August 16, 2004


Honorable Michael O. Leavitt
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, DC 20460

       Subject: Clean Air Scientific Advisory Committee (CASAC) Particulate Matter (PM)
               Review Panel's Ongoing Peer Review of the Agency's Fourth External Review
               Draft of Air Quality Criteria for Particulate Matter (June 2003)

Dear Administrator Leavitt:

       EPA's Clean Air Scientific Advisory Committee (CASAC), supplemented by expert
consultants — collectively referred to as the CASAC Particulate Matter (PM) Review Panel
("Panel") — met in a public meeting held in Research Triangle Park (RTF), NC, on July 20-21,
2004, to discuss follow-on matters related to its ongoing peer review of the two-volume, June
2003 draft document, Fourth External Review Draft EPA Air Quality Criteria for Particulate
Matter (EPA/600/P-99/002, aD, bD).  The current Panel roster is found in Appendix A of this

       This meeting was, in part, a continuation of the CASAC PM Review Panel's review of
the Fourth External Review Draft of the Air Quality Criteria Document (AQCD) for PM in the
current cycle for reviewing the National Ambient Air Quality Standards (NAAQS) for PM. As
noted below, the Panel held extended discussions with EPA staff members on the plans for the
completion of the AQCD for PM. The revised draft Chapters 7  (Toxicology of Particulate
Matter in Humans and Laboratory Animals) and 8 (Epidemiology of Human Health Effects
Associated with Ambient Particulate Matter) of this draft document were provided to the Panel
and the public in mid-June, and a completely-revised Chapter 9  (Integrative Synthesis) was
similarly provided on June 28, 2004.

       The focus of this July  20-21 meeting was for the CASAC PM Review Panel to review
these revised Chapters 7, 8, and 9 and the associated appendices of the AQCD for PM.  After
extended discussion, the Panel concluded that these chapters had been  sufficiently improved that
it could close on Chapters 7 and 8, with the understanding that the Agency's National Center for
Environmental Assessment (NCEA)-RTP will make further revisions as necessary to address the

issues raised both in this report and in the Panelists' individual review comments provided in
Appendix B of this report.

       However, although the Panel considered the rewritten Chapter 9 to represent a very good
initial effort toward developing an integrated synthesis of the information in the AQCD for PM,
the chapter was judged to require sufficient additional revisions that the Panel felt a subsequent
review would be necessary before it could come to closure on this chapter.  In addition, if there is
to be an Executive  Summary (which was not provided in these latest revisions) to the document,
this would also need to be reviewed by the Panel prior to closing on the entire AQCD for PM.
Nevertheless, with  appropriate attention, the Panel thinks that the current version can be brought
to a point where it will be adequate for closure in a relatively short time.

1. Background
       The CAS AC was established under section 109(d)(2) of the Clean Air Act (CAA or
"Act") (42 U.S.C. 7409) as an independent scientific advisory committee, in part to provide
advice, information and recommendations on the scientific and technical aspects of issues related
to air quality criteria and national ambient air quality standards (NAAQS) under sections 108 and
109 of the Act.  Section 109(d)(l) of the CAA requires that EPA carry out a periodic review and
revision, where appropriate, of the air quality criteria and the NAAQS for "criteria" air pollutants
such as PM. The CAS AC, which is administratively located under EPA's Science Advisory
Board (SAB) Staff Office, is a Federal advisory committee chartered under the Federal Advisory
Committee Act (FACA), as  amended, 5 U.S.C., App.

       EPA is in the process of updating, and revising where appropriate, the AQCD for PM as
issued in 1996.  A detailed history of this  current, ongoing review is contained in the Background
section of the CASAC PM Review Panel's report on this subject from the public meeting held in
Research Triangle Park (RTF), NC, on November 12-13, 2003 (EPA-SAB-CASAC-04-004,
dated February 18,  2004). The Panel's most  recent report on this topic (EPA-SAB-CASAC-04-
005, dated March 1, 2004) was prepared following the public teleconference held on February  3,
2004. Both of these documents can be found on the EPA Web Site at: hjtg7/www.eEa.goy/sab_.

2. CASAC PM Review Panel's Ongoing Review of the Revised  Chapters  7-9 of the EPA Air
Quality Criteria for Particulate Matter (Fourth External Review Draft)

Chapter 7 (Toxicology)
       Chapter 7 was deemed to be substantially improved from the prior version. Numerous
previous content issues have been resolved, and the text is much cleaner.   The major problem
that remains is the Chapter's Integrative Summary section. It is missing various key points such
as the fact that Concentrated Ambient Particles (CAPs) used in most of the cited studies do not
concentrate the gaseous phase and ultrafine particles (< 0.1 • m).  Also, the summary needs to
include a mention that the dosimetry modeling predictions describing the doses  in the animal and
some of the human studies were relatively high and the relevance of the results to real world
exposures is uncertain.

       The draft PM AQCD covers relevant  papers published prior to April 2002. The Panel
discussed the inclusion of papers published after April 2002. For consistency, the Panel

recommends that the Reed et al. (2004) paper not be included.  Although the paper adds to the
overall body of information, it does not make any substantive change in the overall evaluation of
the effects of airborne particulate matter from controlled exposures to animals or people.

       The material presented in Chapter 9 on the potential role of particle-bound water on the
toxicity of the particulate components should be in Chapter 7 rather than Chapter 9. In addition,
the study by Morio et al. (2001) presented in Chapter 9 also needs to be introduced in this
chapter. Chapter 9 can then appropriately reference the Chapter 7 material.  The section
summarizing the information currently available should take care to note the limited amount of
information currently available and the need for follow-up studies.

Appendix 7A
       Appendix 7A on rat to human dose extrapolation provides valuable information for
putting into perspective the relationship between various exposure levels and instillation doses
used in animal studies relative to the comparable kinds of exposure levels or doses that would be
needed in humans. All of the calculated values are presented with a high degree of apparent
precision (two or three significant figures). However, it should be recognized that they are the
results of model calculations and in the absence of further validation should  be used and
interpreted with caution.  Thus, it is likely that no more than two significant  digits should be used
and appropriate caveats provided regarding the likely accuracy of the results.

       Unfortunately, some of the appendix examples have failed to find their way into Chapter
7.  Specifically, there is no material in the new Section 7 of Chapter 7 that relates to the
intratracheal instillation studies for the Utah Valley dust used in human and  animal experiments.
The important issue of dose vs. dose rate should be addressed in this context. This is a major
omission that needs to be corrected.  To their credit, the authors have included in Section 7
examples of inhalation studies with CAPs as well as in vitro experiments and the relevant kinds
of interspecies dosimetric comparisons that enable judgments to be made about the potential
extrapolation of the effects seen in the animal studies.

       The instillation studies need to be also covered in Chapter 7 because  of the discussion in
Appendix 7 A in which the authors of the human and animal studies using the Utah Valley dust
incorrectly reported what would be equivalent exposures or typical community exposures
associated with their instillation doses. For example, instead of a single day or up to one week of
exposure for the human instilled doses as stated by Ohio and Devlin, exposures would need to be
on the order of two months. Since EPA has placed great weight on the apparent similarity of
results between studies with the Utah Valley dust in animals and humans, it  is imperative that
Chapter 7 correctly portray the relevance of these instilled doses to real world ambient
exposures.  Similarly, while Table 7-15 is a useful addition to the new Section 7.7, the discussion
in the text surrounding this Table is still not sufficient to put the in vitro doses  into proper
perspective. Doses are reported in terms of nanograms per cell estimated from information in the
publications, and most readers would interpret these doses as being low rather than high.
However, taking into account the density of the cells on the plates and the mass associated with a
single particle, the lowest doses reported  still involve each cell receiving anywhere from about
25 to 400 particles, which is not a small dose. That being said, Table 7-15 helps to establish a
dose response relationship for the effects of PM on alveolar macrophage phagocytosis.

       Appendix 7A contains information on comparative biological responses that belongs in
the main body of the chapter.  It is exactly for the purpose of appropriately comparing the
findings of animal studies to human study results that the comparative dose modeling is required.
The purpose of the appendix is to provide a foundation for judging the value of animal study
data, with the comparison of animal and human responses included in the body of the Chapter 7.
The health response comparisons belong there, not here.

       The Appendix would be  substantially improved by clearly identifying the very few
biological observations that provide the basis for the extensive mathematical extrapolations. The
text should reflect that some of the calculated parameters such as deposited or retained dose per
unit surface area are unlikely to  ever be actually measured. Thus, all of the table and figure
captions should be reviewed and the words "estimated" or "calculated" be placed in front of any
estimated or calculated quantities.

       The Appendix would be  substantially strengthened by including "measured" values for
comparison with "calculated" values.  For example, "measured" lung burden data from the study
of Wolff et al. (1987), [Alterations in Particle Accumulation and Clearance in Lungs of Rats
Chronically Exposed to Diesel Exhaust, Fund. Appl. Toxicol. 9: 154-166] of rats exposed to
diesel exhaust could be plotted in Figure 7A-8.

       Overall, Appendix 7A provides good and thoughtful discussions.  However, it would be
desirable to include some concluding comments after the individual sections as done in the
summary.  Specifically, it would be useful to more fully  emphasize the complexity  of dosimetric
extrapolations, stressing that this is highly dependent on PM parameters, exposure scenarios,
breathing and activity patterns of different species and — not yet achievable by models —
expected differences between responses of a compromised host vs. healthy host.  The summary
does a nice job in this regard,  and conclusions in between individual sections would strengthen

       Page 3, line 22 and subsequently throughout the appendix and figure legends, the Panel
suggest not to use the term "highly insoluble" particles, but rather to call them  "poorly soluble"
particles. Although this term "highly insoluble particles" has been used in a number of earlier
publications, the consensus in the toxicological community is that these particles should more
appropriately be called "poorly soluble particles" because there is no particle which is highly
insoluble (perhaps iridium is the most insoluble particle so far tested).  The key is to make it
clear that particles whose solubility is not a significant factor in relation to the time scales of
clearance or retention; i.e., the time scales of observation.

Appendix 7B
       A new appendix on ambient bioaerosols has been added to Chapter 7 as Appendix 7B in
response to prior suggestions to  move this material from the body of the Chapter. This appendix
inundates the reader with information about  fungi, bacteria, viruses, pollens, plant fragments, etc.
Unfortunately, most of the information indicates the quantities that are present in the air in
various locations and seldom presents effects of this material on humans or animals. If the
Agency is presenting the information on ambient bioaerosols as part of a case for the interaction
or synergism of these materials with other components of the ambient PM, the authors have an

obligation to report effects seen in animals or humans with these materials and at what levels
these effects are seen. Failure to do so provides an open end for an argument of their importance
when such is probably not the case. For example, Section 7B.2.1 describes atmospheric levels of
cellulose and reports them to be typically less than 1 |ig/m3 in the air and in some locations up to
around 6 • g/m3, yet there is no discussion of studies that have been conducted on the effects of
cellulose in animals or humans. In his comments on this chapter, Panelist (and CASAC
member) Dr. Fred Miller has provided about 25 references to studies where cellulose exposures
to animals or humans were in the mg/m3 or greater range with very little if any effects of the
material on human health outcomes. Thus, if there is going to be emphasis on the presence of
biological materials like cellulose, then there needs to be the commensurate discussion of what is
known with respect to health effects.  This disparity in coverage between exposure and effects is
of concern and needs to be rectified. On a broader basis, the treatment of bioaerosols in the
interpretative summary of PM toxicological findings in Section 7.9 falls short of delivering a
punch line in relationship to the constituents of PM that EPA would propose to regulate in any
standards that are revised or added.

       There are a number of specific suggestions for revisions to the Chapter and its
Appendices that have been provided in the Panel members'  comments that should be considered
as they will further improve the presentation.

Chapter 8 (Epidemiology)
       With respect to  Chapter 8, the Panel  concluded that the revised version has now achieved
a much better balance and is a fair representation of the wealth of epidemiological  studies. The
primer on epidemiology has been substantially improved although it still may provide statements
that are too sweeping and these have been noted in the individual comments.  The chapter
properly offers the view that "correct" models can never be identified and that there is always a
potential for residual confounding.  This proposition is hardly unique to studies of particulate
matter  and health and has not been a barrier to the use of observational evidence in other
contexts. Confounding is of greater concern when effect sizes are small, as in this  case.
However, the sweeping generalizations need to be toned down. The chapter begins with a
description of the "Hill" criteria for causality, but these criteria are in fact not uniformly applied,
either in this chapter or in Chapter 9.

       There is again reference to both respiratory hospitalizations and  mortality (p.9-126, L8)
in the description of the Utah Valley steel  mill closure.  The formal study that directly used the
steel mill closure in the design only involved respiratory hospitalizations. Mortality was only
analyzed using a traditional time series design (Pope 1992).  In that paper it was stated that
average deaths per day  were 3.2% higher when the steel mill was open than when it was closed.
The baseline daily mortality was 2.7 deaths/day, which translates to less than a 0.1  death per day
increase with the steel mill open.  The absence of statistical power here  explains why this
"finding" has received little attention, quite correctly, except in this PM AQCD.  The continued
reference to mortality here in the AQCD is not justifiable, at least without appropriate qualifiers.

       The discussion on gaseous pollutant variables as possibly acting as surrogate measures of
some features of PM composition continues to be illuminating.  However, there is too much
made of this point (pp.  9.229-231, and elsewhere) since it remains highly speculative. As has

been noted previously, a more cogent argument can be made that daily variations in the
concentrations of these pollutants serve as measures of unmeasured features of meteorology,
given that meteorology is the primary determinant of daily concentration changes and, arguably,
a more plausible cause of these health effects than the small daily concentration changes in either
PM or the gaseous pollutants.  There needs to be care to provide such alternative interpretations.

       Section 8.4 is very well written and as will be discussed below, it might be better to use
much of this material in Chapter 9 in place of the comparable section that is currently in that

       There was a discussion whether to consider the inclusion of the Hoek et al. paper (2002).
Clearly  it was critical to include work on the GAM reanalyses and the Panel had also requested
the inclusion  of the Hoek et al. paper. However, there are concerns about the use of several of
the exposure metrics such as NC>2 and black carbon in this paper.  The results using these
exposure metrics need to be more carefully caveated or excluded from the discussion.  Another
problem is with the studies of motor vehicles from periods when leaded fuel was still in use.  The
emissions from those vehicles are very different than the emissions from current generation,
catalytic-converter vehicles and appropriate discussion of this issue is needed.  The Panel
believes that the updated results from ARIES presented by Metzger et al. (2004) should not be
included in this version of the PM AQCD, and that appropriate discussion of the problems of
preliminary publication of partial results  such as in the Tolbert et al.  (2000) paper should be
included in the chapter.

