluble particles", which should be replaced with something like "Particles that
          dissolve rapidly after deposition...". Good terminology must be established and used
          for this section and  also in Chapter 5.  The concept of water-solubility doesn't work
          in the respiratory tract. Many substances that are very poorly soluble in pure water
          rapidly dissolve in the lungs when finely divided.
    11.    pg. 4-33, lines 20 to 22: This sentence didn't make logical sense, it should be fixed or
          dropped.
    12.    Section 4.4.3:  Again, the Summary is very poor. It should be dropped, as it detracts
          form the quality of the foregoing chapter. Many other chapters do not have
          summaries, so dropping this one  should not be a problem.
                                            74

-------
Chapter 5.  Possible Pathways/Modes of Action

    1.  The word "soluble" is used throughought, see comment 10. above.
    2.  Section 5.6:  It would be helpful to identify the study populations (e.g. human, rat etc.)
       for each of the effects listed.

Chapter 6.  Integrated Health Effects of Short-Term PM Exposure

1.  Throughout this chapter and Chapter 7, particle measurements are described as "increases"
rather than  "levels".  Increases in levels are not the same as actual levels themselves. Yet, the
epidemiology associations based on increases seem to be used by the EPA to set permissible
levels of PM. A given increase can occur independently the level of PM.  The issue of
considering epidemiology associations with "increases" rather than levels should be discussed
and justified. I recall a pulmonary physician commenting on this issue during a proposal review
session. He indicated that the respiratory tract susceptibility to injury was very changeable and
that a few days of exposure to clean air would make it more vulnerable. Perhaps regulating
short-term "changes" or "increments" in PM measures (count, mass etc.) would be as good (or
better) than regulating low levels.  In any case EPA should address the PM "increases" issue.
E.g. see pg. 6-10, lines 8 and 9; pg. 6-11, lines 24 and 27; and many places elsewhere.  The
possible importance of this issue might be seen in section 6.2.10.8 where the results of
Villeneuve (2006) differed from the other study results.  Also, conclusions based on increases,
such as that on pg. 6-94 (and other pages) should probably indicate that the conclusions may not
apply to absolute levels.
Charge Question 4.

Detail in Chapter 4 is sufficient. Particle translocation is adequately described. The roles of
cough and respiratory tract infections on particle clearance deserves greater emphasis. The
summaries should be made into true summaries of the foregoing material.  Currently, they
introduce some new material. Annex B is ok.
                                           75

-------
Comments from Mr. Rich Poirot

Comments on Chapter 9: Ecosystem and Welfare Effects

General Comments
This chapter summarizes the policy-relevant science that would support review of and possible
revision to the secondary (welfare-based) PM NAAQS.  It includes sections on PM effects on
visibility, individual organisms, ecosystems, materials, and climate.  The section on visibility is
substantially more detailed and informative than those on other welfare effects - about the length
of the other 4 effects sections combined.  I think this emphasis is justified, since it seems
unlikely that a secondary PM NAAQS could be an effective mechanism to ameliorate affects of
PM on organisms, ecosystems, materials or climate. However the relationships between PM
mass (especially PM2.s) and/or PM species and visual  air quality have been well known for many
decades, and a secondary PM NAAQS to protect visibility has been considered in every previous
PM NAAQS review - including the original establishment of TSP standards in!971, when  a
secondary TSP standard of 150 ug/m3 was based in part on (minimal) protection of aviation
safety (maintaining visual flight rule conditions at airports).

The visibility section of this chapter is clearly written  and presents a substantial amount of recent
information, much of which has been generated through the work of the Regional Planning
Organizations (RPOs), and supporting groups at NFS, EPA, and the IMPROVE program, to help
develop the first round of State Implementation Plan (SIP) revisions in response to the 1999 EPA
Regional Haze Rule. Consequently, the focus is primarily on effects of individual fine particle
species (and coarse particle mass) at remote National Parks and Wilderness Areas.

One general criticism then is that relatively little of the information presented relates directly to
causes and effects of visibility-impairing PM in the non-Class-1 rural, suburban and urban  areas
to which a possible secondary PM NAAQS would apply (although conceivably a secondary PM
NAAQS might be exceeded in some of the more polluted Class 1 areas, and could provide  added
mechanisms to assure improvements in those areas). We would also expect to see improvements
in visibility outside Class I areas as  an inevitable result of measures to reduce regional haze
impacts within them. A secondary PM standard and the Regional Haze Rule need not be
mutually exclusive from each  other, although it seems useful to think of them as complementary
rather than as redundant mechanisms.  So it would be  useful to see more information focused on
the differences between urban/suburban and remote area PM composition and visibility effects.

A second general criticism is that the chapter is predominantly focused on the reconstructed
aerosol extinction approach (chemical species measurements combined with generic dry
extinction efficiencies and hygroscopic growth functions) which is the basis of the regional haze
program. Conceivably, a similar approach might be employed for a secondary PM standard -
employing for example the speciation data from the CSN network with a similar or modified
"urban" haze equation.  However, other approaches might also be considered and the chapter
could do a better job supporting those alternate approaches. For example, in reviewing the PM
Criteria document in the last review cycle, it was noted by CASAC (Warren White and others)
that an early draft of the PM CD contained no information showing the simple but strong
relationship that exists between fine particle mass and visibility (especially under conditions
when effects of hygroscopic growth are minimized - such as by focusing on daytime only and/or
                                           76

-------
constrained to hours of lower relative humidity).  As a result, the following plot (with associated
discussion) was added to the final 2004 PM CD.

     Figure 1. Illustration of Fine Mass / Visibility Relationship from 2004 PM CD
                         1600-

                       £  1400-
»  1200-
c
ra
_^  1000-
3
.2  800-
O)
=  600-
                       £   400-
                      0>   200-
                      co
                                                                4m2/g
             »•   x
             '. .^««
         \«
        X
                                  50    100   150    200

                                           PM3 (M9/m3)
                                                         250
                                                               300
                                                                     350
             Figure 4-37. Proportionality of observed daytime haziness to fine particle mass
                       concentration in Los Angeles. Visual ranges are 8-h averages of hourly
                       human observations, plotted as extinction according to Koschmieder
                       formula. Mass concentrations are from 8-h samples collected behind
                       a cyclone with 3-uin cut point. Relative humidities were <, 70%.
             Source: Chow et al. (2002b).
This graph had earlier been included by Warren White in a JAWMA critical discussion on
visibility by Chow et al. (JAWMA 52:973-999, 2002), and attributed in turn to data collected in
1970-71 and published by Samuels et al. in 1973.  One point Warren raised was that the strong,
linear relationship between fine mass and visibility impairment was well known at the time of
the original Clean Air Act. Another point emphasized in that discussion, was that the
information provided by detailed, costly, labor-intensive chemical speciation sampling (which
currently requires use of intermittent filter sampling over long averaging times) was of value
primarily for estimating the effects of aerosol water.  A more efficient regulatory  approach might
take better advantage of current science and technology, by minimizing the importance of water
by constraining the averaging time to daytime-only and/or hours with RH < 70%  - as has been
done in establishing local visibility standards in Denver, Phoenix, Lake Tahoe, etc.

EPA staff and CAS AC took a similar approach in recommending a secondary  PM2.5 standard in
the last review, based on a daylight-only 4 to 8-hour averaging time (also corresponding more
directly to periods when outdoor, sun-illuminated visibility is most important to most people).
An alternative indicator might (if the lawyers allow it) be based more directly  on  optical
measurements, such as by ASOS, transmissometer, nephelometer, or nephelometer +
aethelometer (scattering + adsorption).  To better support consideration of a fine mass or optical
indicators,  it would be useful to include some additional attention in the ISA to discuss the
characteristics, averaging times, strengths and limitations of various aerosol, optical and
meteorological measurement methods; present some illustrative data from urban/suburban areas;
                                             77

-------
and show how effects of water (currently avoided as a component of PM mass), might best be
quantified and/or minimized with alternative regulatory options.

Specific Comments

p. 9-2, line 7:  Add "by particles" after "scattering".

p. 9-2, line 8:  You might rephrase this to "While a large particle scatters more light than a
similarly shaped smaller particle..." or "While a large particles scatter more light than a similar
number of similarly shaped smaller particles..."

p. 9-2, line 19: Add "ed" to "call".

p. 9-3, line 6: Change "and" to "or" or to "and/or".

p. 9-3, line 16: "Western..."

p. 9-4, line 5:  You could add "in rural areas" after "haze".  Note - here and elsewhere in this
section, its not always  clear whether you're referring to urban areas or regional-scale (rural)
concentrations.

p. 9-4, lines 10-20: Need a reference for this study.

p. 9-4, line 21: You could  insert "Perception of or "The value of before "urban visibility"  to
make it clear that you are not referring to urban studies of optics, aerosols, sources, etc.

p. 9-4, line 34 and p. 9-5, line 1: You could delete ", the preferred measure of visibility
impairment", since I don't  think DVs are actually "preferred" by everyone, and also to make it
clear that you don't mean 19 to 25 DV was the preferred measure of impairment....

p. 9-6, line 2:  While "N" is correct, I suspect you mean "Ni".  It might also be noted that Zn is
an element typically most prevalent in the fine fraction, while V often has a relatively high
coarse fraction contribution.

p. 9-7, line 14: Change "is  believed" to "is estimated to".

p. 9-7, line 16 (or elsewhere):  It might be useful to begin with some discussion of EPA staff (&
CASAC) recommendations (& policy decisions) from the last PM review.  This would help lead
to a more "a concise synthesis and evaluation of the most policy-relevant science used to help form  the
scientific foundation for  review of the secondary (welfare-based) NAAQS"

p. 9-7, lines 16 to 18:  Since the "recent" information presented here comes quite predominantly
from regional haze-related  analyses of rural  IMPROVE data, I suggest modifying this sentence
to: In recent years, most visibility research involved characterizing visibility levels and trends over broad
regional scales, improving our understanding of the atmospheric processes and pollutants responsible for
the regional impacts, and attributing visibility-impairing pollutants at remote sites to emission sources,
source types, and regions.
                                             78