       Two panel members and a number of public commenters argued for the inclusion of the
Koop and Tole (2004) paper which addresses the "correct" model issue and questions the
legitimacy of the model selection procedures used in most of the time-series studies that have
been published.  After a lengthy discussion by the panel and the Agency, it was decided not to
include  this paper, but to expand the discussion of the Clyde et al. (2000) paper which raises
some of the same issues.

       In  conclusion, this revised draft of the PM Criteria Document chapter on the
epidemiological  evidence of the health effects of paniculate matter improves on an already
encyclopedic and generally well written review of the scientific literature published since 1996
on this topic.  The authors have adequately addressed the vast majority of CASAC's prior
criticisms and suggestions for improvements of the previous draft document. The remaining
comments in  the Appendix should be relatively easy to address in developing the final version.

Chapter 9 (Integrative Synthesis)
       Chapter 9 has been completely rewritten based on the discussions of the Panel's meeting
in August 2003, the outline provided to the Panel in September 2003 by NCEA, and agreed to by
the Panel in its teleconference in October 2003.  This initial version was a very good draft, but
clearly suffered from time constraints that prevented full consistency in presentation and style.
The Panel felt that the Chapter was too long and that the length was due to too much discussion
of specific studies rather that synthesizing the detailed material presented in the earlier chapters.
The key is that the information presented in the earlier chapters should be synthesized into a
description of the Panel's current level of understanding with respect to the organizing questions.

       The Panel recommends that most references to specific papers be eliminated except in
those limited cases where specific seminal facts are presented that require a reference. This
approach will require more synthesis and less excessive detail that tend to lose the big picture
thread that should be provided in this chapter.  This was felt to be particularly the case for the
toxicological question (Section 9.2.3) which currently represents about one third of the text in
this chapter.  As a result, it is the Panels' recommendation that an effort be made to limit this
chapter to no more than 75 pages.

       For example, the discussion of the potential role of particle-bound water belongs in
Chapter 7 along with the Morio et al. reference.  Then an appropriate summary can be provided
in Chapter 9 as part of a real  synthesis of the information in Chapter 7.

       There was also strong concern for a lack of "bottom line" conclusions in the Summary
and Conclusion sections.  What is the level of understanding with respect to the questions that
are posed as the integrating basis of the Chapter?   The threads of the discussion need to be
pulled together better so that the reader knows what the takeaway messages are. Clear
conclusions will be particularly important if there is to be no Executive Summary as was
discussed as an option.

       The chapter suffers from an overuse and excessive reliance on the term "coherence"
which is used in multiple ways. Although consistency in results among studies, etc. is helpful in
evaluating the totality of the  evidence, the discussion needs to be more even across the questions
and use established evaluation approaches.  The Panel cautions that the EPA  should not be
developing approaches to the evaluation of evidence (i.e., are based on "coherence") that may
differ widely from the approaches taken in other Agency reviews.  Models for evidence
evaluation are available from reports of the Surgeon General ("The Health Consequences of
Smoking," May 2004) and the National Research Council (NRC),  for example.

       Throughout Chapter 9, there is an excessive use of adjectives such as  "considerable,"
"strong," "very," "extensive," etc. such that the reader gets the impression that a "harder sell" is
being made than what may be warranted by the data contained in the first 8 chapters of the PM
Air Quality Criteria Document.  Although the portions  of the Chapter on epidemiology and
toxicology were deemed too  long because of the insufficient level  of aggregation  of information
from the prior chapters, there was concern that many of the  caveats and uncertainties described
appropriately in the earlier chapter are not adequately reflected in Chapter 9.  Thus, more balance
is required in the synthesis and presentation of the integrated information.

       In some cases the descriptive adjectives and hard-sell approach are also evident in
sections that consistently emphasize the absence of effects associated with "crustal" components
of PM2.5, PMio or PMio-2.5. For example on p. 9-44, line 18, it states that "Also of much
importance, all of the above studies that investigated multiple source categories found a soil or
crustal source that was negatively associated with mortality." The irrational implication that we
would live longer if it were dustier needs some additional discussion, and it may be "of much
importance" as it reveals poorly formulated models. In other cases, coarse metals, coarse wood
smoke and/or pollen are proposed as likely causal factors for adverse health effects in arid

southwestern locations where coarse crustal particles and associated soil-borne fungi or bacteria
might provide an equally plausible or more plausible explanation.

       As indicated above, Section 8.4 was found by several of the epidemiological experts on
the Panel to be a good model for much of the epidemiological discussion that is in Chapter 9.

       One specific issue that was given too brief mention in the chapter, and needs some
additional discussion is the timing of PM exposure with acute cardiac-related health effects and
asthma. These findings appear to be potentially important and the results have implications for
the averaging time and the form of the standard with respect to an acute PM2 5 health standard.  It
was the view of the Panel that the  science is not yet adequate to support promulgation of a sub-
24 hour health standard at this time, but these initial results indicate that more attention must be
given to this issue in the future when more time resolved measurement data will be available.  It
would be helpful for this  chapter to provide a short discussion on these findings.

       In general, it was  felt that the welfare effects of PM and the associated desirability of a
possible secondary standard for PM were dealt with in too cursory a fashion. These welfare
effects include: (1) visibility impairments in urban, rural, and park settings; (2) ecosystem
responses to increased atmospheric deposition of the nutrient substances in PM; (3) direct effects
of PM on materials — such as soiling of painted surfaces, exposed textile materials, etc.; and (4)
potential impacts from PM contribution to climate change. This section also needs clear
"bottom-line" conclusions as to what scientific conclusions about welfare effects should be
drawn from the information presented in this AQCD for PM.

       With respect to visibility and some of these other welfare effects that involve monetary
valuation (see page 9-102 line 25,  page 9-105 lines  15-23, and page 9-106 line 3), a broader
approach may be preferred to blend the relevant information and messages from the newer
studies (e.g., the  cited Denver/Phoenix studies that are about perceptions and attitudes about
what constitutes  adverse conditions) and the older economic studies (where monetary valuation
is emphasized in as an indicator of adversity).

       The PM AQCD does little  in Chapter 4 to meaningfully combine information from the
old studies and the newer studies identified to address relevant questions such as "how much or
what characteristics of impairment are adverse?" and "how adverse is it?"  This limitation carries
on in Chapter 9 on page 105.

       On  page 9-106, lines 1-4, it could be  noted that these local visibility standards have (at
least in the case of Denver) resulted in PM2.5 emissions and concentration reductions.  Given that
there is a very good correlation between visual range and PM2.5 concentrations, it would be
helpful to indicate the approximate PM2.5 mass concentrations that correspond to the various
visual range values that are discussed.  It would be useful to  include that "similar threshold
determinations, convergent on a minimal visual range of 40 to  50 km have also been identified in
visibility standards in Lake Tahoe, the Fraser Valley of British Columbia, and state of Vermont.
Thus there  are more than two locations picking similar  "adverse thresholds" and thus suggesting
that this range might be one at which the public feels comfortable about the air quality.  Such a

conclusion would clearly assist EPA's Office of Air Quality Planning and Standards (OAQPS) in
assessing the need for secondary standards to protect visibility.

       Further, regarding visibility impairment, EPA should note that a sub-daily secondary
standard for non-Class I areas may be desirable and feasible, reflecting that: (1) visibility
impairment is an instantaneous effect of PM2 5; (2) daylight visibility is more important to most
people, and is not adequately addressed by a 24-hour standard; and (3) recent local government
initiatives and public value studies indicate that current PM secondary standards are not
providing the desired levels of protection.

       The authors are to be commended for concisely bringing together the key scientific
information/understanding on vegetation and ecosystems from Chapter 4 as well as identifying
the important data gaps and uncertainties which currently prevent relating ambient
concentrations of PM to ecosystem response.  The discussion of the potential application of the
"critical loads concept" in the U.S. opens  a very important philosophical/scientific door for
environmental protection in the future.  However, since Europe has long ago adopted the critical
loads approach, it is time for the United States to consider very carefully why this approach is so
widely used and widely accepted in Europe but has not really been considered very carefully
here in the U.S. in recent years.  It would be useful to bring forward this  information so that there
can be more vigorous efforts to move in this same direction in the near term future. It would also
be useful to include in the environmental effects summary, the growing body of literature on the
effects of crustal particles (and associated soil-borne fungi and bacteria)  deposition on marine
ecosystems — for example, the association between Aspergillus fungi and coral reef decline.

       The section on materials damage is quite weak. Effects from PM and precursors are
significant, but are considered less here than in the 1996 PM AQCD or the National Acid
Precipitation Assessment Program's (NAPAP) State of the Science and Technology documents.
Although there has not been a lot of new work done since these documents,  it does not mean that
materials effects should be described in such meager terms. It might also be better to state
clearly that "federal research funds have not been available to investigate the materials damage
effects of PM and its precursors since the mid-1980s."  The costs associated with materials
interactions with particulate matter could be better articulated. On page 9-119, in the second to
last sentence, the statement could be made broader. For example: "Available data indicates that
airborne particles can result in increases in the frequency of cleaning, maintenance, or
replacement of exposed surfaces and materials, as well as reduced usefulness and enjoyment of
injured materials (as is the case with stone monuments or dirty buildings)."

       It was the Panel's judgment that with careful editing and revisions to address these
general issues described here and the more specific issues presented in the individual comments
it would be likely that the Panel  could close on Chapter 9. Assuming that a revised version of
this chapter and the Executive Summary, if there is to be one, are provided to the Panel by the

end of August, a teleconference meeting could be held in mid- to late September with the goal of
completing the review of the AQCD for PM.  As always, the CASAC PM Review Panel wishes
the Agency well in this important endeavor.

                                             Dr. Philip K. Hopke, Chair
                                             Clean Air Scientific Advisory Committee
Appendix A - Roster of the CASAC Parti culate Matter Review Panel
Appendix B - Review Comments from Individual CASAC Particulate Matter Review Panelists

     Appendix A - Roster of the CASAC Particulate Matter Review Panel

                     U.S. Environmental Protection Agency
                   Science Advisory Board (SAB) Staff Office
                    Clean Air Scientific Advisory Committee
                   CASAC Particulate Matter Review Panel*
Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Potsdam, NY
      Also Member: SAB Board
Dr. Ellis Cowling, University Distinguished Professor At-Large, North Carolina State
University, Colleges of Natural Resources and Agriculture and Life Sciences, North Carolina
State University, Raleigh, NC

Dr. James D. Crapo, Chairman, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO, and Chief Executive Officer (CEO) of Aeolus Pharmaceuticals,

Dr. Frederick J. Miller, Vice President for Research, CUT Centers for Health Research,
Research Triangle Park, NC

Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT

Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard
Medical School, Boston, MA

Dr. Barbara Zielinska, Research Professor, Division of Atmospheric Science, Desert Research
Institute, Reno, NV
Dr. Jane Q. Koenig, Professor, Department of Environmental Health, School of Public Health
and Community Medicine, University of Washington, Seattle, WA

Dr. Petros Koutrakis, Professor of Environmental Science, Environmental Health, School of
Public Health, Harvard University (HSPH), Boston, MA

Dr. Allan Legge, President, Biosphere Solutions, Calgary, Alberta

Dr. Paul J. Lioy, Associate Director and Professor, Environmental and Occupational Health
Sciences Institute, UMDNJ - Robert Wood Johnson Medical School, NJ

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

Dr. Joe Mauderly, Vice President, Senior Scientist, and Director, National Environmental
Respiratory Center, Lovelace Respiratory Research Institute, Albuquerque, NM

Dr. Roger O. McClellan, Consultant, Albuquerque, NM

Dr. Gunter Oberdorster, Professor of Toxicology, Department of Environmental Medicine,
School of Medicine and Dentistry, University of Rochester, Rochester, NY

Dr. Robert D. Rowe, President, Stratus Consulting, Inc., Boulder, CO

Dr. Jonathan M. Samet, Professor and Chair, Department of Epidemiology, Bloomberg School
of Public Health, Johns Hopkins University, Baltimore, MD

Dr. Sverre Vedal, Professor of Medicine, National Jewish Medical and Research Center,
Denver, CO

Mr. Ronald White, Research Scientist, Epidemiology, Bloomberg School of Public Health,
Johns Hopkins University, Baltimore, MD

Dr. Warren H. White, Visiting Professor, Crocker Nuclear Laboratory, University of California
- Davis, Davis, CA

Dr. George T. Wolff, Principal Scientist, General Motors Corporation, Detroit, MI
Mr. Fred Butterfield, CASAC Designated Federal Officer, 1200 Pennsylvania Avenue, N.W.,
Washington, DC, 20460, Phone: 202-343-9994, Fax: 202-233-0643                    gov)
(Physical/Courier/FedEx Address: Fred A. Butterfield, III, EPA Science Advisory Board Staff
Office (Mail Code 1400F), Woodies Building, 1025 F Street, N.W., Room 3604, Washington,
DC 20004, Telephone: 202-343-9994)

* Members of this CASAC Panel consist of:
       a. CASAC Members: Experts appointed to the statutory Clean Air Scientific Advisory Committee
by the EPA Administrator; and
       b. CASAC Consultants: Experts appointed by the SAB Staff Director to serve on one of the
CASAC's National Ambient Air Quality Standards (NAAQS) Panels for a particular criteria air pollutant.

                  Appendix B - Review Comments from
         Individual CASAC Particulate Matter Review Panelists
       This appendix contains the preliminary and/or final written review comments of
the individual members of the Clean Air Scientific Advisory Committee (CASAC)
Particulate Matter (PM) Review Panel who submitted such comments electronically.
The comments are included here to provide both a full perspective and a range of
individual views expressed by Panel members during the review process. These
comments do not represent the views of the CASAC PM Review Panel, the CASAC, the
EPA Science Advisory Board, or the EPA itself. The consensus views of the CASAC
PM Review Panel and the CASAC are contained in the text of the report to which  this
appendix is attached. Panelists providing comments are listed on the next page, and their
individual comments follow.

Panelist                                                                       Page#

Dr. Ellis Cowling	B-3

Dr. Frederick J. Miller	B-6

Mr. Richard L. Poirot	B-23

Dr. Frank Speizer	B-33

Dr. Barbara Zielinska	B-38

Dr. Jane Q. Koenig	B-41

Dr. Petros Koutrakis	 B-43

Dr. Allan Legge	B-49

Dr. Paul J. Lioy	B-50

Dr. Morton Lippmann	B-53

Dr. Joe Mauderly	B-59

Dr. Roger O. McClellan	B-72

Dr. Giinter Oberdorster	B-80

Dr. Robert D. Rowe	B-88

Dr. Jonathan M. Samet	B-90

Dr. Sverre Vedal	B-94

Mr. Ronald H. White	B-99

Dr. Warren H. White	B-102

Dr. George T. Wolff	B-105

                                  Dr. Ellis Cowling

[Note: Sent via e-mail to CAS AC Chair Dr. Philip Hopke, members of the CAS AC PM Review
Panel, and the CASAC Designated Federal Officer (DFO) at 12:22 PM on August 2, 2004]

       In general, I find substantial merit in this First Draft summary of CASAC comments. I
believe these comments provide valuable guidance for NCEA's further efforts to provide a draft
Air Quality Criteria Document for Particulate Matter that can be accepted in full  by CASAC at
its forthcoming conference call  discussion — now tentatively scheduled for some time in mid

       As befits my special particular role in CASAC, my major concerns about the AQCD for
PM have to do with the need for a more balanced treatment in the AQCD for PM of "Welfare
Effects," and the associated desirability of a "Secondary Standard" dealing with PM effects on
various "Air-Quality Related Values."

       These values include: 1) visibility impacts on human enjoyment of scenic vistas
especially in national and state parks, 2) associated economic impacts on our tourism industries,
3) ecosystem responses to decreased solar radiation caused by regional haze, 4) increased
atmospheric deposition of the nutrient and growth-altering substances in PM (including organic,
oxidized, and reduced forms of nitrogen, sulfur, phosphorus, potassium, and the wide variety
organic nutrients of fine and coarse aerosol particles, 5) direct effects on materials such as soiling
of painted surfaces, exposed textile materials, etc., and 6) the need for greater concern during the
next several decades about "smoke management" in light of the greatly increased risk of wild
fires and the corresponding necessity for increased amounts of controlled burning of forests and
natural areas in parks and other recreational areas.

       Greater attention  should be given in the AQCD to these "Air Quality Related Values" in
rural as well as in urban areas.

       Some of the many excellent and readily available photographs, tables, and figures should
be added to the AQCD to illustrate and quantify such welfare effects as: 1) visibility impairment
at scenic vistas and airports, 2) wild fire impacts on the aesthetic values of landscapes, 3)
wildfire impacts on wildlife populations, 4)  economic data on tourism impacts and smoke
management costs and benefits, 5) the successes of urban areas that have adopted secondary
standards for visibility impairment, and 6) changes in populations of aquatic invertebrates or fish
that are induced by atmospheric deposition of the essential nutrient substances in the aerosols
involved in cloud nucleation and precipitation processes.

       With regard to ideas for inclusion in the summary of individual comments deriving from
the CASC "Consultation on Methods for Measuring Coarse-Fraction Particulate Matter (PMc) in
Ambient Air, Based upon Performance Evaluation Studies Conducted by EPA," permit me to
summarize the two points I made  at near the end of this "Consultation" on Thursday July 22.

Point 1)
       EPA (and many other federal research and monitoring organizations) need to guard
against the tendency to allocate so much of the funds used in field measurement campaigns to
"making careful measurements" and that inadequate funds are available for "scientific analysis
and interpretation" to determine what the measurements really mean.
       As described on pages 282-284 in the attachment to this E-mail message, these
cautionary remarks about problems in field measurement programs were suggested originally by
the late Glenn Cass, formerly of Cal Tech and later of Georgia Tech, on the basis of his career-
long experience in various environmental monitoring programs — programs in which too much
funding was allocated to "measurements" and too little to "analysis and interpretation" of the
data.  Please note on pages 283 and 284, the "Fifteen general and specific reasons why this
happens" and the "Thirteen general and specific things that can be done about it!"
       The reference for this published reviewed paper is: Cowling, E., and J. Nilsson. 1995.
Acidification Research: Lessons from History and Visions of Environmental Futures.  Water Air
and Soil Pollution 85:279-292.
       Please also note especially the suggestion in item 9 on page 284 about a "50:50
distribution" of funding allocations between "measurements" and  "analysis and interpretation" of
monitoring data rather than the (90:10 or 80:20 distribution) that is typical of many monitoring
programs in EPA and other agencies.
       But please also note that an even better suggestion was made by Mary Barber,  former
executive leader with the Ecological Society of America's Executive, who  opposed the "50:50
distribution" idea at a recent Whitehouse Conference on monitoring. Mary Barber insisted, and I
agree with her, that it would be even more appropriate to distribute the funding into three rather
than two categories of investments — with equal shares going to "measurements," "analysis and
interpretation," and "outreach and extension of findings" to interested clientele and "customers"
for the results of field measurement programs.
       This problem is so commonplace — not only in this country  but all  over the world — that
I commend these "lessons that are available to be learned" (and perhaps even the "15 reasons
why this happens" and the "13 things to do about it") for inclusion among the "comments from
individual participants" in the CASAC Consultation on PM Measurement Methods.

Point 2)
       EPA should also guard against the tendency to give undue emphasis to "Data Quality
Objectives" in the selection and evaluation of instruments and subsequent implementation of
field measurement programs to the exclusion of concern about "Science Quality Objectives" and
"Policy Relevancy Objectives."
       Experience within the Southern Oxidants Study and other large-scale field measurement
campaigns have demonstrated repeatedly that undue emphasis on "Data Quality Objectives"
often leads to:
1) Serious lack of attention to the  scientific hypotheses and assumptions that are inherent in any
choice of scientific instruments, the appropriateness of the ground-based sites or aircraft
platforms on which the instruments are mounted, the skills of the instrument operators, the data
processing and data-display programs used, and especially the scientific quality of the
conclusions and statements of findings that are drawn from analysis and interpretation of the
measurements that are made; and

2) Equally serious lack of attention to the policy relevancy of the measurements being made —
relevancy to the general or specific enhancements of environmental protection that are the raison
de etre of the public health or public welfare concerns that led to the decision to establish a
monitoring program or undertake a field measurements campaign in the first place.
       In this latter connection, permit me also to call attention to the "Guidelines for the
Formulation of Scientific Findings to be Used for Policy Purposes" which were developed
originally by the NAPAP Oversight Review Board led by Milton Russell. Please find attached to
this E-mail message, an electronic version of these Guidelines which we have adopted and very
slightly adapted for use in formulating policy relevant scientific  findings in the Southern
Oxidants Study.
       The original version of these Guidelines was published as Appendix III of the April 1999
Report titled "The Experience and Legacy of NAPAP."  This was a Report to the Joint Chairs
Council of the Interagency Task Force on Acidic Deposition of the Oversight Review Board
(ORB) of the National Acid Precipitation Assessment Program.
       As indicated in Appendix III:
       "The following guidelines in the form of checklist questions were developed by the ORB
to assist scientists in formulating presentations of research results to be used in policy decision
processes. These guidelines may have broader utility in other programs at the interface of science
and public policy and are presented here with that potential use in mind."
       These guidelines may also be of value as part of the "communication of individual
comments" from the CASAC Consultation on PM Measurement Methods.
Dr. Ellis B. Cowling, Director
Southern Oxidants Study
North Carolina State University
Raleigh, North

                               Dr. Frederick J. Miller

                                                                     Fred J. Miller, Ph.D.
                                                                            July 21,2004
Chapter 7

General Comments

The fourth version of Chapter 7 on the toxicology of PM in humans and laboratory animals
represents an overall improvement from the third draft version. The Chapter's Integrative
Summary section is still missing various important points such as the fact that CAPs do not
concentrate the gaseous phase and that dosimetry modeling predictions indicate many of the
animal and some of the human studies used doses that call into question the relevance of the
results to real world exposures.
   The Appendix 7A on rat to human dose extrapolation provides valuable information for
putting into perspective the relationship between various exposure levels and instillation doses
used in animal studies relative to the comparable kinds of exposure levels or doses that would be
needed in humans. Unfortunately, some of these examples have failed to find their way into
Chapter 7. Specifically, there is no inclusion in the new Section 7 of Chapter 7 that relates to the
intratracheal instillation studies for the Utah Valley dust used in both human and animal
experiments. This is a major omission that needs to be corrected. To their credit, the authors have
included in Section 7 examples of inhalation with CAPs studies as well as in vitro experiments
and the relevant kinds of interspecies dosimetric comparisons that enable judgments to be made
about the potential extrapolation of the effects seen in the animal studies.
   The reason the instillation studies need to be also covered is that there is a clear pointing out
in Appendix 7A that the authors of the human and animal studies using the Utah Valley dust
incorrectly reported what would be equivalent kinds of exposures or typical community
exposures associated with their instillation doses. For example, instead of a single day or up to
one week of exposure  for the human instilled doses as stated by Ohio and Devlin, exposures
would need to be on the order of two months. Since EPA has placed great weight on the apparent
similarity of results between studies with the Utah Valley dust in both animals and humans, it is
imperative that Chapter 7 correctly portray the relevance of these instilled doses to real world
ambient exposures. Similarly, while Table 7-15 is a useful addition to the new Section  7.7, the
discussion in the text surrounding this Table still does not do enough to put the in vitro doses into
perspective. Doses are reported in terms of nanograms per cell estimated from information in the
publications, and most readers would interpret these doses as being low rather than high.
However, taking into account the density of the cells on the plates and the mass associated with a
single particle, the lowest doses reported still involve each cell seeing anywhere from about 25 to
400 particles, which is not  a small dose.  That being said, Table 7-15 helps to establish a dose
response relationship for the effects of PM on alveolar macrophage phagocytosis.
       A new Appendix on ambient bioaerosols has been added to Chapter 7 and appears as
Appendix 7B. This Appendix inundates the reader with information about fungi, bacteria,
viruses, pollens, plant fragments, etc. Unfortunately most of the information indicates the
quantities that are present in the air in various locations and seldom  presents effects of this
material on humans or animals. If the Agency is presenting the information on ambient

bioaerosols as part of a case for the interaction or synergism of these materials with ambient PM,
the authors have an obligation to report effects seen in animals or humans with these materials
and at what levels these effects are seen. Failure to do so provides an open end for an argument
of their importance when such is probably not the case. For example, Section 7B.2.1 describes
atmospheric levels of cellulose and reports them to be typically less than 1 ng/m3 in the air and in
some locations up to around 6 ng/m3, yet there is no discussion of studies that have been
conducted on the effects of cellulose in animals or humans. I have provided about 25 references
for studies where cellulose exposures to animals or humans were in the mg/m3 or more range
seeing very little if any effects of the material on human health outcomes. This disparity in
coverage is of concern and needs to be rectified. On  a broader basis, the treatment of bioaerosols
in the interpretative summary of PM toxicological findings in Section 7.9 falls short of delivering
a punch line in relationship to the constituents of PM that EPA would propose to regulate in any
standards that are revised or added.

Specific Comments

p. 7-3,1.  17          The  authors refer to analyses contained in Appendix 7A. Some of the
                    highlights of those findings should be presented in the main body of the
                    chapter, perhaps to the extent  of a  page of description of the more
                    notable results. Otherwise, most readers will skip the Appendix and not
                    really be informed of the salient points that the Appendix makes relative
                    to the interpretability of several studies that EPA has  deemed key for
                    concerns about the potential effects of PM exposure to humans.

p. 7-4, ]j 1           The  authors continue to fail to point out that CAPs systems do not allow
                    for concentrating any gaseous components. This means that potential
                    interactions present in the ambient atmosphere are not captured in the
                    same proportionality via the CAPs studies. This point should be captured
                    in the interpretive summary for Chapter 7.

p. 7-28              In Section 7.2.3, there has been no attempt to bring in information from
                    the extrapolation appendix to put the ROFA exposures into perspective.
                    Based upon the extensive calculations that were done by EPA for
                    Appendix 7A, this certainly could  and should be  done because it will
                    clearly demonstrate that most  of these studies are not providing
                    information relevant to anywhere near real world exposure levels.

p. 7-45,1.15        The  study by Bouthillier used resuspended Ottawa UAP ambient PM on
                    the size of 4-5 mm in MMAD. As Appendix 7A shows (Figure 7A-4)
                    only about 60% of the material would have been inhalable and most of
                    this would have been trapped in the nose. Whether this supports or does
                    not support a lack of effects seen in the study is not as important as the
                    need to put some of these kinds of study results into perspective from a
                    dose viewpoint.

p. 7-49, last  1|        The  point made here relative to the fact that the high exposures contained
                    in many of the studies may overwhelm lower dose mechanisms and that
                    the high dose instillation studies may  actually produce different effects

                    on the lung than inhalation exposures at lower concentrations or doses is
                    an important point. However, this point does not make its way to the
                    interpretive summary section of Chapter 7 (i.e., Section 7.9), but it

p. 7-50,1J 2          This discussion of the properties of ROFA and the metal content that
                    leads to toxicity is a better description of the body of evidence around
                    ROFA than is contained in Section 7.9 on p. 7-24 where ROFA effects
                    attribution are discussed. Consider reworking of the material on  p. 7-24
                    to better reflect that contained on p. 7-50.

p. 7-118             The discussion of potential importance of particle charge may be
                    somewhat superfluous because ambient aerosols to which humans are
                    exposed have come to Boltzman equilibrium. Charge would only be
                    important for freshly generated material such as in a workplace
p. 7-123,1. 9         This paragraph  describing the Churg et al. study is excessive in length.
                    The same points could be made in a shorter paragraph.

p. 7-126,1. 1         In the final version of this chapter, EPA should make an attempt to not
                    have major sections begin on a separate page but rather conserve space
                    by having them flow immediately from the conclusion of the preceding

p. 7-126,1. 14       The statement that the laboratory generated particles may be of limited
                    value because of uncertainties in extrapolating between laboratory
                    generated and ambient particles seems to have escaped comment in
                    earlier versions of this chapter. This particular sentence is too strong a
                    statement in the view of this reviewer since dosimetry models allow for
                    an interspecies dosimetric adjustment of any findings. The more relevant
                    point the authors were probably trying to convey was that studies of
                    individual particles may not provide the overall picture compared to the
                    complex mixture of exposures that occur in urban aerosols.