-------
p. 9-7, line 25 and first 2 lines on p. 9-8: Break this into a few sentences or otherwise fix the grammar.

p. 9-12, line 14:  Change "are" to "is".
p. 9-12, lines 16-17:  True, although NC>2 measurement data are often not available, are typically
of questionable quality (include other oxidized N species), and may not represent the locations -
such as in NOx plumes and in layered hazes - where NC>2 effects are most evident.

p. 9-14, line 11: You could change "theoretical" to "Mie theory", as less sophisticated
"theoretical" calculations are certainly possible - and are routinely conducted for the regional
haze program.

p. 9-16, line 3: Add "s" to either "term" or "inflate"

p. 9-16, line 21:  Add "s" to "aerosol" or add "particles" or "effects" after "aerosol".

p. 9-18, lines 12-14:  Can you provide some quantitative estimates of the underestimation of total
extinction due to coarse mode hygroscopic species at Brigantine or elsewhere (i.e. is this
important?). Also, if we are concerned with effects of anthropogenic pollutants on extinction, is
there any visibility consequence of replacing NaCl with NaNOs?

p. 9-18, line 29:  Change "are" to "is".

p. 9-20, line 18:  Delete the first "components".

p. 9-20, lines 19-22:  Seems like either too much detail  or too little. You could just report the
values they recommended, but if you also report the different values they "determined" from
their review, then I think additional explanation is needed to justify why they felt compelled to
recommend numbers different from what they determined.

p. 9-21, lines 16-23 (or elsewhere nearby): It might be  informative to comment here  on possible
problems in applying this (remote site) equation to urban data - for example, organic mass might
be < 1.8 * OC, and we might also not expect the higher OM concentrations to be well-aged.

p. 9-25, Figure 9-7 (and about the next 20 map figures): The VIEWS system is an excellent tool
and these color contour maps are very informative. However, a major limitation has been the
lack of user control in selecting the color contour levels. I think the system "auto-scales" these
to display maximum  information content for each map from the (10) colors employed.  This is
fine for seeing spatial pattern in an individual map, but makes it difficult to compare maps.  I
don't think its critical for this ISA, but it would be useful for many future applications if options
were provided for the VIEWS user to specify the contour levels and intervals (i.e. the maximum
contour level, number of contour intervals, and ranges of the bins).

p. 9-27, Figure 9-9, and in text discussion: The Agency has decided that effects of S  and N
deposition (including particle phase) will be addressed in the review of secondary SOx and NOx
NAAQS and not in this PM NAAQS review, while the  secondary aerosol effects of S and N
compounds - such as on visibility - will be addressed in the PM NAAQS review, but not in the
secondary SOx and NOx NAAQS review.  Since ammonium sulfate and ammonium  nitrate have


                                           79

-------
similar dry scattering efficiencies and hygroscopic growth functions and combined account for a
majority of visibility impairment over much of the country, it might have been interesting (& still
would be) to break out a separate section on (for example include a map here of) the visibility
effects of combined S + N aerosols. Such a discussion could be copy-pasted in both the PM and
secondary SOx & NOx ISAs & REAs. A reason for this is that it might help support
consideration of SOx - NOx NAAQS metrics that might have been based on a combined sum of
gaseous and aerosol S and N compounds, that might have had (and been justified by) beneficial
effects of S & N emissions reductions on both deposition-related and aerosol-related effects.

p. 9-35, lines 8-20 (and following figures):  Line 9 indicates that Figures 9-22 and 9-23 show
total carbon, but those figures actually show fine soil (the total carbon figures are missing).
There needs to be some discussion  of the differences in the CSN TOT and IMPROVE TOR
carbon methods - indicating that OC & EC don't have the same meaning at urban & rural sites,
although the missing total carbon comparison would be meaningful.  Possibly the CSN OC & EC
data could be adjusted to make it "IMPROVE-like" before combining it with the IMPROVE data
in Figures 9-19 and 9-21 - otherwise I don't think you should be combining the disparate data.
As indicated before, using the same scales for paired IMPROVE and IMPROVE + CSN maps
would be helpful.  It would also be useful to delete the few urban IMPROVE sites (Washington,
Phoenix, Puget Sound, etc.) from the otherwise rural  IMPROVE-only maps (if possible).

pp.  9-41, 9-43 (Figures 9-26, 9-27): A discomforting feature of the haze trends in these figures is
that the 10-year period 1995 - 2004 is the period of the Title IV Acid Rain SO2 emissions
reductions in the eastern US.  Yet on the cleanest and haziest 20% days respectively, 6 of 14 and
8 of 14 eastern sites show no significant improvement. This doesn't inspire confidence in
successful outcomes of future controls. I suspect this may be partly related to the fact that a large
portion of the acid rain SO2 reductions occurred in the first (1995) year of the Title IV program
(and of this trend analysis) and due to spending of banked allowances etc., the actual emissions
reductions over the 1995-2004 period have been relatively small. See figure below from EPA
"Acid Rain Program 2007 Progress Report" http://www.epa.gov/airmarkets/progress/arp07.html.
Possibly some discussion of this "non-trend" could be added here.

                  Figure 2: S02 Emissions from Acid Rain Program Sources
       20
        is   17.3                   I Phase I (1995-1999) Sources   • All Affected Sources
                16']  15.7         • Phase II (2000 on) Sources    — Allowances Allocated
                             12.5           12.5

                             "II
                                   -7.1 — 7.0	7.0|

                              5.4  5.5   5.3  49
10.6  1Q2  10.6  10.3  io.2
   9.6 • 9.5 • 9.5 • 9.5H 9.5=9.5 — 9.5
         0
           1980  1985  1990  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007

                                            Year
     Source: EPA, 2008


                                          80

-------
p. 9-45, Figure 9-29:  Although there are cautionary remarks in the text on p. 9-44 describing this
figure, as well as a warning that this "may be misleading" in the figure caption, I'm suspicious of
this figure - generated from the "Causes of Haze" website - that shows increasing trends
(significantly so at most sites) of ammonium nitrate extinction on the haziest days at all western
IMPROVE sites over the period 1994-2003. Extraction of a similar plot from the COH website
for the cleanest 20% days shows an opposite trend of decreasing ammonium nitrate extinction on
the cleanest 20% days at these  sites - and I'm just finding this hard to believe,  and  suggest
double-checking the calculations.  If these opposite trends are correct, they would seem to
warrant additional exploration  (for example doing same for central and eastern sites) and
offering some explanation(s) of causality).

Figure 3.  Nitrate Extinction Trends at Western IMPROVE Sites on 20% Haziest Days
(left) and 20% Clearest Days (right), 1994-2003 from Causes of Haze Website  (
http://www.coha.dri.edu)
              COH Nitrate text on 20% Haziest Days
COH Nitrate bext on 20% Clearest Days
  Legend

  WN Slope
   •ft 0.064 - 0.093

   ^ 0.094- 0.170

   fa 0.171 -0.520

   4> 0.521 - 1.509
  WN PValue
   • 0.21 -1.00

     0.11 -0.20
     0.06-0.10

   • 0.00-0.05
                             Legend

                             BNSIope
                              ft  0.011 -0.033

                              §  0.034-0.070

                              •ft  0071 -0.109
                                                                            ft
                             At* 0231 -0439

                             BNPValue
                              •  0.21-1.00
                                OL11-QJO
                                0.06-0-10
                              •  000-0.06
I notice that similar trend information, including estimated extinction calculations on worst 20%
& best 20% days by species, site & year can be extracted from both the VIEWS and (for WRAP
sites for 1994-2003) the Causes of Haze websites. Comparing the 2 separate sets of calculations
(using the RHR1 equation on VIEWS),  I note that the results are generally very similar for most
species at all sites, but there are substantial differences - especially in the first few years - that
seem likely to influence trend calculations, directions and significance levels. I don't know
which one is right (if either are), but note they are different. An example of the data for the
Lassen Volcanoes NP  site is pasted in Figure 4 below.
                                             81

-------
p. 9-45, lines 1-2: I don't understand what "at any relative humidity" means here.  Haze
contributions are only proportional to species concentrations at very low humidity. At higher
RH, the hygroscopic species contribute proportionately more, relative to their mass fractions.
Figure 4. Estimated Light Extinction on Worst 20% Days for sulfates and nitrates at Lassen
   Volcanoes IMPROVE site, 1994-2003, as extracted from the Causes of Haze (COH) and
                                     VIEWS websites
        10

         9

         8

         7 -

         6

         5

         4

         3

         2 -

         1

         0
£,----&
—e—Sulfate_Ext (COH)
- *•- Sulfate_Ext (VIEWS)
-A- Nitrate_Ext (COH)
  D  Nitrate_Ext (VIEWS)
             1994    1995   1996    1997   1998    1999   2000    2001    2002    2003

p. 9-46, lines 6,7: Nitrate does not "dominate" in the summer in the Midwest.

p. 9-48, Figure 9-30 and several subsequent figures (9-32 through 9-36). You might add the time
periods for which these plotted analysis results apply to the figure captions.

p. 9-57, line 24: Add "s" to "location".

pp. 9-57 through 9-64:  I have a hard time following this discussion and interpreting these
figures. A bit more explanation up front would be helpful.  Are all these figures really
necessary? Also, I note that a large majority (22 of 30 pages) of the discussion and illustrations
in this "causes of haze" section 9.3.4.5 pertain to the WRAP RPO region (and to aerosol source
apportionment in WRAP class 1 areas only), while only 7 pages are devoted to results from the
other four RPO regions combined.