p. 7-155,1.18       While the Ottawa UAP and ROFA studies are at unrealistic exposure
                    levels and probably don't have much relevance for evaluating current
                    ambient PM levels, the authors should note that these studies were
                    conducted in healthy animals. Thus, the potential certainly exists for
                    lower exposures in compromised animal models to show effects at much
                    lower exposure levels.

p. 7-160,1. 4         It is incorrect to refer to instilled material as exposure concentrations. It
                    was an administered dose of 3mg/kg in the particular sentence. The
                    entire Chapter 7 should be checked for instances of inappropriate
                    wording of exposure concentration when really the correct terminology
                    would be administered dose.

p. 7-164,1. 25       Would suggest extending this sentence instead of stopping at "are
                    identical" to state "are identical because of variations in individual

                    breathing patterns, lung anatomy, and particle deposition fractions.

p. 7-166,1. 10       The paragraph is a good example of how dosimetry models can be used
                    to make interspecies comparisons of potential effects after adjusting for
                    dose differences. However, in this particular paragraph, the authors
                    should note that their statements about the ng/m2 of delivered material
                    are related only to the insoluble part of PM since the dosimetry model
                    upon which these doses were calculated considers only insoluble
                    particulate matter. Thus, a caveat should be included to indicate not only
                    this but that the accounting of the soluble  fraction could lead to a more
                    complex statement of potential differences between rats and humans.
                    Moreover, in this paragraph, the authors should be more cautious about
                    ascribing significance to a 60-500% increase in PMNs in the rat
                    compared to about a 40-fold lower dose (mass/m2) yielding a 267%
                    increase in humans. The basal level for the percent increase greatly
                    influences the calculated percent increase

p. 7-167,1. 26       Please clarify if the statement "soluble and insoluble components" in
                    parenthesis is made to imply PM2.5 was the soluble and PMiothe
                    insoluble. If this was what was intended, it is an incorrect
                    characterization because both the fine and coarse modes contain soluble
                    as well  as insoluble components although there is a greater percentage of
                    constituents of the fine mode that are soluble..

p. 7-168,1. 25       It should be 500 ng instead of 500 ug.

p. 7-169,1. 2         Unless the abbreviation for right angle light scatter is used subsequently,
                    it should be deleted.

p. 7-169,1. 3-4      Again, |j,g should be used in place of ug.
p. 7-170             Table 7-15 provides an interesting set of information concerning
                    phagocytosis effects of particulates in humans and rodents. The title of
                    table could probably be shortened to something such as "Interspecies
                    Comparisons of Particle Effects on Alveolar Macrophage Phagocytosis".
                    For the Goldsmith et al. study entry with 0.4 as the estimated dose per
                    cell in terms of nanograms, the estimated  percent filling of 140% is an
                    impossible value and should either be deleted or asterisked to note that it
                    is not possible to obtain such a value. In the Van Eden portion of the
                    table, there appears to have been a data entry for the 0.02 ng value where
                    the as percent filling is 2.121. If this is indeed computed,  it certainly
                    should be rolled back to 2.1%. However, there is no data  entry with 0.2
                    ng, and this value for percent estimated of cell filling needs to be

p. 7-171,1. 7-8      The units here should be ng/cell not mg/cell if the entries in Table 7-15
                    are indeed in units of nanograms per cell.
p. 7-181             Add an "s" to sampler.

p. 7-182,1. 23        The authors state that the studies they have discussed provide strong new
                    evidence indicative of ambient PM having mutagenic properties. To this
                    reviewer, this is an overstatement of the body of data and the
                    interpretation of such, particularly since most of the work involved
                    transform cells whereas normal human epithelial and other lung cells are
                    not easily mutated.

p. 7-183,1. 21        Insert a comma after assay and remove the left-hand parenthesis in front
                    of Houk putting it  in front of the year and also removing the comma after
p. 7-186,1. 3         This reviewer would hardly call studies conducted in the late 1980s as
                    more recent studies.

p. 7-200,1. 5         The authors use the words extensive credible evidence relative to the
                    biologic plausibility and potential mechanisms relating to the ability of
                    ambient PM to be  associated with lung cancer. This seems like an
                    overstatement and a better characterization would be that they provide
                    some credible evidence.

p. 7-202,1. 1         In various places, the authors have indicated that rats clear particles
                    faster than humans and this is the reason for the higher doses. While this
                    is indeed the case relative to clearance, rats  also have a lower deposition
                    fraction than do humans for most particle sizes, which in itself also leads
                    to higher exposure levels being needed in animals compared to humans.
                    This should be pointed out here and in various other places in Chapter 7
                    and in Appendix 7A.

p. 7-203,1. 23        The assumption by Ohio and Devlin of an average ventilation of 15L/min
                    is not supportable  either from EPA's documentation of typical activity
                    patterns and ventilation rates nor from the published literature.

p. 7-203,1. 27        Why have the authors of Chapter 7 not put the statements by Ohio and
                    Devlin of comparable exposure levels into perspective as was done in the
                    extrapolation appendix on page 7A-45 where it was clearly shown that
                    between 44  and 65 days would be required for a person to deposit the
                    instilled dose that was used in these studies? An integration of the results
                    from the extrapolation appendix have not found their way into the
                    interpretations in Chapter 7. This needs to be corrected.

p. 7-206,1. 25        The caveat about relevance of ROFA studies to real world exposures is
                    quite weak here compared to what is expressed on p. 7-161. The tone in
                    the earlier sections is what should be conveyed in Section 7.9.

p. 7-210,1. 3         Strike "previously".
p. 7-211,1. 2         Insert "of after production.

Appendix 7A

Specific Comments

Table of Contents
In the Table of Contents, Multiple Path Particle Deposition Model should be referred to as
Multiple Path Particle Dosimetry Model.

Entry 7A.5          Should read Dosimetric "Calculations" instead of "Calculation".

Figures/Figure Legends
In the list of figures and in the figure legends themselves, suggest replacing Human with Humans
and Rats for Rat.

p. 7A-3,1. 1         Remove the hyphen in inter-subject.

p. 7A-3,1. 12        Suggest rewording to be "in a lung region may be expressed as".
p. 7A-4,1. 1         Insert a comma after comparisons.

p. 7A-5,1. 14        Should be exposures not exposure.
p. 7A-6,1. 20, 22     Again, deposition should be replaced by dosimetry in the definition of
                    multiple path particle dosimetry model.
p. 7A-6,1. 24        It should be RIVM not RIM.

p. 7A-8,1. 5         Should be macrophages not macrophage.

p. 7A-8,1. 17        As stated, the sentence after the i.e. is incorrect. What the authors are
                    probably intending to say is that mucus itself is not reabsorbed in the
                    respiratory tract. However, mucus secreting cells are located throughout
                    the conducting airways and there  is secretion at every level, even though
                    there is reabsorption of water, otherwise there would be total occlusion
                    of the trachea with a cumulative flow of the secretions from the lower
                    respiratory tract.

p. 7A-13            In Panel A where mass is presented, what does the number 100 in
                    parenthesis stand for under the title of Resuspended?
p. 7A-14            Insert "of after distribution.

p. 7A-15,1. 5        Suggest adding in parentheses reference back to Chapter 6 where it was
                    noted that neither the ICRP nor the MPPD model should be used for
                    particles < .01 due to not taking into  account the effective actual
                    dispersion of particles that size.
p. 7A-18,1. 7        In the section on retention, it would be worth noting that clearance is
                    indirectly represented in the curves because you are examining the
                    amount retained and therefore the amount compared to the previous time
                    is the amount that was actually  cleared. Some readers may find it

p. 7A-19
p. 7A-20,1. 2-3

p. 7A-23

p. 7A-25
 Section 7A.4.6
p. 7A-26
 Section 7A. 5
p. 7A-27

p. 7A-27

p. 7A-30
p. 7A-31

p. 7A-37

p. 7A-42,1. 10

p. 7A-45,1. 10
confusing to be interchanging clearance with retention.

The legend for the figures is incorrect in that it does not delineate the
postexposure time in hours only represents the value after 1 hour such
that m/nio for the first hour represents the rise in concentration during the
hour of exposure. The figure needs to be corrected.

The figure contains both tracheobronchial and alveolar data so the
sentence here should be TB or A region in both places where currently
only TB region is specified.

Delete "of in the last line of the figure legend.

This material does not need to be a separate  subsection. It could be
combined with the section on  normalizing factors and a new title given.
This is particularly the case since no text is really contained with the
section and it is merely a summary table.

For the entire section on comparing rats to humans, the authors should
make clear that the various scenarios they have modeled invokes the
assumption that "a particle is a particle" and that the dose mass can be
aggregated over multiple size  modes. In reality, it is extremely unlikely
that this assumption could be met since the PM is comprised of many
types of particles each having their own size distribution. Rather the
authors should make it very clear that the kinds of calculations presented
are for illustration of how you would go about computing dose if
definitive information on speciation, size distribution, etc. were  available
such as may be forthcoming from the EPA PM  supersite monitoring

In the first sentence  after equation 1 Ib, strike "the variety of.

For the example where  the EqER ranges from 0.9 to 5.5, suggest putting
in parentheses, "see first section of Table 7A-7a" to help orient the

For the EqER values contained here, it would help the reader to  locate
the areas of the table if Table 7A-7a and 7b had roman numeral  sections
added to them such as I, II, III, IV, and V so for example, deposited mass
might be I, retained mass in the TB might be II, surface area of particles
retained in the alveolar  would be III, etc.

Add "of after values.

On last line of page, insert "of after 10 years.

Insert a comma after species.

Isn't the point being made using insoluble particles that the computations
of Ohio and Devlin relative to the  amount of time that a person would
have to be exposed is incorrect? Whether the appropriate dose
represented the soluble  or insoluble fraction is not really the issue. The
authors could illustrate  this difference by adding an additional sentence

p. 7A-46,1. 16

p. 7A-50,1. 21

p. 7A-54,1. 27
p. 7A-54,1. 30
p. 7A-56,1. 4

p. 7A-58,1. 25

p. 7A-59
wherein they only use 20% but considered it insoluble just to see what
kind of time period exposures would have to be to get the dose that Ohio
and Devlin instilled.

Insert "to" after equal.

Should be overload is "affected".

The statement concerning not being able to simulate chronic retention in
humans except under conditions of overload for rats is not correct. This
statement is only true for high exposure levels in humans. However,
ambient exposures of the 10 to 20 ng/rn3 that are being considered as part
of the PMio revisions should be able to be estimated by comparable rat
equivalent exposure levels.

The thrust of the ILSI Workshop (Inhal. Toxicol. Vol. 12, 2000) makes it
quite clear that overload of poorly soluble particles is dependent on
coexistent active inflammation and cell proliferation, which is a high
dose phenomenon, and that effects such as lung cancer would not be
expected at typical ambient exposure levels. The same could be said for
various noncancer effects.  The authors are making a significant and
speculative "leap of faith"  that rats in lung overload may simulate the
response of susceptible or impaired humans. Since this topic is not
treated in the main body of the chapter, which it should be if there was
any evidence for the speculation made by the authors, this material
should be omitted or substantially toned down.

It should be "affected" by PM.

I believe the authors are intending to used criticized instead of critiqued.

The Conclusion section should contain several additional bullets that
specifically relate to the instillation studies of the Utah Valley dust and
the disparity in results compared to the authors as far as equivalent doses.
In addition, the CAP studies and their closer similarity for extrapolation
purposes should be highlighted.
Appendix 7B

Specific Comments
p. 7B-9
The Celenza study did not characterize the concentration of pollen or any
size distribution. So how can the author contend that following a
thunderstorm the increase in ER visits is due to pollen levels? This
reviewer questions the value of including this study. If one of the
publications cited on p. 7B-14 concerning allergens following
thunderstorm and asthma-like effects has quantitative data on
concentrations or particle sizes, this reviewer would suggest one of these

                    references be included in the table rather than the Celenza publication.
p. 7B-9             The title of the table is Respiratory Effects of Pollen/Fungi in PM
                    Exposures. However, there is little if any information in the table relating
                    to the PM exposures so that the table could just as well be labeled
                    Respiratory Effects of Pollen/Fungi and Ozone or Respiratory Effects of
                    Pollen/Fungi and Sulfur Dioxide, etc. The point being that the table more
                    reasonably should be labeled Respiratory Effects of Pollen/Fungi and Air

p. 7B-10            In an effort to be complete, the authors have gone overboard by including
                    the Pieckova and Kunova (2002) study. This experiment did not measure
                    concentration or particle size of any of the filamentous fungi in their in
                    vitro study of chicken tracheal rings. Yet the report is that it stopped
                    tracheal ciliary movement. If such conditions were prevailing in humans,
                    the epidemiology would be a whole lot clearer than it is. This study adds
                    little value in the opinion of this reviewer and should be deleted.

p. 7B-12            Section 7B.2.1 on atmospheric levels of cellulose and other plant debris
                    markers indicate that the cellulose levels, while being in some locations a
                    significant constituent of PM, are themselves only a few ng/m3. In
                    contrast to this, the toxicological studies on cellulose in the mg/m3 range
                    show very limited responses if any. These studies have been done in both
                    animals and humans. If the agency is invoking the contribution of
                    bioaerosols in an interactive form with PM, then there is an obligation to
                    put the known effects of bioaerosols into perspective. This has not been
                    done. To assist the agency relative to cellulose, a number of citations are
                    provided that have studied the effects of cellulose in animals or humans.
Adamis, Z., Tatrai, E., Honma, K. and Ungvary, G. (1997). In vitro and in vivo assessment of the
   pulmonary toxicity of cellulose. J. Appl. Toxicol. 17, 137-141.
Adamson, G. M., Muhle, H., Creutzenberg, O., Bellmann, B., Dasenbrock, C. (1999). Effects of
   different cellulose-containing respirable samples in the lung of Fischer 344 rats. SOT99.
Anderson, R. L., Owens, J. W., and Timms, C. W. (1992). The toxicity of purified cellulose in
   studies with laboratory animals. Cancer Lett. 63, 83-92.
Scientific Committee on Toxicity, Ecotoxicity, and the Environment (CSTEE). (2002). Opinion
   on risk to human health from chrysotile asbestos and organic substitutes. 35th CSTEE plenary
   meeting, Brussels, December 17.
Cullen, R. T., Searl, A., Miller, B. G., Davis, J. M. G., and Jones, A.  D. (2000). Pulmonary and
   intraperitoneal inflammation induced by cellulose fibres. J. Appl.  Toxicol. 20, 49-60.
Cullen, R. T., Miller, B.  G., Clark, S., and Davis, J. M. G. (2002). Tumorigenicity of cellulose
   fibers injected into the rat peritoneal cavity. Inhal. Toxicol. 14, 685-703.
Davis, J. (1996). The toxicity of wool and cellulose fibres. J.  Occup.  Health Safety 12, 341-344.
Ericsson, J., Jarvholm, and Norin, F. (1988). Respiratory symptoms and lung function following
   exposure in workers exposed to soft paper tissue dust. Int.  Arch. Occup. Environ. Health 60,

Hadley, J. G., Kotin, P., and Bernstein, D. M. (1992). Subacute (28 day) repeated dose inhalation
   toxicity of cellulose building insulation in the rat. The Toxicologist 12, 225.
Harrison, P. T. C., Levy, L. S., Patrick, G., Pigott, G. H., and Smith, L. L. (1999). Comparative
   hazards of chrysotile asbestos and its substitutes: a European perspective. Environ. Health
   Per sped. 107,607-611.
Kraus, T., Pfahlberg, A., Zobelein, P., Gefeller, O., and Raithel, H. J. (2004). Lung function
   among workers in the soft tissue paper-producing industry. Chest 125, 731-736.
Lanes, S. F., Cohen, A., Rothman, K. J., Dreyer, N. A., and Soden, K. J. (1990). Mortality of
   cellulose fiber production workers. Scand. J. Work Environ. Health 16, 247-251.
Love, R. G., Smith, T. A., Gurr, D., Soutar, C. A., Scarisbrick, D. A., and Seaton, A. (1988).
   Respiratory and allergic symptoms in wool  textile workers. Br.  J. Ind. Med. 45, 727-741.
Milton, D. K., Godleski, J. J., Feldman, H. A., and Greaves, I. A. (1990). Toxicity of
   intratracheally instilled cotton dust,  cellulose, and endotoxin. Am. Rev. Respir. Dis. 142,
Muhle, H., Ernst, H., and Bellmann, B. (1997). Investigation of the durability of cellulose fibres
   in rat lungs. Ann. Occup. Hyg. 41 (suppl. 1), 184-188.
National Council for Air and Stream Improvement (NCASI). (2004). Characterization of
   particulate matter in workplace atmospheres of paper making and converting operations.
   Technical Bulletin No. 873, NCASI Regional Center, Gainesville, FL, March.
Pauly, J. L., Allaart, H. A., Rodriguez,  M. L, and Streck, R. J. (1995). Fibers released from
   cigarette filters: an additional health risk to  the smoker. Cancer Res. 55, 253-258.
Pauly, J. L., Stegmeier, S. J., Allaart, H. A.,  Cheney, R. T., Zhang, P. J., Mayer, A. G., and
   Streck, R. J. (1998). Inhaled cellulosic and plastic fibers found in human lung tissue. Cancer
   Epidemiol. Biomarkers Prev. 7, 419-428.
Rosenbruch, M., Friedrichs, K. H., and Schlipkoter, H. W. (1992). Health significance of the
   fibrous asbestos substitutes used in the production of fibre-reinforced cement. Zbl.
   Arbeitsmed. 42, 355-362.
Solet, D., Zoloth, S. R., Sullivan, C., Jewett, J., and Michaels, D. M. (1989). Patterns of mortality
   in pulp and paper workers. J. Occup. Med. 31 (7), 627-630.
Tatrai, E., and Ungvary, G. (1992). The aetiology of experimental  fibrosing alveobronchiolotis
   induced in rats by paprika dust. Br. J. Ind. Med. 49, 494-498.
Tatrai, E., Adamis, Z., Bohm, U., Meretey, K., and Ungvary, G. (1995). Role of cellulose in
   wood dust-induced fibrosing alveo-bronchiolotis in rat. J. Appl. Toxicol. 18, 45-48.
Thoren, K., Jarvholm, B., and Morgan, J. (1988). Mortality from asthma and chronic obstructive
   pulmonary disease among workers in a soft paper mill:  a case referent study. Br. J. Ind. Med.
   46, 192-195.
Warheit, D. B., Snajdr, S. L, Hartsky, M. A., and Frame, S. R. (1998). Two-week inhalation
   study in rats with cellulose fibers. In Advances in the Prevention of Occupational Respiratory
   Diseases. (Chiyotani, K., Hosoda, Y., and Aizawa, Y., editors), pp. 579-582. Elsevier, Tokyo.
Warheit, D. B., Hartsky, M. A., and Webb, T. R. (2000). Biodegradability of inhaled p aramid
   respirable fibre-shaped particulates: representative of other synthetic organic fibre-types? Int.
   Arch. Occup.  Environ. Health 73 (Suppl), 575-578.
Warheit, D. B., Hartsky, M. A., Reed, K. L., and Webb, T. R. (2001). Biodegradability of
   inhaled para-aramid respirable-sized fiber-shaped particulates: mechanistic in vivo and in
   vitro studies.  Toxicol. Appl.  Pharmacol. 174, 78-88.

Warheit, D. B., Reed, K. L., and Webb, T. R. (2001). Man-made respirable-sized organic fibers:
   what do we know about their toxicological profiles? Ind. Health 39, 119-125.
Warheit, D. B., Reed, K. L., Pinkerton, K. E., and Webb, T. R. (2002). Biodegradability of
   inhaled p-aramid respirable fiber-shaped particulates (RFP): mechanisms of RFP shortening
   and evidence of reversibility of pulmonary lesions. Toxicol. Lett. 127, 259-267.

p. 7B-14,1. 8        Delete "which" and replace it with "that" or else insert a comma after

p. 7B-19,1. 3        The 3 on O should be a subscript, not a superscript.

p. 7B-20,1. 4        Do the authors really mean macro organisms or do they mean micro

p. 7-24,1. 1          The study by Vogelzang describes effects of endotoxin related to decline
                    in pulmonary function obtained by regression analysis. There is no
                    discussion of what variables were controlled for in the regression and this
                    should be noted to insure that other variable were controlled for.

p. 7B-24,121        The Zock study showed a decline in FVE1 of 3% during a shift a potato
                    processing plant. This decline may have been statistically significant but
                    certainly doesn't meet the American Thoracic Society criteria of
                    clinically  significant change. The question arises as to whether Zock
                    looked at multiple days and found the same decline or was this a one
                    time event?
p. 7B-28,1.15       Check the ±147ng/m3 for fidelity with the original publication.  If this
                    number is indeed correct, it shows significant lack of control of the
                    experiment and certainly in this reviewer's mind draws into the question
                    the results.

p. 7B-29,1. 20       The results of the phadioatop test being statistically significant when
                    analyzed for age represents a post hoc analysis and can only be
                    considered as exploratory.  It is troubling to this reviewer that many of the
                    studies appear to analyze for every possible combination and are not
                    based upon a testable hypothesis.
Chapter 8

Specific Comments

p. 8-217,1. 5         The description of the results for which models would be selected using
                    the BIC is misleading. As described in public comments (Moolgavkar, p.
                    8 at the bottom), the authors should specifically state that the AIC and
                    BIC only both give a RR of 1.05 when models that included PM were
                    included as a covariate. Otherwise the top 25 models had under AIC an
                    estimated RR of 1.05 and under BIC an estimated RR of 1.015 when the
                    Bayesian Model Averaging approach was used. This is a much different

                     depiction of the results compared to the way the authors of Chapter 8
                     described this study. Any changes should also find their way to Chapter

p. 8-217,1. 18        It should be p < 0.05 not p < 0.05.

p. 8-219,1. 19        The discussion indicates that the GAM-default vs. GAM-stringent
                     analyses for the most part do not lead to significant changes in the risk
                     estimates. Later on in the synthesis part of the chapter, it is argued that
                     the multi city studies have greater precision. Yet there is no note that in
                     Table 8-36 the NMMAPs 90 city study went from being statistically
                     significant for the PM effect in the GAM-default analysis compared to
                     the GAM-stringent analysis wherein the effects of PMi0 were no longer
                     significant. To this reviewer, this is a striking discrepancy of the
                     interpretation of the overall body of work given the reliance later on
                     multi city studies being viewed as more precise and carrying greater
                     weight. That this loss in significance may be due to a biased downward
                     value in the standard error may be part of the explanation for this but on
                     the next page it is stated that the  GLM with natural splines that was used
                     in the reanalysis study, which I presume is the GLM entry in Table  8-36,
                     does not have a bias of the standard error; yet for this GLM analysis of
                     the NMMAPs studies, the PMio effect is still not statistically significant.
                     This situation is further confused by the plot of the NMMAPs data in
                     Figure 8-15 where the error bars around it show that all three analyses
                     methods are statistically significant. So which is correct, and what is this
                     reviewer missing?

p. 8-223,1. 30        Shouldn't it be p-values > 0.05 not > 0.5? And how does this compare
                     with the statement on page  8-228 by the HEI committee that the effect
                     estimates were quantitatively smaller in the original studies  but that the
                     overall  effect of PMio on mortality remained.

p. 8-274,1. 21        Insert "than" after "less" so it should read "scale of less than 100  miles".

p. 8-281              The authors of the epidemiology chapter should give  consideration to
                     shortening the titles of the various tables and figures.  For example, the
                     title to Table 8-40 is typical of excess wording.

p. 8-305,1. 5          The statement here implies that the model used by Pope included
                     interaction terms for smoking and pollution versus not smoking and
                     pollution such that the effects of smoking as a main factor could be
                     eliminated and yet the potential interaction with pollution be detected.
                     Were these terms indeed included in the followup analysis of the  ACS

p. 8-306,1. 25        The discussion here of individual vs. population threshold is somewhat
                     masking the issue relative to the  Clean Air Act. The Clean Air Act does
                     not require that every single member of the population be protected but
                     rather sensitive subgroups of the population. Thus, the concept of
                     threshold is still relevant even though it is a difficult task to identify

                    whether a threshold exists.

p. 8-307,1. 26       As this reviewer has previously noted, the grid search involving 5 |j,m/m3
                    increments is too coarse a grid in relationship to the fine particle standard
                    of 15 |j,m/m3. The increments represented 33% of the standard and a
                    realistic search should have had much finer grid increments on the order
                    of 1-2 |j,m/m3.  The grid size is also a function of the roughness of the
                    response surface being examined. Either the authors of this chapter
                    should acknowledge these concerns or else provide information that
                    refutes it.

p. 8-308,1. 14       How was the hypothesis of linearity more formally examined using AIC
                    values across models? Was a statistical test applied to distinguish if two
                    models were significantly different? The question of changes in the AIC
                    for really being interpretive is analogous to the results of an analysis of
                    variance where Bernhard showed that an F value had to be four times the
                    computed value before one could really discuss that statistical
                    significance was at a lower level of probability than what had been stated
                    a priori. In short,  a couple of sentences should be added here explaining
                    how the AIC was used from  a statistical viewpoint.
Chapter 9

General Comments

Given that CAS AC provided consultation to the Agency and basically endorsed the outline of the
integrated synthesis chapter, it would be hard to fault the Agency for the structure of the chapter.
That being said, there is considerable imbalance between the sections that address new
information on the broader set of questions that are the section titles for 9.2.1, 9.2.2, 9.2.3, etc.
Not all of this imbalance is due to the amount of additional information that has been developed
since the 1996 review, but may be due to different subsection authorship and writing style.
Common throughout Chapter 9 is an excessive use of adjectives such as "considerable",
"strong", "very", "extensive", etc.  such that the reader gets the impression that a harder sell is
being made than what may be warranted by the data contained in the first 8 chapters of the
Criteria Document.
   The integrated synthesis contains a number of technical errors in bringing forward material
from the other chapters, some of which are noted in my Specific Comments. The authors of
Chapters 8 and 9 have continued to not address the concern I expressed relative to the
Brunekreef (1997) study purporting a reduction in life time expectancy due to PM exposure
increase of 10 |ig/m3  and the calculation of a reduction of about 1.3 years for the entire
population's life expectancy at age 25 using 1969-1971 life table data. This calculation needs to
be redone using a life table not 30 years out of date. Moreover, the vastness of this decrease is
clearly not detected in the few long term mortality epidemiology studies that have been
conducted. This is but one example of where CAS AC or public comments on technical matters
have largely been ignored in further drafts of CD chapters.

   I still get a sense of "selectivity" in Chapter 9 compared to an analysis of "here is what we
know now about PM health effects in 2004 compared to what we knew in 1996 and here are
areas of PM effect analyses where we honestly can't say what is going on".
   I would have thought that the integrated synthesis would focus more within the sections on
whether the information sheds new light as to the level of the standard, the averaging time, the
value, etc. As an example, the section on the discussion of lag times ( would benefit
from a consideration  of a moving average for a standard since the results for the various lag
times can vary considerably, although the  authors make the point that a greater span of time
should provide a more stable estimate of the overall magnitude of potential PM effects and that
shorter lags may underestimate the PM effect.
   Some consistency in the structure of each of the subsections should be imparted by a technical
editor. Currently, some sections include references while others do not, and the bottom  line of
the conclusions on the incremental information provided by studies since the  1996 review is not
always clear. In addition, the bottom line is evident in some of the sections but not in others.