p. 9-74, lines 29-31:  While I understand there are some good arguments and strong opinions to
the contrary, I don't concur that the existence of primary health standards necessitates the
separation (& removal) of "wellbeing" components (i.e. perceived health effects), from so-called
                                           82

-------
"aesthetic" components when attempting to assess perceived levels of adversity or to evaluate
alternative levels at which secondary standards might be set.  First, this requires an assumption
that the primary standard has in fact been set at a level below which all health effects are
prevented. With the TSP primary standard of 260 ug/m3, or PMi0  of 150, or PM2 5 of 65 or 35
ug/m3, any observers who perceived health effects at or just below the levels of the primary
standard would have been correct, and any regulators who argued that any perceived health
effects were irrational and should be discounted at those (health-protective) levels were wrong.
Second, there's a required assumption that its possible to separate a perceived health concern
from the overall emotional response to the visual effects of polluted air. I'm not so sure this is
really possible (or necessary), nor do I understand what a purely "aesthetic" response to impaired
visibility would mean. Third, I'll argue that there can be a perceived health effect which is
different from the actual health effect, but which is nonetheless an important source of stress,
malaise, and/or emotional discomfort to the viewer. An analogy would be the strong perception
of "unhealthiness" associated with certain "chemical" odors.  This perception may occur at
ambient concentrations of such chemicals which are not (currently) demonstrated to cause actual
health effects. The perceived health effect unavoidably contributes to the perception of a foul,
objectionable odor, which adversely affects the quality of life for those exposed to it.

pp. 9-76 - 9-98  (or in a subsequent section):  In (or after) summarizing these various urban
visibility valuation studies, it might be useful to summarize the urban visibility standards  that
have resulted in part from some of these studies - or from other public policy decision-making
processes (such as  for California, statewide and for Lake Tahoe), as well as noting, the scientific
and practical reasons for the details of the indicators, averaging times, levels and forms of these
standards  (such as constraints to daytime-only, and to lower RH levels,...).

It can also be noted that almost all of the preceding discussion was based on results taken from
remote, Class 1 areas (which are already "protected" by EPA's Regional Haze Rule.  Then
there's a jump to "urban" valuation and preference studies, but no  mention of such valuation
studies in remote Class 1  areas, nor any discussion of visibility effects or valuation in the  (non-
Class 1, non-urban) suburban and rural areas which compose most (>95%) of the US land area.

p. 9-84, line 2:  Since effects of PM (and its precursors and transformation products) on materials
are also related to deposition, you might also add a mention of that to this intro on deposition. It
might also be appropriate to mention here the PM species for which deposition is of concern for
a materials damage or ecological perspective.

p. 9-84, line 25: you might insert "as well as" after "soil," and before "plant" to make it clear
that your not talking about "base cations from soil, plant and insect fragments,  ..." etc.

p. 9-85, lines 1-7:  This discussion of radiative effects doesn't really belong in the "Forms of
Deposition" section, unless you are referring to alteration of radiative effects from reflective
snow and ice surfaces as they are modified by deposition of light absorbing particles.

p. 9-89, lines 6-15: I'm surprised we need to go to Italy to find references on primary vs.
secondary PM components.  Also, I doubt that Cl" is really an important secondary PM
component in coastal or any other areas. Its predominantly a primary PM component, and often
                                            83

-------
lost from particle phase, prior to deposition. Secondary organics are also important - especially
in urban areas. Further, the deposition rates of fine particles are substantially less than for larger
particles, such that the dry deposition of particulate matter components in urban areas is likely
more heavily influenced by coarse particles, gaseous precursors (or former particle components
that have disassociated from particle phase) than it is by fine particles - even if the fine particle
fraction tends to increase during episodes.

p. 9-90, lines 15-16 and Table 9-3 on p. 9-89:  Would fit better in the "Mercury" section, 9.4.5.2.

p. 9-98, lines 21-22:  Although anthropogenic V is typically heavily enriched in fine particles
influenced by oil combustion sources, there can also be substantial concentrations of V in soils,
and deposition is likely often dominated by natural sources in some areas.
p. 9-102, lines 1-6: There seem to be some inconsistencies here. For example line 5 describes a
global  [Hg(0) I presume] reservoir "with a residence time of a couple of years", while lines 31-
32 on p. 9-101 refer to an Hg(0) "residence time of about a year". Also I doubt that Hg(II) in gas
phase has a residence time of "months" or that Hg(II) associated with fine particles has a
residence time "similar to Hg(0)" of 1 or 2 years.

p. 9-106, lines 17-18: Its not clear what "much of the burden" refers to.  If it resides in the
atmosphere, to what organisms is it a burden there? If you're referring to the "atmospheric
burden",  I think pretty much all of it resides either dissolved in fog or clouds or as liquids or
solid aerosols.

p. 9-106, line 27: You could add "and other mining activities" after "limestone quarries".

p. 9-110, lines 1  &2,  and generally throughout this section. Sometimes the terms "metals and
"heavy metals" are not always used consistently here and sometimes just "trace elements" would
be a better term.

p. 9-111, line 3:  should be "exposures .. .have" or "exposure has".

p. 9-112, line 9:  Change "of to "to".

p. 9-113, lines 12-22: I suggest replacing the term "tire dust" with "road dust" in lines 12 14 and
19, and using the term "tire wear" in line 16. For the most part, the dust you're referring to is  not
predominantly composed of tire fragments, and its suspension to the ambient air is more driven
by the  aerodynamic action of moving vehicles rather than the direct action of tires on the road.
p. 9-113, lines 19-22: What's the meaning  of "particles embedded in tire dust"?  If Adachi and
Tainosho used the term "tire dust",  explain what they meant by the term (how did they collect
it?).  Is there any explanation for the two different compositions of "brake dust"?

p. 9-132, lines 14-15:1 don't disagree with your assertion that fine mode particles generally have
a larger effect on climate than large particles, but I think you need more explanation than "sizes
close to the wavelengths of visible light" before proclaiming "therefore..."
                                           84

-------
p. 9-136, lines 4, 5: This seems inconsistent with your citation to Forster et al., 2007 on p. 9-144
indicating a total direct aerosol RF of-0.04 + 0.04.

p. 9-138, Table 9-4 (or elsewhere in this chapter): It would be interesting to see the US subtotals
or fractions of the global emissions in this table - or otherwise provide some indications of the
extent to which relatively small modulations in the relatively small US contributions to the
global aerosol budget might be expected to influence climate.

p. 9-141, lines 1-3: You claim 3 ways but only list 2.

p. 9-142, lines 5-12: The point isn't clear to me here, but I assume any comparison of the
effectiveness of reducing BC vs CO2 or CH4, depends on the proportionate size of the reductions
in the different pollutants.

p. 9-142, lines 27-30:  Not sure what the point is in  describing the different assumptions on
(internal vs external) aerosol mixtures.  Was the range of results -0.39 to -0.78 W/m2 due to
which mixture assumption was used, and if so, which assumption produced the largest result?

p. 9-144, lines 3-10:  "These" in the second and third sentences doesn't agree with "Progress" in
the first sentence, and you only cite one inter-comparison.  I also don't understand why "coarse
aerosol fractions are largely responsible for variations in natural emissions". Do you mean
"coarse aerosol fractions are largely dependent on variations  in natural emissions?"

p. 9-147, Figure 9-57: This is an interesting graphic, but hard to read.  I'd like to see it larger
with a (much) shorter caption, which refers the reader to the text for a detailed explanation.

p. 9-155, lines 7, 8: Should be "role ... is" or "roles ...are".
                                            85

-------
Comments from Dr. Ted Russell

First, while I am sure I will note be the only one to comment on this, an 800 page document that
starts with "... a concise review, ..." might be open to some criticism for what is meant by
"concise." I recognize it could be longer, but I also think it could be shorter without loosing
much in terms of its support of the possible revision of the NAAQS. EPA should resist making
the ISA much larger in response to CASAC or other review. My ISA has travelled many
thousands of miles to get read.  On the other hand, it does pack a lot of information. As to where
the document might be trimmed a bit, speaking as an air quality engineer type, there are parts of
that section that might be put in an Annex. The main  information should be to provide a
foundation for understanding the health and welfare impact results later on.

Chapter 2

Chapter 2 on the Integrative Health Effects Overview is an important chapter, and reasonably
well done. The integrative summary, along with the conclusions, provide a solid and informative
overview and highlight the key points.  Inclusion of just the causal and likely to be causal
relationships make sense because those are what will drive the consideration of revising the
primary NAAQS. I would use exactly the same terminology on page 2-11 and in the Tables (I
would use causal and likely to be causal).  While I like what has been laid out in terms of
linkages between PM mass and the health endpoints, I think that the discussion of PM
components and sources linkages to health outcomes is insufficient.  I would have liked more of
a summary similar to what was done for mass linkages, or a statement  as to why such is not
there. Also, as the Chapter is now 25 pages long, and that this is the chapter that will get read by
those not wanting to read the other 800 pages, I think  a wrap-up page or two would be called for.
If you do decide to add more on impacts that do not make the top two classes of causality, I
would make that section short, focusing on what might be important in the future, and if it is an
area where there is a strong chance that as uncertainties are reduced, that area might influence
revision of the NAAQS.

While I like  how they have laid out the health endpoints and linkages, I do not believe that the
non-health effects endpoints should be left out of this  chapter. The ecosystem/welfare endpoints
need not get as much attention, but relegating all of that discussion to a single chapter at the end
of the ISA sends the wrong impression and might lead one to miss that we are extremely certain
about some of the welfare effects.  There are some very important effects that are not included in
the health assessment (that have possible health impacts via climate modification).