Specific Comments

p. 9-1,1. 23          Suggest adding at the end of the sentence "or to establish a new standard
p. 9-13,1. 10         Add to this list the anatomy of the respiratory tract of the individual.
p. 9-19,1. 7          Insert PM exposure after short-term.

p. 9-13,1. 14         It should be PMio-2.s in this sentence.
p. 9-24,1. 10         Most of the text on this page is redundant with the previous page. The
                    percent increased risk discussion should be combined with the discussion
                    on the previous page about the degree of association be it positive or
                    negative and the extent of statistical significance of various effects with
                    the three PM indicator variables.
p. 9-35,1. 7          For the analysis discussion in Section and brought forward to
                    Chapter 9, the creation of six l-in-6 day data sets and the resulting
                    inconsistency of PM mortality associations appears to be used  as an
                    argument for restricting analyses to those that only have data for every
                    day and do not attempt to  do any extrapolation of values based upon
                    atmospheric models or whatever. This is a  potentially very important
                    finding because EPA should assess how many studies used 1 and 6-day
                    values compared to those that have every day analyses and determine the
                    overall magnitude of effect when the data sets are more complete. The
                    bottom line is that the l-in-6 day sampling pattern is adding uncertainty
                    to the PM effects and represents the kind of uncertainty addressed by
                    Koop and Tole (J. Environ. Econ. Management 47, 30-54, 2004). In
                    addition, the sites for which the data are available every day could be
                    used informatively to establish the lag time or basically the averaging
                    time which is one of the important factors that has to be considered in the
                    development of any PM standard.

p. 9-46,1. 8           This reviewer is under the impression that the broader toxicologic
                     community views the monocrotoline treated rat model as not very useful
                     for studying the effects of PM due to the severity of the state of the
                     animals before being challenged with PM exposures as well as the fact
                     that the model acutely tries to develop in animals what is a chronic
                     process in humans.

p. 9-50,1. 27         Why restrict the discussion here to TB cell exposures? A more logical
                     statement is lower respiratory tract cellular exposures as instillation
                     clearly allows for particles to be delivered throughout the lower
                     respiratory tract, albeit to a greater extent of variability than by

p. 9-53               The same cautions concerning interpretations of the results shown in
                     Table  9-4 are warranted in the integrated synthesis chapter as this
                     reviewer has conveyed for the original presentation of the information in
                     Chapter 7.

p. 9-54,1. 5           This paragraph needs to be put in perspective relative to the actual
                     number of particles that a given cell would be exposed to because the in
                     vitro doses are misleading in that they correspond to tens to hundreds of
                     particles per cell and thus are still in the extremely high dose mode.
p. 9-54,1. 30         The study by Ulrich et al. used such high instillation doses that the
                     results of showing an increase in plasma fibrinogen by Ottawa UAP are
                     an absolutely high dose phenomenon and to this reviewer are not worthy
                     of being included in the integrated synthesis chapter.

p. 9-55,1. 4           The same comment as above can be attributed to the ROFA studies
                     involving inhalation of 10-15 mg/m3.

p. 9-55,1. 14         The characterization of the Ottawa UAP study in rabbits is incorrect in
                     saying that the material  caused in increases in atherosclerotic lesions. As
                     described in Chapter 7 on page 156, the study showed a progression of
                     atherosclerotic lesions and not an increased number of them.

p. 9-57,1. 2           Delete the first "in" that appears before BAL.

p. 9-57,1. 16         Again, the integrated synthesis does not put the in vitro studies into
                     perspective relative to number of particles that each cell would see.
                     Calculations show that these numbers would range in the tens to many
                     hundreds of particles per cell and clearly still represent a high dose

p. 9-58,1. 1           Delete "strong". What is the criteria by which one would use the
                     adjective  strong in this instance?

p. 9-60,1. 28         We hardly need the new studies to know that some types of particles are
                     more toxic than others. This understanding has been present for at least
                     30 years.

p. 9-61,1. 20         Change "dosimetric" to "dose metric".

p. 9-76,1. 23         In this paragraph, the authors should make clear that they are talking
                    about aerodynamic diameters at least with respect to the 1-2 |j,m
                    particles; i.e., the density of the particle needs to be taken into account
                    via the aerodynamic diameter.

p. 9-77,1. 9          Ozone is a bad example to describe toxic material in the particle bound
                    water aspect of a droplet given the extreme insolubility of ozone in

p. 9-77,1. 17         This paragraph somewhat overstates that trends in deposition and
                    patterns of deposition seen in diseased or compromised lungs in that the
                    experimental studies have not used a wide range of particle sizes, and the
                    authors should indicate the size range for which these statements are
                    considered to be appropriate.

p. 9-78,1. 19         Again, remove ozone from this list of agents.

p. 9-78              Delete "the" at the end of the line. In addition, the statement that they
                    likely contribute to some types of ambient PM health effects exposure is
                    probably an overstatement given the description of effects and the
                    concentrations of these bioaerosols needed to produce any changes (see
                    Appendix 7B). Such is at least the case for cellulose given that ambient
                    levels are typically only a few ng/m3 whereas effect studies have used
                    mg/m3 quantities and seen very little changes in health outcome in either
                    animals or humans.

p. 9-83,1. 18         Once again, the authors of the integrated summary have not brought
                    forward the  results from Appendix 7B which show that the instillation
                    doses are not equivalent to higher level exposure concentrations that
                    could be experienced in the time frame for which the original authors
                    contend. Rather, the doses correspond to 6-9 weeks of elevated exposure
                    and as a result are not reflective of the community-wide exposures that
                    individuals would be expected to receive.

p. 9-84,1. 28         Exposures over several weeks to PMi0 is an incorrect statement
                    compared to requiring 6 to 9 weeks of exposure as determined on p. 7A-
                    45 of the extrapolation appendix.

p. 9-85,1. 12         Perhaps a more accurate representation of the extent to which exposures
                    of humans at higher ambient levels compare with doses in exposures
                    used in toxicological studies would be to indicate on this line instead of
                    "are often not necessarily" changed to "are sometimes not necessarily".

p. 9-85,1.31         Again, delete ozone from the list.

p. 9-90,1.14         The reference to the figure whereby dose is stated on a per kg
                    bodyweight basis being much higher can't really be determined from the
                    figure  since  it merely relates to the inhalation rate as a function of age. If
                    deposition rates are different in children than adults, this would be a

                    major driver in dose comparisons between children and adults. The
                    statement is made later in the paragraph that the discrepancy is even
                    greater when viewed on a per lung surface area basis. This is indeed
                    correct, but why not show this kind of data in a figure rather than the
                    inhalation rates from the Layton (1993) publication?

p. 9-99,1.  19         It seems inappropriate to this reviewer to characterize childhood asthma
                    as a less serious condition compared to other health outcomes later in

p. 9-102,1. 8         The entire section on welfare effects is interesting reading but really does
                    not come down to any bottom line relative to potential changes of any
                    existing welfare standards or addition of new ones. Aside from the
                    obvious positive impact on visibility by reduction of fine particles, the
                    rest of the synthesis really does not provide a punch line as to the
                    benefits that could most likely be achieved with fine mode PM
                    reductions. The variability between air sheds and ecosystems and the
                    amount of Nr doesn't seem to support the conclusion that one might
                    expect — namely that reduction in Nr would result in an improvement in
                    the ecosystem.

                                Mr. Richard L. Poirot

Chapter 9 Comments, R Poirot

Overall, this is a very good first shot at an inherently difficult target.  I don't think major
substantive changes should be required to finish this chapter and finalize the CD.

9.2.1:  The first question considers the continuing distinction between fine & coarse particles,
provides strong support for such a distinction, and in several cases (for example, new
information on characteristics of particles in the "intermodal" - 1-2.5 um - size range), further
strengthens the selection of 2.5 um as the division point.  By contrast, discussion or additional
justification of the 10 um upper cut-point for "coarse particles" seems notably absent.  Its been
several decades since the 10  um cut-point was carefully considered, and revisiting this decision
seems especially important if 2.5 is now firmly established as the lower cut-point - thereby
excluding the fine tail of the  coarse mode from the newly defined "PMio-2.5" category, which will
also exclude the coarse tail (& in some cases most of the  whole dog) of the coarse mode particles
from consideration. There are several indications throughout the chapter that crustal particles
(which typically compose the majority of the coarse mass) appear to be uniquely not associated
with adverse health effects (and I think generally the mass - without regard to the  composition -
has virtually no relevance to effects  on environment).  Yet we appear poised to develop and
implement costly new ambient measurement programs specifically focused on quantifying the
mass of particles in the PMio-2.5 size range.  Is this wise?

p. 9-5, lines 19-20: Could delete "re" from "resuspended", delete "in dry dusty areas", and delete
"in cities". The fine tail of coarse-mode suspended soil particles exists in humid areas too, and
since coarse mode particles are less  abundantly emitted in humid areas, their presence there often
originates elsewhere, such that a proportionately larger fraction of larger coarse-mode particles
has been removed during transport.  Fine (but coarse-mode) sea salt particles and their
derivatives exist everywhere near oceans (not just cities)  and are often observed many hundreds
of miles inland.

p. 9.7, line 1: ditto. When using terms like "fraction of inhaled particles",  "most accumulation mode
particles" and "fractional deposition", I assume you intend "fraction by number" and not by
mass? If this is the case - or not - it would be helpful to  state clearly, since few other discussion
sections relate specifically to particle count. Possibly it could be informative to comment on the
associated mass fractions (in Figure 9-4, for example).

pp. 9-21&22, Figures 9-5 and 9-6: Are excellent data visualization & communication
applications!  I don't suppose there would be any easy way to distinguish between those studies
that included 1, 2, 3, 4,... different PM-size ranges and/or effects metrics?  Also, here and in
following pages, is it absolutely necessary to use different units (i.e. per 50 ug/m3 for PMi0 vs.
per 25 ug/m3 for PM2.5?  If it is, could you provide a brief explanation why?

pp. 9-26, lines 28-31 & top of p. 9-27: If available, it would be informative to cite the coarse
mass associated with the crustal and "metals" factors in Phoenix.  Surely the "metals" (which
may well be the causal factor here) represent a small fraction of the coarse mass (yet we're
about to embark on a new (and fairly useless) coarse mass measurement program).  This is not necessarily the place to address this unasked/unanswered question, but a
detailed discussion of (short-term) lag periods clearly implies significant response from short-
term exposures. The relevant question that might be posed somewhere is something like: Are
health effects of PM most strongly  associated with short-term exposures, long-term exposures, or
both?  This lengthy, multi-caveat discussion feels like a re-summary of issues better
addressed in preceding chapters (6  & 7). Too many "howevers"; too few simple declarative
sentences.  Overall, a very informative discussion, clearly stated!

p.  9-62, line 20: By "fine fraction" do you mean "accumulation mode", or everything less than
2.5? If its everything less than 2.5, it can't have less surface area than ultrafines - which it
includes. If its "accumulation mode", it does not necessarily have a "much larger particle
number" than ultrafines.

p.  9-63: This discussion of acid aerosols - which for the most part is nothing new - might benefit
from a pointer to, or some additional discussion of the section discussion of the
influence of aerosol water - which  is relatively new and informative. All things (RH) being
equal, a more acidic aerosol will draw more water to aerosol phase than its neutralized
counterpart and also further increase the solubility of some gasses, organics & metals (&
decrease solubility of others).  Discussions of aerosol water and bio-aerosols  seem new, helpful, and directly
responsive to previous review comments. I've previously commented extensively on visibility effects sections and won't repeat -
except to emphasize that differences between PM-2.5 mass and visibility effects are
predominantly due to aerosol water - present in ambient  aerosol and intentionally removed by
the "artificially dried" regulatory, instrumental definition of "fine mass".

p.  9-103, line 7: Add "artificially dried" before fine mass to make it clear you are not referring
to  the mass of fine particles in the ambient air - which has already been affected by the
combination of relative humidity and the hygroscopic characteristics of the aerosol.

p.  9-104, lines 1-10: Since  the last review, neither the algorithms & parameters used to calculate
light extinction nor the measurement methods used to quantify visibility effects from aerosol
species have changed for the (IMPROVE) program which is now the basis for calculating haze
effects in class 1 federal areas. The cited Malm (2000) analysis showing improved performance

by considering sulfate ammoniation (at acidic GRSM) relates again primarily to a better estimate
of ambient aerosol water content.

p. 9-104, line 15: add "rural" before "West"

p. 9-106, linesl-4:  could also add that these local standards have (at least in Denver) resulted in
PM2.5 emissions & concentration reductions.  Could also add that "similar threshold
determinations, convergent on a minimal visual range of 40 to 50 km have also been identified
in visibility standards in Lake Tahoe, the Fraser Valley and State of VT." Two locations picking
similar "adverse thresholds" is a coincidence; 5 is a convergence.

9.3.2: A substantial improvement over previous "ecological effects" discussions! But the
question remains ambiguously avoided regarding whether eco-effects of PM include: PM
deposited as (dry) PM only; PM (formerly present as PM in the ambient air) deposited by dry,
wet & occult process; or PM & PM precursors deposited by various mechanisms. Much of the
discussion relates to the latter (cumulative effects of PM and its precursors) definition, which
seems important to include here -  for later consideration of costs/benefits of attaining standards.

9.3.3: Not much content here, but pleased to see that EPA still (or again) considers climate
change as potential environmental effect of anything...

9.3.4: Materials damage section is especially lame. Effects from PM & precursors are
significant, but less considered here than 1996 CD  or NAPAP SOST. Might be better to state
clearly that "federal research funds have not been available to investigate the effects of PM and
its precursors since the mid-1980s."
Supplemental Comments on PM CD Chapters 8 and 9, R. Poirot 7/26/04

Several sections of chapters 8 & 9 (for example & summarize health effects
associations with different chemical components and/or source categories on PM in various size
fractions. These discussions are clear, detailed, helpful and informative - and (I think) the results
could conceptually be presented, integrated, summarized, etc. in two general ways:

    1.  to indicate that many or most all of the major PM mass-contributing species or source
       contributions have been individually shown to be injurious (this adds considerably to the
       use of PM2.5 or PM10-2.5 mass as a regulatory metrics, regardless of different PM
       mixtures in different regions).

    2.  to indicate that some species or source categories appear to be more harmful or less
       harmful than others (potentially this might lead to species-specific standards or source-
       specific priorities in the implementation phase).

Based on discussions at the CASAC PM CD review, subsequent discussions at the PM coarse
monitoring methods review and on a re-reading of relevant sections  of chapters 8 and 9,1

encourage EPA to more heavily emphasize the former (#1) use of this information and de-
emphasize the latter (#2). General reasons include:

    a.  Adverse health effects are associated with many different species and/or source-specific
       contributions, although these associations are not always consistent among studies.
       Taken in the aggregate, they clearly show adverse effects from many species, but
       individually no one study is definitive.

    b.  The species and/or source-specific health associations are not sufficiently  strong or
       consistent in their findings to support species-specific standards or to prioritize (or
       exculpate) species or sources for future controls at the present time - and to do so would
       require choosing among or rating studies which show contrary effects (a much more
       difficult argument to support than #1).

    c.  Epi. studies associating specific source categories with effects (or non-effects) are limited
       in number, and have generally have relied on "factor analysis" approaches (such as PCA
       with Varimax or Procrustean rotation) which are not currently considered state-of-the-
       science (poorly constrained and potentially yielding many different "equally correct"
       answers) and require subjective interpretation of the resulting sources. These results are
       then often further interpreted and commented on in the CD in a highly speculative

Specifically, I think the  chapter 9 integrated synthesis should de-emphasize or present counter
examples in sections where specific source categories are  identified as uniquely benign.  This
seems most evident for the contributions of "crustal" emissions to PM2.5, PM10 and/or PM10-
2.5.1 think this is especially not helpful in considering any coarse particle standards, since
crustal material (and its  associated anthropogenic chemical or biological contaminants) is
typically a large fraction of coarse mass at most times & places. For example following are
several examples where I think the potential effects of "crustal particles" are unnecessarily (&
speculatively) de-emphasized:

On p. 9-44, lines  18-19: "Also of much importance,  all of the above studies that investigated
multiple  source categories found a soil or crustal source that was negatively associated with
mortality". Here,  its not entirely clear why this is "of much importance" (compared to what?), or
what "all of the above studies" refers to (the preceding paragraph, page, section?). The consistent
finding of a negative association (and implication we would live longer if it were  dustier) is a
consistent indication (to me) of a poorly formulated  model(s). It is also inconsistent with the
many studies (mostly cited in the CD) which do show effects associated with coarse particle
mass, and with the rather extensive bodies of literature on adverse effects from both the
inorganic components of crustal material (silicosis, pneumoconiosis, etc.), as well as with the
extensive and growing literature on diseases associated with soil-borne fungi or bacteria
(Coccidioidomycosis, etc.). I've listed some references grouped in these 3 general areas  at the
end of these comments.

Several features of the (rather outdated) receptor model approach taken by the studies which I
assume are referred to in "all of the above studies" are important. First, all multi-elemental

measurement techniques, and especially the most common XRF, coincidentally quantify a large
number of elements which are of predominantly crustal origin (Si, Al, Fe, Ca, Ti, etc. - much
more so than for any other source category). For this reason, a "crustal" or "soil" factor is nearly
always identified in virtually all receptor model applications. The (rotated eigenvector) factor
analysis approach which I think was used in all of the above studies seeks first to account for the
collective variance of all the species used as input, and so typically (prior to rotation) the first
component, explaining a maximum of the total variance tends to be "crustal" (even though these
elements together typically account for only a small fraction of the fine mass).  Subsequent
rotational schemes (Varimax, Procrustes, etc.) then redistribute the variance in ways that require
highly subjective decisions by the modelers. These models also require (can only find) sources of
fixed, unique chemical composition and variable, unique contribution. Soil  itself has a highly
variable composition but tends to be more alkaline in the West than in the East, very alkaline in
areas with calcarious bedrock, and different yet again in the  Sahara Dust and Asian Dust which
often result in the highest soil contributions in the Eastern US and West coasts respectively.
These more distant dust events also tend to have much smaller particle size distributions than
"local dust" emissions, as the larger particles are more readily removed during transport. Crustal
material can become heavily contaminated with anthropogenic S, N,  OC, EC, salt and metals -
both as it is deposited & resuspended from roadways or as it undergoes chemical reactions
during transport. Conversely, many other sources also contain "crustal impurities" (coal fly ash
for example), and so when one obtains a "pure crustal source" from a factor analysis its not
entirely clear what that source actually represents. If the rotation is oblique, the sources are
required to be uncorrelated, and it's therefore highly probable that the "crustal" source will (to
the extent local sources contribute) be a good indicator of high wind  speeds, since this will lead
uniquely to high emissions & concentrations of dust which will be uncorrelated with all other
(gaseous &) particulate pollutants. While high dust concentrations that also build up under
stagnation conditions (from road dust emissions) or dust from more distant  origins will tend to
get mixed into other modeled sources. Quite possibly the consistent finding of negative health
associations with dust just reflects windy days when folks stay indoors and  the air is otherwise at
its cleanest. For example:

On p. 9-27, lines 1&2, we learn that "new studies have shown no increases  in mortality on days
with high concentrations of wind-blown dust (crustal particles), using PM 10 concentrations and
data on wind speed as indicators of dust storm days." Which new studies? I think the (not
unreasonable) use of wind speed as a dust surrogate is telling, as dust emissions (especially the
maximum concentrations) are uniquely associated with high wind speeds - which in turn will
tend to minimize concentrations of all other (fine) particle and gaseous components - assuring
minimal chemical reactions between crustal particles and other species. High concentrations of
crustal particles and chemically associated contaminants (on the surface of coarse particles) from
MV, SO2 or smelting activities would also reach high concentrations (as would many other
gaseous and PM pollutants) on local stagnation days with low mixing heights - but would not be
considered with this "wind speed" surrogate (nor would dust of distant origin). Potentially
outdoor activities are curtailed on very windy, "local" dusty days, windows are closed, inhalation
efficiency of coarse particles likely decreases with wind speed, and the spatial representativeness
of "central site monitors" diminishes. Conversely, the lengthy Section discussion of
"Adjustments for Meteorological Variables" includes factors like temperature and humidity that

might tend to exaggerate assumed PM effects, but makes no mention of wind speed - which
might tend to diminish such effects.

On p. 9-27, lines 3-6, it is postulated that cardiovascular mortality in Phoenix may be due to the
metal rather than crustal content of coarse particles. Yet on p. 8-63, lines 22-28 it's indicated that
"... (Smith et al., 2000) indicate that coarse particle-mortality associations are stronger in spring
and summer, when the anthropogenic metal (Fe, Cu, Zn, and Pb) contribution to PM10-2.5 is
lowest, as determined by factor analysis." In this case, the seasonal association of effects when
crustal, not metal, coarse particles are greatest is attributed speculatively to "biogenic processes
(e.g., wind-blown pollen fragments, fungal materials, endotoxins, and glucans) of the particles
during spring and summer". It is also specifically emphasized that the authors "observed that the
implication that crustal, rather than anthropogenic elements, for the observed relationship with
mortality was counterintuitive." Thus a finding that does not fit the theory is discredited.

Emphasizing the potential importance  of coarse biological content is reasonable, but on p. 8-326,
lines 8-17, its indicated that "Reasons for differences among findings on coarse-particle health
effects reported for different cities are  still poorly understood, but several of the locations where
significant PM10-2.5 effects have been observed (e.g., Phoenix, Mexico City, Santiago) tend to
be in drier climates and may have contributions to observed effects due to higher levels of
organic particles from biogenic processes (e.g., endotoxins, fungi, etc.) during warm months."

Here, I can understand how dry climate can and does lead to increased emissions and
concentrations of coarse crustal material (and any biological material it contains), but I'm not
sure why or if its logical to expect arid climates (and associated sparse vegetation) to have
uniquely higher pollen, endotoxin or fungi emissions & concentrations than humid areas - where
wind-blown dust emissions would tend to be suppressed by precipitation, and where pollens,
pollen fragments and fungi might be relatively more abundant.

I think a more logical  explanation could be effects from soil-associated fungi, which for the most
part become airborne only as the soil becomes airborne during "natural" dust storms and/or as
modified by human agricultural activities (tilling harvesting, grazing, etc.) and on & off-road

For example, the geographically-focused incidence  of "Valley Fever" specifically caused by
caused by the fungus Coccidioides sp., which grows in soils in areas of low rainfall,  high
summer temperatures, moderate winter temperatures, and which is emitted in direct association
with the soil that supports it, would seem like a more logical causal or contributing factor than
some non-soil-related biogenic contribution from pollen or more benign fungi in general. See
also the references on other soil-related fungal or bacteriological effects on human & animal
health, crops, aquatic ecosystems, etc.  - for example Garrison et al. (2003).

On p. 8-326, lines 17-21 it is indicated that "in  some U.S. cities (especially in the NW  and the
SW) where PM10-2.5 tends to be a large fraction of PM10, measurements, coarse thoracic
particles from woodburning are often an important source during at least some seasons. In such
situations, the relationship between hospital admissions and PM10 may be an indicator of
response to coarse thoracic particles from wood burning."

        Spuikil DNriluiliun of Valley  Inner
         Source' hup  \\ v, >;.,.'.. alkn te\ erc^m
                        •  •   ~V  ~\        "<
                        '  !    f '  i,   f"  V-!>
                        , —A  ! i  / ,/ •""> I*
However, since wood smoke concentrations are VERY predominately < 2.5 um, it seems
illogical that wood smoke should be the likely causal factor for coarse particle effects in areas
that have high coarse:fine ratios. I also question whether the NW has a high coarse:fine ratio and
why the (dusty, crusty) SW would tend to have a uniquely high coarse wood smoke contribution
(compared to all northern areas where space heating demands and fuel wood supplies are
greater). This also seems inconsistent with the "counterintuitive" Phoenix results indicating
highest coarse PM effects in the spring & summer. I'm getting picky here, but again it looks like
trying too hard to show "it must be anything but crustal emissions"...

On Coarse PM health effects associations in general:

Becker S, Soukup J. Coarse(PM(2.5-10)), fme(PM(2.5)), and ultrafme air pollution particles
induce/increase immune costimulatory receptors on human blood-derived monocytes but not on
alveolar macrophages. J Toxicol Environ Health A. 2003 May 9;66(9):847-59.

Becker S, Soukup JM, Sioutas C, Cassee FR. Response of human alveolar macrophages to
ultrafme, fine, and coarse urban air pollution particles. Exp Lung Res. 2003 Jan-Feb;29(l):29-44.

Becker S, Fenton MJ, Soukup JM. Involvement of microbial components and toll-like receptors
2 and 4 in cytokine responses to air pollution particles. Am J Respir Cell Mol Biol. 2002

Soukup JM, Becker S. Human alveolar macrophage responses to air pollution particulates are
associated with insoluble components of coarse material, including paniculate endotoxin.
Toxicol Appl Pharmacol.  2001 Feb  15;171:20-6.

Ostro BD, Broadwin R, Lipsett MJ. Coarse and fine particles and daily mortality in the
Coachella Valley, California: a follow-up study. J Expo Anal Environ Epidemiol. 2000

Ostro BD, Hurley S, Lipsett MJ. Air pollution and daily mortality in the Coachella Valley,
California: a study of PM10 dominated by coarse particles. Environ Res. 1999 Oct;81(3):231-8

Sheppard L, Levy D, Norris G, Larson TV, Koenig JQ. Effects of ambient air pollution on
nonelderly asthma hospital admissions in Seattle, Washington, 1987-1994. Epidemiology. 1999

Cifuentes LA, Vega J, Kopfer K, Lave LB. Effect of the fine fraction of paniculate matter versus
the coarse mass and other pollutants on daily mortality in Santiago, Chile. J Air Waste Manag
Assoc. 2000 Aug;50(8): 1287-98. Burnett et al. 2000

Lin M, Chen Y, Burnett RT, Villeneuve PJ, Krewski D. The influence of ambient coarse
particulate matter on asthma hospitalization in children: case-crossover and time-series analyses.
Environ Health Perspect. 2002 Jun;l 10(6):575-81. Sheppard L et al. 1999

Pozzi R, De Berardis B, Paoletti L, Guastadisegni C. Inflammatory mediators induced by coarse
(PM2.5-10) and fine (PM2.5) urban air particles in RAW 264.7 cells. Toxicology. 2003 Feb

Kleinman MT, Sioutas C, Chang MC, Boere  AJ, Cassee FR. Ambient fine and coarse particle
suppression of alveolar macrophage functions. Toxicol Lett. 2003 Feb 3;137(3):151-8.

Monn C, Becker S. Cytotoxicity and induction of proinflammatory cytokines from human
monocytes exposed to fine (PM2.5) and coarse particles (PM10-2.5) in outdoor and indoor  air.
Toxicol Appl Pharmacol 1999;155: 245-52.

Schins RP, Lightbody JH, Borm PJ, et al. Inflammatory effects of coarse and fine particulate
matter in relation to chemical and biological constituents. Toxicol Appl Pharmacol. 2004; 195(1):

Shi T, Knaapen AM, Begerow J, et al. Temporal variation of hydroxyl radical generation and 8-
hydroxy-2'-deoxyguanosine formation by coarse and fine particulate matter. Occup Environ
Med.2003;60: 315-21.

Greenwell LL, Moreno T, Jones TP, et al. Particle-induced oxidative damage is ameliorated by
pulmonary antioxidants. Free Rad Biol Med.  2002;32(9):  898-905.

Mar  TF, Norris GA, Koenig JQ, et al. Associations between air pollution and mortality in
Phoenix, 1995-1997. Environ Health Perspect 2000; 108: 347-53.

Castillejos M, Borja-Aburto V, Dockery D, et al. Airborne coarse particles and mortality.
Inhal Toxicol 2000; 12 (supplement l):61-72.

On dust -associated inorganic components & effects:

Gift, J.S., & Faust, R.A.: "Noncancer Inhalation Toxicology of Crystalline Silica: Exposure-
Response Assessment," J. Expos. Analysis Environ. Epidemiol. 7(3^:345-358 (1997).

Wright, G.W.: "The Pulmonary Effects of Inhaled Inorganic Dust" (Chapter 7). In Patty's
Industrial Hygiene and Toxicology, 3rd Rev. Ed., Vol. 1 — General Principles. New York: John
Wiley & Sons, 1978. pp. 175-176.

M. Schenker (2000) Exposures and Health Effects from Inorganic Agricultural Dusts,
Environmental Health Perspectives Supplements, Volume 108, Number S4 August 2000.

International Agency for Research on Cancer: I ARC Monographs on the Evaluation of
Carcinogenic Risks to Humans, Vol. 68 — Silica, Some Silicates, Coal Dust and Paraaramid
Fibrils.  Geneva, Switzerland: World Health Organization, IARC Press, 1997. pp. 41-85.

U.S. Environmental Protection Agency/Office of Research and Development: Ambient Levels
and Noncancer Health Effects of Inhaled Crystalline and Amorphous Silica: Health Issue
Assessment (EPA/600/R-95/115). November 1996. pp. 7-1-7-6.

On dust -associated biological components & effects:

Buchwaldt, L., ... and C.C. Bernier. 1996. Windborne dispersal of Colletotrichum truncatum and
survival in infested lentil debris. Ecology and Epidemiology 86:1193. Gloster, J., R.F. Sellers,
and A.I. Donaldson. 1982. Long distance transport of foot-andmouth disease virus over the sea.
The Veterinary Record 110(Jan. 16):47. Griffin, D.W., V.H. Garrison, etal. 2001. African desert
dust in the Caribbean atmosphere: Microbiology and public health. Aerobiologia  17(June

Griffin, D.W., C.A. Kellogg, and E.A. Shinn. 2001. Dust in the wind: Long range transport of
dust in the atmosphere and its implications for global public and ecosystem health. Global
Change and Human Health (September). O'Hara, S.L.,  etal. 2000. Exposure to airborne dust
contaminated with pesticide in the Aral Sea region. Lancet 355(Feb. 19):627.

Prospero, J.M. 1999. Long-term measurements of the transport of African mineral dust to the
southeastern United States: Implications for regional air quality. Journal of Geophysical
Research 104(July 20): 15917.