In Chapter 2 (page 2-4) an important point is glossed over: PM2.5 is a major fraction of PM10 in
many urban  areas. This likely drives much of the findings reported here.  This is also true at the
top of page 2-6 (it is not surprising that PM10 and PM2.5 are highly correlated much of the time
and in many locations). On page 2-6 they note various methods to measure temporal variation in
PM mass and components, but it does not include filter -based techniques anywhere in this
section. For completeness, there should be a section just on measurements (one paragraph)
before the discussion of spatial or temporal variability to discuss both integrated and
continuous/semi-continuous methods, together.  On page 2-7 they state factually that current
inventories overestimate primary components etc... the verdict is still out, and more goes in to
                                           86

-------
this than just volatility issues. This comment can be removed from here, though should be
discussed in Chapter 3. We are quite unsure of our emission estimates from the various sources,
particularly for organic species.
Chapter 3

Like past ISA's, I like the Source-to-Human Exposure approach to cover that aspect of the
review.  It is potentially very efficient, though as indicated above, this chapter might be reduced
a bit, keeping in mind what the key considerations are later on (e.g., possible mechanisms and
their linkage to specific components; spatial and temporal variability and epidemiologic studies;
components and epidemiologic results, etc.), and that the chapter will be used for the exposure
modeling as part of the risk assessment. I recognize that I might be asking for a bit of
clairvoyance to see exactly where the risk assessment might go.  The Scope and Methods
Documents helped figuring out part of that (see below).

As noted above, one should tread a bit more cautiously about stating that the volatility effects
have led to specific under/overestimates in the inventory.  This process might be discussed, but
there are many other factors that might lead to over/underestimates. (Also, 3-12,118 is a bit
awkward in that the emissions do not estimate anything).

Section  3.7.2.3: This section needs to be expanded (o.k., forget one of my first comments above).
Exposure modeling will be key to the Risk Assessment, and either a thorough exposition of the
models to be used should be contained here, or such will be required as part of the RA.

Like my comment above, Section 3.8.1.4 should be moved up to be inclusive of all measurement
techniques that have been used extensively enough to be included in such a summary (e.g., filter-
based techniques). Make a section  3.8.1.2

Section  3.8.2.3 is important and needs to make its points very strongly. It should answer the
question (or lay the foundation for answering the question) as to whether confounding is a likely
issue just based upon the correlations between pollutants and in the exposure studies. It actually
only goes halfway.  It provides a strong statement about the indoor-outdoor PM source exposure
confounding, but not if a gaseous species may lead to confounding. It ends on a negative note
that may be of secondary importance (what it says is important in terms of assessing the epi
studies for the gaseous species impacts), but not if the gaseous species may confound the PM
results.  .

Chapter 9.

This chapter suffers from not being as completely developed as it could/should be.  There is no
integrative summary.  It is also very weak on assessing how anthropogenic emissions impact
PM10-2.5 (and the resulting welfare effects) over natural emissions in urban areas, and how this
relates to apportioning visibility reduction.  It should also be more specific about confidence, and
probably more bold, and how uncertainties/certainties should impact consideration of possible
revisions to a secondary NAAQS.  This is particularly true for climate. While there are
                                           87

-------
uncertainties in climate relationships, but there are topics for which we are relatively confident.
This is particularly true for the direct effects.  Looking at the question as to what might drive
revision of the NAAQS, climate change should be part of that, if not play a central role. I am
concerned that this chapter in its current form leads to a lack of consideration of climate impacts
(both ways). This section needs to make the link between climate change and health impacts
(which, I agree, are uncertain).  It is not clear what the section on POPs adds.  (I would suggest
focusing the climate section.)

A couple details:

What is meant by "warmed molecules can have long lifetimes and be transported large
distances..."

Given the issues that have come up in both Scope and Methods documents as to how to treat
PM10-2.5, the PM ISA needs to do a more comprehensive job on that topic.  The additional
information and analysis includes how well it is treated in CMAQ,  and the evaluation of CMAQ
for simulating PM10-2.5. It also includes how well we can distinguish between the
anthropogenic and natural PM10-2.5 fractions, and how the two correlate.  Further, given that the
SM-Welfare proposes to use scaling of the source apportioned PM2.5 to estimate the
corresponding PM10-2.5, the components of that process should be discussed here in greater
detail.

Another topic that I worry about from the point of insufficient coverage is climate. While there
are many uncertainties as to how aerosols impact climate, there is enough known that this topic
should be influencing how we consider revising the secondary standard, and possibly even the
primary standard.

Specifics:
Fig 3-1: In the caption, need to do a bit better job as to what the dashed lines versus solid lines
mean.
Maps (e.g., Figs 3-19, 21 etc.) A bit more info than just the  roads would be helpful (e.g., rivers).
Fig. 3-30. I don't think you mean Pittsburgh in the figure caption.  I think it is Philadelphia. Did
you mean to switch cities?
Fig. 3-2: Should note anthropogenic.
Page 3-18: Cations and Anions.
Page 3-122: Note sure why they think that the Voyageurs site was not impacted by fires.
Fig. 3-69... Why change colors from prior figures?
Table 3-23: Be consistent in how you distinguish between mod and obs (use mod; obs in the
column headings).
                                           88

-------
Comments from Dr. Frank Speizer
Chapter 1
General Comment: This chapter outlines both the historical Clean Air Act activities as related to
Particles and provides a discussion by which criteria can be applied to establish causality. It
seems that after the several iterations that have taken place in trying to get this right, the Staff has
put together a summary of the issues that are hard to fault, save one issue mentioned below. .
One question might be whether there is a better way to summarize the weight of evidence in
terms of consensus building. A recent discussion at the Intergovernmental Panel on Climate
Change adopted seven verbal expression of certainty.  I list them here for purposes of discussion
only to  see if CASAC should consider them in our deliberations:
Verbal Description Percent Certainty
Virtually Certain
Very Likely
Likely
More likely than not
Unlikely
Very Unlikely
Exceptionally unlikely
Considered more than
99% likely
More than 90%
More than 66%
More than 50%
Less than 33%
Less than 10%
Less than 5%
 Intergovernmental Panel on Climate Change, 2007
 The chapter reads well and needs only minor editing
Added here are the factors in Chapter 1 for causation: (for purposes of discussion)
Table 1-3. Weight of evidence for causal determination.
Health Effects
Causal
relationship
Evidence is sufficient to conclude that there is a causal relationship
between relevant pollutant exposures and the health outcome. That
is, a positive association has been observed between the pollutant
and the outcome in studies in which chance, bias, and confounding
could be ruled out with reasonable confidence. Evidence includes, for
example, controlled human exposure studies; or observational studies
that cannot be explained  by plausible alternatives or are supported by
other lines of evidence (e.g. animal studies or mode of action
information). Evidence includes replicated and consistent high-quality
studies by multiple investigators.
Likely to be a
causal
relationship
Evidence is sufficient to conclude that a causal relationship is likely to
exist between relevant pollutant exposures and health outcome but
important uncertainties remain. That is, a positive association has
been observed between the pollutant and the outcome in studies in
which chance and bias can be ruled out with reasonable confidence
but potential issues remain. For example: a) observational studies
show positive associations but copollutant exposures are difficult to
address and/or other lines of evidence (controlled human  exposure,
animal, or mode of action information) are limited or inconsistent; or b)
animal evidence from multiple studies, sex, or species is positive but
limited or no human data  are available. Evidence generally includes
                                             89

-------
replicated and high-quality studies by multiple investigators.
Suggestive of
a causal
relationship
Evidence is suggestive of a causal relationship between relevant
pollutant exposures and the health outcome, but is limited because
chance, bias and confounding cannot be ruled out. For example, at
least one high-quality study shows a positive association but the
results of other studies are inconsistent.
Inadequate to
infer a causal
relationship
Evidence is inadequate to determine that a causal relationship exists
between relevant pollutant exposures and health outcome. The
available studies are of insufficient quantity, quality, consistency or
statistical power to permit a conclusion regarding the presence or
absence of an association between relevant pollutant exposure and
the outcome.
Suggestive of
no causal
relationship
Evidence is suggestive of no  causal relationship between relevant
pollutant exposures and health outcome. Several adequate studies,
covering the full  range of levels of exposure that human beings are
known to encounter and considering susceptible or vulnerable
subpopulations,  are mutually  consistent in not showing a positive
association between exposure and the outcome at any level  of
exposure.

Specific Comments:
Section 1.5, 1.11  Discussion of factors evaluated for causation':, line 28:
defines cause, in  contrast to statistical association Not clear that this belongs here as part of the
discussion.  Almost all of the factors entering causation with regard to population data are in fact
statistical.

Page  1.15-16:  Confounding needs to be defined more clearly, just as effect modification is.

Chapter 2
General Comment:
This is a useful summary if it indeed truly reflects what is contained in subsequent chapters.
There will need to be some cross checking to be sure that the major finding are all summarized
here.  I would likely take issue with the classification of cancer.  I think it will need to be broken
down to lung cancer and other cancers as I think the ACS data would at least be "suggestive".

Specific Comments
Page 2-5, line 13: Assumption is that "diel" is short for diurnal but needs to be spelled out first
time.

Page 2-6, end of paragraph at line 23:  Suggest a statement is needed here that the composition of
PM10 changes  by region and that the distribution and ranking as sources of wild fires, road dust
and EGUs would vary by region.
                                             90

-------
Page 2-18, Paragraph on Cardiovascular Morbidity: There is an issue here in that using 1 year
data for classifying exposure doesn't make biologic sense for a chronic effect. It needs to be
made clearer that the 1 year figures are " surrogates" for long term exposure (over several years).

Page 2-23, Section 2.3.3.:  This section coming after the Chronic effects section above needs to
be clarified that all the effects assessed in table 2.1 are in fact acute (or semi-acute) effects.
Although the table indicates inadequate data for a relationship to cancer there is no discussion in
the text.  In addition the newer data from ACS would suggest that the classification might be
moved up a notch to suggestive.
Chapter 3
General Comment:  The authors of this chapter have done an exceptional job in summarizing a
large body of data and presenting in a readable fashion that makes a very reasonable logical case
for the factors that move particles from source to measuring sites, including factors of size,
chemistry and transport. Trends over time although not large are impressive. Perhaps more
could have been included on the effects of wind direction, although I doubt that much useful data
is available that would alter conclusions reached.  The last few sections of this chapter that relate
to estimates for human exposure and population or community effects are logical and well
summarized.  I personally would have like to have seen a little more in interpretation of the
directionality and factors affecting the directionality of the surrogates used to estimate population
exposure. What factors lead to over or underestimation, and perhaps how results might be
adjusted for these factors. Perhaps this will come up in later chapters on Dosimetry or Health
Effects.
Specific Comments:
Page 3-1, line 11: The use of the word "uncontrollable"  For describing background pollution
level would it not be better to describe this level as "natural occurring"?  Uncontrollable could be
in the eyes of the beholder, since for some the cost of controlling could predict what was
uncontrollable (i.e. Too expensive to control, or technologically difficult).

Page 3-2, last line 34:  Some further explanation of the range of values at National Parks is
indicated. If the natural occurring background is < Iug/m3 than what is being defined as
background in the range of 3-63 is clearly being impacted by long range transport. This makes
for a confusing definition of background.  Are we saying that background is natural plus long
range transport and therefore varies regionally depending on down wind sources or is
background some thing constant, but not achievable?

Page 3-3, line 20: Is it really seasonal differences in sources, or is it differences in photochemical
activities by season. One suspects that latter, partly depending upon region where winter heating
and summer cooling might result in match emissions from stationary sources and mobile sources
between seasons not varying that much and dominate variable being amount of sunlight and
temperature.

Page 3-17. Paragraph ending line 15:  This section provides well describe logic of what the FRM
for PM10 is doing. However, in so doing it provides information that 50% of the particles
                                           91

-------
greater than lOum are getting collected. I think some mention of an "upper limit" needs to be
indicated. Is it not the case that there is an absolute cut off at 15 or 30um?  And if so (or if not) a
justification of what is getting through, particularly when focusing on mass, needs to be offered
to deal with the mass not being dominated by these extra large particles.

Page 3-30, Figure 3-4. Need to change monitor indicator from rectangle to circle, if I am reading
figure correctly. Ditto figure 3-5.

Page 3-32, Tables 3-3 and 3-4: Need to provide footnote as to why data are "na" for New York
City.

Pages 3-40 to 3-46.  These pages contain a series of tables and figures that indicate the
distributions of PM measures across the 15 sites over the 3 year period.  Although the reader can
look at these tables ad draw their own conclusion, text contained that describes them does not
indicate any particular finding. It simply describes what is in the tables. I would have thought
that some summaries in text would be appropriate (like the 75%tiles are all above the annual
standards or although some values exceed 1000 ug/m3 at the max, the 95%tile do not ever
exceed 60ug/m3 for the 3 years 2005-2007 or Phoenix and Denver are the only cities with
PM10-2.5 with annual means 20 or higher, etc).  See text descriptions on pages 3-51-3-53 for
Figures 3-9-3-13 which give a much better assessment of what is presented.

.Page 3-87, Figure 3-42. The text indicates that this figure is to be replaced. I look forward to
seeing the replacement, particularly because this figure and the accompanying text to my mind is
critical to our understanding of the effects of distance from source, particularly as related to road
traffic for ultra-fine particles.

Page 3-109, Figure 3-58. Typo in Vertical Axis.

Page 3-130, Figure 3-70. Should this be divided into two figures (if data exist) since there is a
real mixing of children, who contribute almost all of the in school time, with adults who are
more likely to be out and about or indoors in a different setting?

Page 3-132, after line 5: A further conclusion here would be useful. If ambient concentration is
better than personal  exposure as a surrogate  for ambient exposure is ambient exposure less likely
to over or underestimate true exposure.  Maybe this will come up in the next section, but with the
relatively high correlations in table 3-27 some statement is necessary.
ADDED note after page 3-160.  This section is well written and summarizes well  the issues
needed to be considered.  However, I would still like to see a more interpretive statement of the
bottom line. Do ambient measures over or underestimate population human exposure and if so
in what direction and under what circumstances or what influences the direction of the over or
under estimation?

Page 3-168, line 7: Is this last sentence true for all regions? I would have thought that broadly
in the Midwest or western plains we know more about PM10-2.5 as indicated in some of the
tables.
                                           92

-------
Chapter 5
General Comment:
       This chapter presents a number of mechanisms of injury but does not appear to be
presented in an even fashion. In some cases the mechanism is being discussed from a basic
pathophysiologic perspective with no particular relation to PM exposure, in others it is more
focused on the effects of PM. I think this is ok, since in the next chapter it has been promised
that the focus will be on PM. This might be helped by some cross referencing on the last 2 pages
where there are a number of bullets (some of which could be combined) with added references to
places in Chapter 6 and 7 where they are discussed.  One issue that seems to be missing (and
may be missing from the subsequent chapters) related to potential reproductive outcomes.
Although there is not a lot of data there are several studies pointing in this direction so potential
mechanisms probably should be considered.

Chapter 6
General Comments:
Reasonably well done job of integrating toxicology, human studies and epidemiology for each of
the exposure scenarios discussed as well as for each of the outcomes of interest. However, the
text gets overwhelming when it gets to the summaries. The initial parts  of presenting the level of
causal certainty is fine, but then the details provided on the various disciplines data (epi, human
clinical, and tox) seem to get somewhat contradictory. For example, the discussion of short term
respiratory morbidity general conclusion seems to be contradicted with discussions of specific
symptoms or pulmonary function, which does apply to all groups equally.  Surely that is not the
place to be bringing out the subgroup variation, unless the purpose is to  confuse the reader. It is
also quite redundant as all of the details are previously discussed in the chapter.  I would have
left more of the toxicological studies in the appendix only.

Specific Comments
Page 6.3, paragraph beginning line 11: This is an appropriate  discussion of time-series, but
somewhere in the paragraph it needs to be stated more explicitly that there is generally a
relationship between PM and mortality that is seen in these studies.  Then it becomes appropriate
to discuss the statistical issues and issues of threshold

Page 6-4, line 23:  Should this be "linear non-threshold"?

Page 6.7, line 11:  Subtle, but would change "do" to "may"

Page 6-18,  line 30:  The reason one expects a  decline in HRV with exposure is because of the
Framingham data that suggests lower HRV predicts  CVD.  However, there are some who
suggest that change in HRV in either direction represents a response. See top of page 6-21  on
animal studies. It therefore may not be appropriate to call the response "inconsistent".

Page 6.24, line 11:  typo; ditto line 12

Page 6.28, line 5:  "discrepancy" is a value judgment that would appear unnecessary.  I would suggest
something like variability in the several studies.
                                           93

-------
Page 6-54, lines 12 and 14:  Please check whether 100ug/m3 is correct. (Would have thought it should be
10ug/m3.)

Page 6.85, line 19:  Issue with the word "consistent" In fact looking at figure 6.2 the results is
really quite consistent.  All but one show a positive effect, albeit that not all are statistically
significant. With the subsequent data on heart failure admissions and ER visits it I suspect the
summary could be a little more positive than as expressed.

Page 6.88, line 23:  I would agree with the conclusion as written but it begs the question of what
is the summary of the ambulatory arrhythmia data. Will this come up later? If so probably
should indicate to be discussed in section 6? If not should be discussed here.

Page 6.90, line 27 end of summary. There is a particular issue with regard to  stroke.  Some
comment need to be made that the appropriate lag structure is simply not known. The 0-2 days
may not be appropriate as the onset of symptoms may be much longer than is reflected in
administrative data bases on date of admission. Most of the studies cited do not include any
specific query about symptom onset.  This is particularly true for the embolic disease which
make up about 80+% of the events and in which the events are likely to be acute than for the
hemorrhagic strokes.

PAGE 6.94 SUMMARY: Logic well presented with discussion of toxicological data to support
mechanisms for the conclusion of likely causal.

6.94, line 26 to end of 6.96: this  section seems out of place.  It is a summary of the details used
in the second paragraph on page 6.94  and therefore might all come before the last paragraph on
page 6.94 (which is the conclusion for the section on PM10.
Ditto next several sections.

Page 6.93, Section 6.2.11  Summary and Conclusions
Pollutant     Chapt statement      Reviewer position
PM10        Likely causal        agree
PM2.5        Causal               agret
PM10-2.5     Inadequate evidence  +/-  agree
Ultrafme      Inadequate evidence  agree
Because PM2.5 is contained in PM10 and particularly because PM10-2.5 in inadequate some
discussion need to be here as to what the evidence is that makes PM10 likely causal separate
from effects of PM2.5. Even is the argument is that for many of the studies it is the only data we
have. Further at least for Ultrafme, (and  perhaps for PM10-2.5 it needs to be  stressed that it is
the lack of sufficient study rather than either inconsistent studies or clearly negative studies.

Respiratory effects
Figure 6.5 and 6.6 and Table 6.7: Although all the information is in these figures  and table, (and
is given in the text) it is somewhat frustrating in that I would like to see the odds ratios along side
the pollution measures. How to do this?  Maybe it is done in the Appendix and I will check.  If
there however, a cross reference is necessary.  If not, it probably ought to be there.
                                           94

-------
Page 6.117, lines 18-22: Potential problem in that the mass data for PM2.5..0i contains all of the
count data NC.oi-.i. Therefore the attenuation would be expected.  This is hinted at in the
statement that the correlations with the gasses is high but probably needs to be better spelled out.

Page 6.124, line 2:  typo; ditto page 6.134, line 4

Page 6.152, line 5-6.  The entire tox section (now that it is written) probably could have been a
lot shorter.  It reads too much like the old CD, and the writer could have been asked to shorten.
As indicated here much of the tox material presented was done at high doses. And thus is less
relevant and therefore relegated to the appendix.  .

Figures 6.9-6.11. I find it useful in these figures to have provided the range of estimates from
the 2004 CD. However, in presenting these individual and multicity studies for the most part
there is an overlap  with 1 as presented? I assume there will be a combined risk assessment
analysis presented  in subsequent document but it would be useful to have it here as well.

Several of the figures: May want to put into the legend of these figures that they are expressed
as Excess Risk % (if that is what they are)!

Page 6.182, line 8.  This sentence needs to be qualified.  It appears to apply only to asthmatics
and the elderly.
Summary and Conclusions
Pollutant      Chapt statement                                 Reviewer position
PM10         Short term Respiratory Morbidity
PM10-2.5     Short term Respiratory Morbidity
PM2.5        Short term Respiratory Morbidity
Ultrafme      Short term Respiratory Morbidity
PM           CNS Symptoms or Disease

PM10         Short term mortality
PM2.5        Short term mortality
PM10-2.5     Short term mortality
Utrafine      Short term mortality
likely causal          agree
suggestive of causal   agree
likely causal          causal
inadequate evidence   +/- agree
inadequate evidence   agree

likely causal          agree
likely causal          causal
suggestive of causal*  +/-agree
inadequate evidence   agree
* Might take issue with fact that although fewer studies, with regard to cardiovascular disease the
measure of PM10-2.5 appear to be where PM2.5 were in 2004 and might move up to likely
causal with a few more studies,  (not likely to go the other way).

Page 6.247, Table 6.16 starting this page
I  do not find this a very useful table as a summary of Epidemiologic Studies for source
apportionment. First many of the studies reported are not epidemiology (e.g. the first one is
mice). Second there is lots of NR or na reported in the table.  I do not get the message from the
table. Suggest redue more specifically or leave out.
                                           95

-------
Chapter 7
General Comments: There are several difficulties with the way results are presented in this
chapter.  First of all most of the contrasts in exposure that are discussed are presented as
increases in pollution  when in fact they are not increases but differences between regions.
Secondly and perhaps more important results are reported as differences in hazard ratios or rates,
albeit mostly with c.v.s given, that can only be described as null  effects.  This is carried
throughout the presentation of the individual outcomes. In particular this relates to intermediate
outcomes related to CVDs.  In fairness generally there is a summary statement at the end of the
section that indicates the author recognizes this but it needs to be part of the description of the
results rather than what appears to be a pushing of the results.  To anticipate what is coming it
might also be discussed  as to the role of these intermediate outcomes in the process of
identifying health outcomes.  If anything there is a prior that is mentioned early on that chronic
inflammation is a major pathogenesis pathway. In fact since the markers used for estimating this
effect are not related to air pollution, it suggests that some other  alternative mechanisms should
be invoked.
Specific Comments:
Page 7.1, line 1:  Minor point. Should the order be changed in this sentence?. The chapter
reviews, integrates, and summarizes rather than summarizes first.

Page 7.2, line 12: I am not sure that all would agree with this. Chronic inflammation may be a
component of Atherosclerosis but certainly is not the sole cause.

Page 7.2, line 24: Define CAC if first time used.

Page 7.2-7.3: Although useful to define the clinical tools used to assess atherosclerosis, what is
missing is the any quantitative assessment of the coefficient of variation of theses tools.  This is
particularly important since the next sections indicates small differences resulting from deltas in
PMof 10ug/m3..

Page 7.4, two paragraphs.  There is something terribly wrong between these two paragraphs.
The magnitudes  of the deltas in the first paragraph are virtually zero and none are significant. To
discuss them as "increases" is ludicrous. On the other hand deltas in the second paragraph are 50
to 100 times higher and are significant, albeit the PM measures are different. Is there a decimal
point error in one or the other?

All of the epidemiology section on Atherosclerosis:  This entire section needs to be rewritten.
Except for the one study from Germany all of the other data presented is definitely null.  To
describe any of the findings as showing "increases"  "associations"  etc. is misleading the reader.
In addition, since these are cross sectional studies to describe differences in PM levels as
increases in also not correct.

Page 7.9, line 13: change "increase" to  "difference"
Page7.15, line 13:  again change increase to difference
                                            96

-------
SAME PAGE, In line 15-16 in contrast to lines 13-14 is the reference to NON-FATAL CHD or
MI? Otherwise these are two opposite statements about PM10. Alternatively, one statement
may be about PM10 and other PM2.5.  Please check.
Line 23, again change increase to difference
Page 7.17, line 6: ditto
Line 13-15:  Two problems.  What does it mean after adjusting for clinical characteristics to not
find a prevalent difference in CHD?  If one of the clinical characteristics is prevalence of disease
this is what would be expected.  Second the deltas reported with the c.v. given are really null.  To
describe them as difference is misleading.
Line 21:  Again, these findings are null. (Alternatively to be consistent with the way others are
being reported one would say that a 10ug//m3 difference in PM was protective!
Lines 28-29:  What does this mean given the previous sentences reflects a null?

Page 7.18, line 20:  Don't need OR and % difference. (They are the same numbers)
Page 7.21, line 4-5:  This conclusion needs to revisited. Most the studies reported are in fact
null, albeit that they are in the direction that might lead to this conclusion. The important issue is
that the direction of the effects is supported by the toxicological work. This needs to be factored
in, somehow, to the logic that allows the author to reach this conclusion.

Page 7.32, Figure 7.4  As presented this figure is more confusing than enlightening.  It mixes up
both cross sectional effects with longitudinal change. It is not clear what arrows going in both
directions means. The text that describes some of the finding is much clearer.  I would
recommend leaving Figure out or expanding greatly and perhaps even making it into multiple
figures.

Page 7.33, sentence beginning line 1.  Not clear what is meant by pulmonary function in first
year of life, surely not the measure reported.
Paragraph beginning line 16. Not clear how exposure in 1st and 2nd trimester is separated from
3rd trimester.
Summary results
Pollutant      Chapt statement                                 Reviewer position
PM10         Long term CVD morbidity   suggestive of causal                 agree
PM2.5        Long term CVD morbidity   likely causal                       agree
PM10-2.5     Long term CVD morbidity   inadequate evidence for yes or no    agree
Utrafine       Long term CVD morbidity   inadequate evidence for yes or no    agree

PM10         Long term Respir Morbidity  likely causal                       agree
PM10-2.5     Long  term Respir Morbidity No data for conclusion on causality +/-agree
PM2.5        Long  term Respir Morbidity likely causal                       causal
Ultrafine      Long  term Respir Morbidity inadequate evidence            suggestive

Section 7.4.2 Toxicology of reproductive outcomes. Much of this could have shortened or left to
the appendix. As indicated before this section  none  of the mechanisms seem applicable to the
epidemiologic findings.  Therefore it seems the details of the toxicological data are not really
helpful and just makes the chapter too long.
Pollutant      Chapt statement                                 Reviewer position
                                           97

-------
PM10         Reproduct & Developmental suggestive of causal         agree
PM10-2.5     Reproduct.& Developmental inadequate evidence         agree
PM2.5        Reproduct & Developmental suggestive of causal         agree
Ultrafme      Reproduct & Developmental No data for conclusion on causality agree

Section on Cancer: This section is extremely disappointing with regard to the presentation of the
epidemiologic data. The author seems much more comfortable presenting a discussion of the
toxicology, carcinogenesis and mutagenesis even to the point of discussing in vitro studies. The
epi is extremely naive. For example, on Page 7.87-88, Paragraph on Breast Ca and TSP:  These
data are presented in a crude fashion with out discussion of possible confounders. The ORs are
in fact considerable higher than any reported with smoking, where the exposures to potentially
toxic agents are considerably higher than might be seen from environmental TSP. A more
careful reporting therefore is warranted.  Further there is no discussion of any other studies.
This appears to leave out the Pope analyses of the Amer Cancer Society study, which does
indicate a rather well confirmed risk about a 40%  excess risk of total cancer with higher risks for
lung cancer.

I would therefore disagree with the summary of inadequate data for an estimate  of causality.  I
would suggest at least suggestive.

Page 7.117, sentence beginning line 7: This needs to be reworded. The fact that the new studies
focused on CVD effects rather than respiratory effects doesn't make the respiratory effects any
less important (except there are less of them). I think the authors know this but the way they
have said it reads as though a change in thinking has occurred. This will come up again in next
chapter.
Pollutant      Chapt statement                                Reviewer position
PM10         Overall Chronic mortality    suggestive of causal*        likely
PM2.5        Overall Chronic mortality    likely causal                agree
PM10-2.5     Overall Chronic mortality    inadequate evidence         +/-agree
Ultrfme       Overall Chronic mortality    not considered              suggestive
*Might be argued that there are sufficient evidence to raise to likely

Chapter 8
General Comment:
The  authors have effectively laid out the potential susceptibility and vulnerability characteristics
and then  systematically discussed each.  One might take issue with the conclusions that almost in
every case because there is some inconsistency in the results reviewed that the uncertainty that
results leads to the position that more research is needed. Albeit it true that should not leave the
document without indicating more or less to what degree the argument is made.  In fact in most
cases the reasons the particular characteristic is on the list is because there are some groups in
which there is more than chance operating that a particular subgroup is more susceptible or more
vulnerable. Chapter needs a concluding summary paragraph.
Specific Comments:
                                           98

-------
8.14, line 5:  This statement may be true for "in association with exposure to PM" but is not true
as stated in general.  Clearly those with COPD are more susceptible to exacerbations of COPD.
Suggest qualify sentence more specifically.
                                           99

-------
Comments from Dr. Helen Suh Macintosh
Question 1: The framework for causal determination and judging the overall weight of evidence
is presented in Chapter 1.  Is this framework appropriately applied for this PM ISA? How might
the application of the framework be improved for PM effects?

The two-step framework for causal determination and weight of evidence is sound and is
appropriately adapted to PM by allowing this framework to be applied separately for different
sized particles  (PMio, PMio-2.5, PM2 5, and ultra-fine particles), different adverse health/welfare
outcomes, and short and long-term exposure periods. Although implicit to this framework, the
explicit consideration of causality by major particle sources or components should also be
considered, even though there is currently insufficient evidence to make any determinations.  In
addition, there should be some integration of evidence across particle sizes as well, especially
given that PMio includes PM2.5, PMio-2.5, and ultrafine particles and that many recent health
studies (especially of intermediate health markers)  no longer examine PMio by itself. In this
regard, it may be preferable to include discussion and consideration of PMio after that for PM2.5
and PMio-2.5, with findings used as evidence in support of or contrary to those for PM2 5 and
PMio-2.5.  Else, causal determinations for PMio should be made in part based on determinations
for PM2.s and PMio-2.5-  This reorganization may help to explain apparent inconsistencies in
causal determinations (as discussed further below).

Question 3: To what extent are the atmospheric chemistry and air quality characterizations
clearly communicated, appropriately characterized, and relevant to the review of the PM
NAAQS? Does the information on atmospheric sciences and exposure provide useful context
and insights for the evaluation of human health effects ofPMin the ISA?

a. Is accurate and appropriate information provided regarding PM source characteristics,
   techniques for measuring PM and its components policy-relevant background PM, and
   spatial and temporal patters of PM concentration? Are the analyses and figures presented in
   Chapter 3 effective in depicting ambient PM characteristics?

Chapter 3  presents PM composition, sources, and spatial and temporal trends for PM, its
associated size fractions and its major components  in a clear, concise, and well organized
manner, with trends shown for different spatial (neighborhood, urban, regional and global) and
temporal (hourly, daily, and yearly). Specific comments are below:

•  For clarity  all figures and tables should explicitly state the averaging time of the measured
   concentrations.
•  As was done for PMio and PM2.s, maps illustrating the location of speciation monitors
   relative to population density should be included, although perhaps just in Annex A.
•  A table showing a summary of the  exposure-related factors for each of the selected 15 cities
   should be added.  These factors may include population density, number of large stationary
   PM sources, some metric of traffic congestion,  percent of population living within 100
   meters of a busy road (e.g., class Al or A2), percent of people living in apartments or single
   family homes, and central or room AC prevalence.
                                           100

-------
•  The Figures 3-14 through 3-18 were very informative and effectively showed the variation in
   PM2.5 composition by region and by season; it would be helpful to have the sample sizes
   noted somehow as it would help determine the comparability and robustness of the city-to-
   city comparisons.  The discussion of how composition may vary within a city and near
   sources should be expanded.
•  Although noted, interpretation and discussion of Figure 3-43 should emphasize that even
   though neighborhood scale variability in PM2.5 was rather low, it may not be the case for
   PM2.5 components, as concentrations of locally generated components,  such as EC, have
   been shown to vary substantially over very short distances. The implications of spatial
   variability in the components to the interpretation of epidemiological studies should be
   explained.
•  Diurnal variation in hourly EC, OC, SC>42" and other major particle components should also
   be presented and discussed, especially in relation to particle number concentrations.
•  It would be interesting to include some discussion or graphic of sites that are in violation of
   the annual but not 24-h NAAQS for PMio and PM2.5, specifically related to whether these
   sites exhibit a seasonal or spatial daily concentration pattern.
•  Although introduced in Section 3.5.1.3 and discussed later in 3.7.2.3, methods  used in source
   proximity studies, such as land use regression models,  should be mentioned in  the section
   Estimating Source Contributions to /^(Section 3.5.4)
•  The discussion regarding PMio-2.5 measurement  devices should be expanded, especially with
   regard to the precision of various methods (including difference, flow rates, etc.)
•  Composition of PMio-2.5 should be discussed.

b.  Is the evidence relating human exposure to ambient PM and errors associated with PM
   exposure assessment presented clearly, succinctly, and accurately?  Are there PM exposure
   issues that should be expanded, shortened, added or removed?

While relatively complete, this section is generally not as cohesive and clear as the previous
section. It would benefit from an upfront section that briefly introduces the concept of exposure,
its relation to ambient levels, and its importance in the interpretation of ambient concentrations
and health risks. This introduction could then be followed by a brief description of the sections
or discussions to follow. In addition, for consistency it would be preferable for this section to
follow the pattern established in other parts of the ISA, with each topic discussed for each PM
size fraction (e.g., PM2.5, PMio-2.5, PMio, ultra-fine particles,  and components) and also for acute
and chronic averaging periods. Alternatively, the exposure section could follow the general
structure of the Human Exposure section in the Conclusions (Section 3.8.2), which is very clear,
well written, thoughtful and concise.  In addition, the section appears to focus primarily on PM2.5
measured over 24-h periods (although the specific PM size fraction is often not mentioned). It
would be preferable to include evidence or information of exposures for PMio and PMio-2.5 as
well and also for PM exposures (of all size ranges) measured over shorter and longer time
periods.

Specific comments are below:
                                           101

-------
•  Page 3-131, lines 7-3-132 is too complicated, long, and confusing.  Its main points are
   important, making their clear and simple discussion even more necessary. It is possible that
   this section is misplaced, adding to the confusion.
•  Section 3.7.2.3 PM Exposure Modeling should include a discussion of time-weighed
   microenvironmental models and GIS-based spatial smoothing models.
•  The section entitled "Outdoor Exposure to Local Sources" is misplaced and should be moved
   to Section 3.5.4.
•  Figures 3-74 and 3-75 should be clearly annotated with regard to the location,  study
   population, and time period of the study
•  Section 3.7.4 Exposure Assessment andSocioeconomic Status is misplaced and should be
   moved to the Section 3.4.2.2.

c.  To what extent does the Panel find Annex A appropriate, adequate, and effective in
   supporting the ISA?

Annex A contains a voluminous amount of data, which takes away somewhat from its
effectiveness. Further, several figures and tables are contained in both the ISA and in the Annex,
probably mistakenly, but they should only appear in only one place.  Regardless, the Annex adds
important, supporting concentration and exposure information to the ISA. Figures that could be
condensed into tabular form include:

     •   city-specific maps of monitoring site by population density and SES, which while
        beautiful are numerous; data could be placed into a single table each for the PMi0, PM2.s
        and speciation monitors.
     •   city-specific PM2.5 composition pie charts, with data expressed in the table as a range if
        there are multiple pie charts for a given city

To help sort through the data, it may also be important to categorize findings further.  For
example, studies summarized in Section A.4.1 Exposure Assessment Study Findings should be
grouped first by general topic area and then sorted alphabetically by last name of the first author.

Question 7: Strength, consistency, coherence, and plausibility of evidence for health effects of
PM?

The  chapter on chronic health impacts should discuss the relation of chronic and acute impacts,
for example whether they occur via similar or different biological pathways. The  causal
determinations for the various PM measures are appropriate and supported by the  available
scientific evidence.  Some allowance may be necessary for determinations of causality for PMi0
either now or in the  future, as an increasing number of health studies appear to no  longer
examine the health risks associated with ambient PMi0.  As a result, additional evidence of
PMio-associated acute or chronic health impacts may be difficult to obtain or may have to be
pieced together from PM2.5 and PMi0.2.5.

As mentioned above, causal determinations for PMi0  are not always the same as those for its size
fraction groups, PM2.5 and PMio-2.5, which can present seeming inconsistencies.  For example, if
short term effects of PM2.5 are causal with relation to  cardiovascular morbidity, why isn't the
                                           102

-------
same determination made for PMio given that (1) PM2.5 generally comprises a substantial
fraction of PMio and (2) correlations between 24-h ambient PMio and PM2 5 tend to be strong.
The practical implications of this and other seeming inconsistencies should be noted.
                                          103

-------
Comments from Dr. Sverre Vedal

For each chapter, numbered points that are considered to be more major, general or substantial
are followed by a subsection of more specific points.

Chapter 1. Introduction.

1.  Organization.  The flow of topics is generally good, but there are some places where
consolidation and change in ordering might help.
Examples:
1-15, line 21. This starts the discussion of confounding, but this is continued and completed only
   later (line 33 and p. 1-16) after it is interrupted by bringing up the notion of effect
   modification. Line 25 starts a listing of the "multiple ways" PM can be associated with
   outcomes, but only lists three ways. There are other ways of getting a spurious association:
   e.g., any number of non-pollutant factors that are associated either in space or time
   (depending on the study design) with PM. [By the way, I'm not sure effect modification has
   a place in a discussion of causation. Also, the definition of effect modification here is
   awkward and weird - effect modification simply means that exposure effects differ across
   some subgroups or strata of the population.]
 1-16, line 32.  The measurement error discussion here should follow immediately after the
   multi-pollutant modeling discussion (line 11, etc.), rather  than being interrupted by a
   discussion of stratified analysis.

2.  Causality criteria and grades (Table 1-3).
i. Although the Hill criteria probably need to be brought up, they have limited utility, as
evidenced by all of the qualifiers in the description of each criterion.  The desire to avoid
"criteria" is understandable, but replacement with "aspects" (1-18, line 8) is awkward - how
about "characteristics"? "corollaries"?
ii.  The weight of evidence grades (1-22) are problematic.  Some suggestions:  1) drop
"reasonable" as a qualifier to confidence for the causal grade  -  "reasonable" to me is not strong
enough to allow a categorization of causal; 2) I'm not sure I see much difference between
"likely" and "suggestive" here; 3) why is there no "not causal"  or "likely not causal" category vs.
"suggestive" of not causal only? i.e., there is no symmetry here. In general, I don't think there's
much to be gained by trying to make these grades parallel the 5-level categories of EPA
Guidelines for Carcinogen Risk Assessment (footnote 1-21).
iii. See3.il, below.

3.  Things not discussed.
i. It is important in this chapter to bring up publication bias: how it happens, the evidence for  it,
its impact on both the assessment of causality and the exposure-response relationship (slope and
shape), and how it's countered.
ii.  There are also other soft criteria that have been floated recently (e.g., the loannidis series of
papers) that highlight features influencing whether published  findings turn out to be true; these
could bear on our assessment of the published scientific literature (i.e., the literature being
evaluated here) and the determination of causality.  A few points that should increase suspicion
and vigilance: small effects sizes, large number of potential comparisons, flexibility in outcomes
                                           104

-------
and modes of analysis, potential for conflict of interest, a "hot" scientific field. I bring this up
with some hesitation, but decided in favor of it when I again noted the many instances in the air
pollution health effects literature where this perspective might have merit.
iii. What is the solution?  Replication of specific findings is important and should be heavily
weighted. To avoid selective reporting of positive findings, an attempt should be made to report
and compare all findings  (all exposure periods and lags, all endpoints, all subgroups, etc.),
although this is admittedly a tall order.

Specific.
1-12, line 17. It is not clear that the type of important evidence would vary by pollutant - it
   would if this were based on the availability of evidence for different pollutants, but I would
   think the important evidence should be pretty much the same.
1-12, line 23. Sometimes there can be an intervention in observational studies: e.g., vitamin
   interventions with ozone, or fish oil with PM.
1-12, line 29. Toxicology studies are also useful in enhancing plausibility of exposure effects.
1-13, line 7.  The  definition of counterfactual is vague.  We are concerned with exposure. We
   observe what happens with exposure; a counterfactual is what we observe happens when
   there is no exposure, all else being equal.
1-15, line 15. While this  is true,  clinical studies also have the potential of overestimating effects
   when exposures used  (concentration or intensity) are seldom experienced in the real world.
1-16, line 15. "discongruities?" - the meaning is opaque here.
1-18, line 16. It's not clear how analogy would not apply to the gaseous pollutants.
1-20, line 5.  Coherence also refers to comparison of effects across study designs and across
   study endpoints.
1-20, line 27. I would argue that not meeting the temporality criterion, when it can be assessed,
   rules out causality.
1-23, line 1.  I don't know what the authors are getting at here in comparing incidence and
   severity.
1-25, line 3.  This discussion is unclear.  If the point is that estimated mean population effects do
   not reflect more extreme effects in individuals that would be more clinically relevant, then
   please say that. If that is not the point,  start again.
Chapter 5.  Modes of action.

Figure 5.5. The emphasis here is on initial pulmonary oxidative stress and inflammation which
"spills over" systemically. There should now also be an arrow going directly from autonomic
modulation to systemic oxidative stress and inflammation, based on the Gonzalez-Flecha work.
Chapter 6.  Acute health effects

1.  Selective highlighting of positive results (a perennial issue); see Ch.l comments (point 3.iii).
   In the review of individual study findings, as in previous CDs, there continues to be an
obvious tendency to report the positive findings, even though these might be one of among
several endpoints.  For example, in the review of controlled human exposure studies on HRV,
                                           705

-------
findings from some generally negative studies (Peretz, Gong [6-18, line 23), etc) are summarized
by highlighting the one or two positive findings selected from a host of the many possible
findings from among the many time windows in which HRV was measured and the many HRV
measures that were analyzed.
   One flagrant example of this tendency is the summary of the Diez-Roux MESA study of
CRP (6-52, line 28) which found no PM effects for lags up to one week. When the averaging
period was extended up to 30 or 60 days, a small effect on increased CRP was observed. Effects
observed over those long averaging periods are probably uninterpretable, however. This is
reported as a "positive" study in the ISA.  This finding is highlighted in the section summary (6-
103, line 8). Here the result was characterized as being "stronger" at these long averaging times,
when in fact it was only present at these averaging times.
   Having said that,  compared to earlier CDs, a better job is done in the ISA of summarizing
effects on a comprehensive  array of endpoints, regardless of whether these were highlighted by
the original authors.

2.  In  reviewing the toxicological findings, it might be better to distinguish instillation studies
from inhalation studies.

3.  The assessments of causal determination  (section 6.2.11; pp. 6.98-109) are not particularly
transparent. For example, what specifically  is the rationale for claiming a causal relationship for
PM2.5, but only that  such a relationship is likely for PM10? And, why is PM10-2.5  merely
insufficient?

Specific.
6-2, line 12. Differences in exposure measurement error may also contribute to between-city
   heterogeneity.
6-3, line 34. Transfer of effects is  only a minor aspect of the measurement error issue in the
   2004 CD, but is the only one touched on here in this overview.
6-8, line 1. Might add how  the ApoE"A mouse model of atherosclerosis differs from  natural
   human atherosclerosis, eg, no coronary atherosclerosis, etc.
6-8, line 19. There is no reference to the Rowan paper.
Section 6-2.
I would consider including discussion of cardiovascular mortality here, even though  Section 6.5
   summarizes mortality in general.
6-38,  line 1. Need to include in this summary the BAD vs. FMD finding.
6-55,  line 24. Why is discussion on a study of oxygen saturation included in the section on
   systemic inflammation?
6-58,  line 14, What is described here is not really hematopoiesis, simply changes in  peripheral
   blood cell counts.
6-59,  line 18. I would say the evidence here is limited and certainly inconsistent.
6-61,  line 20. Although even for vWF, some epi studies showed increases while others showed
   decreases.  Also, only the Liao 2005 study is referred to, not the 2004b or 2007).
6-83,  table 6-2.  The PM2.5 MI Zanobetti and Schwartz reference should be 2006, not 2005.
6-88,  line 13. How "more successful?"
6-106, line 3. It seems that  the Mills (2008)  study of DE that was almost entirely UFP should be
   included as a human clinical study here.
                                          106

-------
6-109. I suggest that a section on PM chemical components (i.e., 6.2.11.5) be included here in
    addition to just sections on the PM size fractions, although that is presented late in the
    chapter.
6-110, Table 6-5. I would not include the 1999 and 1998 studies of Boezen and Forsberg in this
    table.  This table is for more recent studies.
6-112, line 4.  Should be Table 6.7, not 6.6.
6-215, line 19. PM10 findings were not robust to NO2 in the Canadian multi-city study (6-206,
    line 27).
6-225, line 26. Regarding Villeneuve 2003, this is not a cohort.  Line 29 indicates a 5.4%
    increase in CV mortality for same day lag, but Fig 6-24 instead shows the lag  1 effect.  Also,
    as opposed to what is stated, the PM2.5 effect is not quite significant (see Table 4 of the
    Villeneuve 2003 paper). For Fig  6-23 regarding PM2.5, for total mortality at same day lag,
    the effect should be negative (Table 3 of Villeneuve 2003), for CV mortality for lag 1 the
    effect should be at least as negative as -1.0%, and for respiratory mortality for lag 1 the
    graphed effect looks close to 3%, but this is not correct either. I would recommend checking
    the math that converts to the  10|j,g/m3 increment and ensure that the graphed effects are
    correct.
6-251 (section 6.6.2.4). Several of the toxicological endpoints described here are  associated with
    long-term exposure and belong in Ch.7 only.

Chapter 7. Chronic  health effects

1.  This chapter presents a reasonably balanced presentation of findings, with exceptions.  The
cardiovascular and birth outcomes sections are particularly good in this regard.

2. Regarding birth outcomes, specifically, one point that I did not see a discussion of was the
potential for confounding by season.  Air pollution and birth outcomes vary by season.
Matching on time of the year of birth, for example, does not adequately address this.  In both
cohort studies and studies that use birth certificate data, the proper controls for preterm births, for
example, are not births occurring at the same time of the year as cases, but rather gestations of
approximately the same duration. The influence of this potential source of confounding needs to
be considered for the relevant studies.

3.  The Respiratory Effects section (7.3) is an exception to the generally excellent nature of this
chapter.  There is a tendency to report primarily positive endpoints from studies with multiple
endpoints, as well as a tendency to report as positive findings that are unquestionably negative;
e.g., p.7-28, line 33 on asthma in the Kim 2004 study. The Pulmonary Function (7.3.2) section,
especially, is poorly written and presents extraneous information (e.g., Oftedal study description
p.7-33). Unjustified claims are made; e.g., about lack of reversibility and about the meaning of
different lung function measures (p.7-34, lines 18-21).  There is too much relaying of individual
authors' interpretation of their own findings. Unjustified summaries are made; e.g., Pulmonary
Inflammation (p.7-50), especially regarding woodsmoke effects (line 22).  The overall
conclusions regarding causality in this respiratory section, however, are reasonable.
                                           707

-------
4.  My preference would be to include discussion of cohort studies that include cancer mortality
endpoints (eg, Pope 2002) in the section that is currently limited to cancer incidence (section
7.5), rather than placing that discussion in the following section on mortality.

5.  Regarding mortality, an example of how further analyses modify initially-reported dramatic
findings is provided by the Beelen 2008 study of traffic-associated mortality in the Netherlands.
Effects were dramatically lower than those estimated based on an earlier portion of the cohort
(Hoek 2002). The Beelen study suggests that traffic-associated effects may not be greater than
those due to PM itself, a finding at odds with that in the initial report.  This perspective is not
provided here, but is appropriate in a discussion of source- and component-specific effects. On
another issue relating to mortality, section 7.6.7 on exposure time windows does not include the
Krewski  et al assessments on this question using the ACS cohort.

6.  I would dispute that an expert elicitation provides much if any additional evidence or support
(7-115, line 18).

Specific.
7-4, line  11. It is not appropriate to report estimated increases in CAC here in the Diez-Roux
    study as if these were meaningful, in light of the wide CIs that essentially center on 0, or no
    effect. At the least, these should be reported as null findings.
7-16. It is not clear why Table 7-1 shows only some selected endpoints, eg, for Puett et al.
7-21, line 9. "Modification" does not belong here.
7-32, Fig. 7-4. MMEF is synonymous with FEF25-75.
                                           108

-------