Snyder LS, Galgiani JN. Coccidioidomycosis: The Initial Pulmonary Infection and Beyond.
Seminars in Respiratory and Critical Care Medicine 1997; 18:235-247.

Galgiani JN. Coccidioidomycosis. In: Remington JS and Swartz MN ed. Current Clinical Topics
in Infectious Diseases. 1997, pp. 188-204.

Morbidity & Mortality Weekly Report, "Coccidioidomycosis- Arizona, 1990-1995. Dec. 13,
1996/vol. 45:1069-1073.

Einstein HE, Johnson RH. Coccidioidomycosis: new aspects of epidemiology and therapy.
Clinical Infectious Diseases 1993;16:349-356.

Galgiani JN. Coccidioidomycosis. Western Journal of Medicine 1993;159:153-171.

Galgiani JN. Coccidioidomycosis. In: Rakel RE, ed. Conn's Current Therapy. Philadelphia: WB
Saunders Co., 1996, pp. 188-190.

Galgiani JN, Ampel NM. Coccidioidomycosis in human immunodeficiency virus-infected
patients. Journal of Infectious Diseases 1990;162:1165-69.

Hall KA, Copeland JG, Zukoski CF, et al. Markers of Coccidioidomycosis before cardiac or renal
transplantation and the risk of recurrent infection. Transplantation 1993;55:1422-24.

Kwon-Chung KJ, Bennett JE. Medical Mycology. Philadelphia, Lea and Febiger,  1992, pp. 356-

Pappagianis D. Marked increase in cases of Coccidioidomycosis in California: 1991, 1992, and
1993. Clinical Infectious Diseases 1994;19(Suppl 1):S14-18

Pappagianis D, Zimmer BL. Serology of Coccidioidomycosis. Clinical Microbiology Reviews

Schneider E, et al: A Coccidioidomycosis Outbreak Following the Northridge, Calif Earthquake.
JAMA 1997;277:904-908.

Pappagianis D: Coccidioidomycosis, In DiSalvo A (Ed) Occupational Mycoses. Philadelphia,
PA, Lea & Febiger, 1983, pp  13-28.

V. H. Garrison, E. A.  Shin, W. T. Foreman, D. W. Griffin, C. W. Holmes, C. A. Kellogg,  M.  S.
Majewski, L. L. Richardson, K. B. Ritchie & G. W. Smith (2003) African and Asian Dust: From
Desert Soils to Coral Reefs, BioScience . May 2003 / Vol. 53 No. 5, pp. 469-480

                                     Dr. Frank Speizer
Review of Chapter 7
Submitted by F.E. Speizer
       Summary Comment:

Section 7.201
       I found the summary (section 7.9) difficult to read. This is mainly because, although it professes
to be interpretive on toxicologic findings, from pages 7.201 to the end of section 7.92.2, it is mostly a
catalog of the findings in animals both positive and negative with little interpretation of the relevance to
the human mechanisms of interest. The next section on seems to have been written by a different
voice in that the toxicologic findings are presented in a much more interpretive mode to explain, where
possible,  the human finding reported by the epidemiology.  It just may be that the data are more clear for
the last half of the summary than the first half, but I doubt it; particularly since most of the agents
mentioned have been studies  far less than those mentioned in the first half. Examples of sentences that
are not helpful:

                      p. 7.206, lines 25-27
                      p. 7.206, lines 28-29
                      p. 7.207, lines 23-26
                      p. 7.208; lines 24-25 (what is good of studies negative in animals but positive in

This is also apparent in some of the tables. It seems that the authors are losing their focus. The purpose
of the toxicology is only partly to understand the mechanism in animals, but surely to understand it when
it is relevant to humans.

Side Issue:
       I really doubt Gohio and Devlin (2001) statement that average ventilation of active person is 15
/min for 24 hours. If I give an active person 2 hours at 30 /min; 8 hours of sleep at 5 /min, and 14 hours
of wake time at 10 /min (generous) that only comes to 10 /min for the 24 hours..
Review of Chapter 8
Submitted by F.E. Speizer

General Comment:

This is a much more readable draft that the previous one. The format of reviewing the 1996 document
followed by the new material seems to work well. I think the chapter is ready to be signed off on with
very minor modifications the staff should be able to handle.  Most of the comments below are for editing
and clarity. I found section 8.4 a particularly useful summary that for the most part reads like it belongs
in chapter 9.

Specific Comments:

Page 8.11; Line 29
       Change "eliminate to "control for"

Section 8.1 to page 8.17
       Although this discussion is accurate in what it lays out and tries to do, it is still not clear to me
that it belongs in a criteria document. It would be much more reasonable for a preamble to the staff paper
in that it provides a framework for interpretation rather than a criteria for inclusion. Having been done,
leave it in, but consider not including it in future C.D.'s.

Page 8.61; Summary paragraph, beginning Line  16
       Perhaps one additional point should be added regarding size fraction, in that there appears to be a
regional component to the apparent course fraction response, in that in western cities the PM10_2.5,
measured or indirectly calculated, appears more likely to be significantly associated with total and CVD
mortality. This could be related to somewhat lower PM2 5 levels as fractions of total in west vs. east.

Page 8.106 at Line 11
       Need to provide a brief summary statement on AHSMOG study.

Page 8.114
       Where comparisons are made among studies, suggest add one on VA sites that match ACS sites.
Still not clear what pollution averages were used in VA and how adjustment for "background" modifies
these averages.

Page 8.116; after Line 4
       Add issue of selectivity of the cohort as being at greater risk of hypertensive disease and, as
indicated by Lipfert, possibly depletion of at risk subjects.

       Remarkable consistency of results seen for cohort studies.  The difference seen in the later Lipfert
studies for PM2 5 is really the only inconsistency  and may relate to competing risk in an aging population;
e.g., as risk goes up in older population, may be harder to see effects of pollutant.  Opposite  to the
increasing relative effect of smoking as risk of dying in ex-smokers/non-smokers goes down, relative risk
of smoking goes up.

       Chronic cohort studies summarized well and give consistent result for PM2 5. The results for
PM] 0-2 5 much more tentative and inconsistent.

       No effects of PMi0-2,5 on CVD morbidity or ER visits.

Page 8.146; Line  18
       Change to be specific to indicate that there was a significant association with  PMi 0-2,5 for
Ischemic heart disease, but not for dysrthythmias or heart failure.

Page 8.151; Line 26-27
       Statement not true.  There appears to be  a clear and consistent finding that PMi0 and PM2 5 are
both associated with acute CVD effects. What is not clear is the effect of PMi0-2,5, which is inconsistent.

Page 8.161; Line  1-6
       This  represents too negative a statement for the results seen. It is true that the results across
studies are different, but what is consistent is that there are changes in electrophysiologic parameters that

are associated with PM.  They may be different in different studies, but that should not negate their
importance and therefore it is inappropriate to say that no conclusion is possible.

Page 8.165; Section beginning with Line 5
       Need to specify consistent PMio andPM25 results. No consistent results for PMi0-2,5.

Page 8.165.
       Add section on electrophysiology.  Results are significant, but because the interpretation are not
yet understood, doesn't mean the effects should be ignored. The effects are just as interesting as are the
blood characteristic effects (see line 13.)

Page 8.189; Line 25
       Add sentence: "However, there were consistent significant decrements in PEFR for those regions
where the mean levels of PM10 exposures were 47  /m3 or higher."

Page 8.193; Line 10
       Add:  Again for those communities with higher average levels of PM10 exposure the results were

Section 8.4
       Is a well-organized interpretive summary of the entire chapter.  It reads quite well and potential
with other summary components might become a major section of Chapter 9.  Having not seen Chapter 9
yet, I am not sure that one will not be redundant for the other, at least as far as the epidemiology is
Chapter 9 Integrative Synthesis
Review submitted by F. Speizer

General Comments:  The chapter is organized in a much better way than previously. However,
the last part of chapter 8 appeared to be a much more readable section that integrated and
summarized much of the material. I would have liked to have seen much of that more directly
incorporated.  Here, we seem to have several writers who have taken different approaches to
summarize the data.  In some cases we get brief summaries without references, in others we get
detail with specific references or specific references to earlier chapter segments.  What this does
is to emphasize some findings and de-emphasize others. A single science editor needs to go
through (and shorten) the whole chapter. The content seems appropriately reviewed and the
overall emphasis seems right.  Therefore staff may be able to move forward to closure, given the
input without further  CASAC review.
Specific Comments:

Page 9.19, line 2:  Typo "known"

Page 9.23, line 14: Should this first line PM2.5 be PM10-2.5?

Page 9.24, line 5:  this should be Figure 9.6.

At the end of this section it seems to me that a comment is warranted that for both mortality and
morbidity the effects of PM10-2.5, although not always significant was generally larger in
magnitude for respiratory effects vs. cardiovascular effects.

 Page 9.28, line 1-2:  More detail in needed here on the VA study.  It is my understanding that
the analyses done in the earlier years showed a significant or at least consistently positive result
and it is only the later years when potentially susceptible subjects had been lost that no effect
was seen.

Section, page 9.37-38:  This is an interesting argument.  A convincing story is given for
making any estimate of a population threshold almost impossible to interpret, and then the
discussion proceeds to try (unsuccessfully) to get beyond linear. Not sure this make sense. Why
not simply leave it at a population threshold is irrelevant?

Page 9.39, line 21: Change to read: ".... morbidity effects, particularly as related to respiratory
diseases. Little evidence."

Page 9.45, lines 3-12: this is an indication of serious editing needed in this Chapter. This entire
paragraph is a repeat in content of what is said earlier in the chapter.

Pages 9.45-9.54. All  section labeled under .3:  This section seems far too long and
detailed for an "interpretive synthesis".

Page 9.56, last sentence beginning on line 6. It seems to me that the toxicology is not being done
to find a  causal relationship. The purpose is to understand mechanisms that explain the
associations that have been found.  The tone needs to  be changed to reflect this.

Page 9.67, para starting line 11:  This paragraph on CAPS, in contrast to those around it contains
only a generic discussion of CAPS rather specific citations of findings as is done for the methods
discussed in both preceding and following paragraphs.  This is an example of the lack of
consistent editing in putting together the chapter. Why cite specific experiments for one and not
the other unless a value judgment is being made? If so then  should be specified. If not a
consistent approach is required.

Section Mechanisms of action
       This section is far too  important to be limited to the space and detail presented,
particularly in contrast to the discussion before on both the aerometrics and the toxicology.  This
minimizes the importance of what the basic issue is.

Page 9.85, line 2. At end of sentence should add: "primarily because specific studies with
particles  of this size simply have not been performed."

Page 9.95, par In this paragraph there are a number of places when extremes are
suggested that really  are not justified. Line 5, "most" important factors  ...; line 10, potentially
"great" importance. I think these need to be toned down (not eliminated).

Page 9.98, Table 9.6. May I ask that the authors check again on the source for this table.  I am
concerned that the last two columns may be reversed in that I would have anticipated for all of
these conditions, not just acute bronchitis and pneumonia, that ages 65 + would have had higher
rates than 45-64 unless this represents a reporting bias.

                                Dr. Barbara Zielinska

             Review of revised Chapters 7, 8 and 9 of PM Criteria Document

                                   Barbara Zielinska
                                Desert Research Institute
                                    Reno, NV 89512

Comments on revised Chapter 9, "Integrative Summary"

   In my opinion, this chapter represents a significant improvement in comparison with the
previous version.  However, I have some concerns, which are  detailed below:

   1.  This chapter is still too long and not uniform.  Some sections truly present the synthesis
       of the information from previous chapters, whereas some other sections repeat the results
       of the individual studies. Some sections give the references to the individual studies,
       some not.  I suggest referencing those papers only that were not referenced in the
       previous chapters (this would greatly reduce 22 pages of references). I think that a
       consisted editing of Chapter 9 would be very useful.
   2.  The Chapter is lacking a summary concerning the ambient trends in PMio and PM2.5
       (when available) concentrations since the last 1996 review (i.e. some summary of
       Chapter 3). Also, a short summary concerning the changes in chemical composition of
       PM over the years would be appropriate. In my opinion, this is important information that
       would put the exposure data into a proper context.
   3.  Section concerning the coherence of evidences, discusses the Utah Valley study
       in great details. Although this study is certainly very valuable in showing the relation
       between health effects and emissions from a very specific source, i.e. a steel mill, it does
       not necessary relate to the general ambient PM in the U.S. As I pointed out in my
       previous reviews, the fact that the hospital admission was greatly reduced after the steel
       mill closure, indicates that  not all  ambient PM is created equal and that the PM health
       effect depends greatly on its sources and chemical compositions, not only mass
       concentration. So,  it is not necessary "coherent" with the results of epidemiological
   4.  In addition, as I pointed out before, the Utah Valley instillation study used the extracts
       from the 10 years old PMi0 filters. The filters were stored in plastic sleeves at room
       temperature and humidity and they might be subjected to significant artifacts over the
       years of storage in these inappropriate conditions.
   5.  Although Section that discusses the inhaled particles as potential  carriers of
       toxic agents is very interesting, it  doesn't really belong to this Chapter, but rather to
       Chapter 7. Only a short summary should be included in Chapter 9.
   6.  Minor comments:
       Is Appendix A necessary?
   -   Page 9-23, line 14:  This should be PMi0-2.5 rather than PM2.5.

   -   Page 9-30, line 30-31 and page 9-31, line 1-2.  The statement that gaseous co-pollutants
       act as surrogate of PM is highly speculative

   -   Page 9-65 says that the organic constituents remain of concern regarding PM health
       effect due in large part to the contribution of diesel exhaust particles to the fine PM
       fraction.  However, the organic compounds emitted from other combustion sources, such
       as gasoline vehicles, are not very different... In addition, little is known about the
       atmospheric transformation products of different emissions. One would argue that
       organics remain of concern because some of them are known air toxics, regardless of the

On a positive note, I found the discussion regarding caveats and limitations in interpretation of
human and laboratory animal PM exposure data in Section (approaches to experimental
evaluation of PM health effect) very valuable. However, this seems to be a stand-alone section,
and its conclusions are not reflected in any of the subsequent discussion.  On the contrary, the
Chapter uses excessively such adjectives as "considerable", "strong", "extensive",  etc., when
presenting the evidences of PM health effects.

Comments on revised Chapter 7, "Toxicology of Particulate Matter in Humans and
Laboratory Animals"

   In general, I found this chapter significantly improved in comparison with the previous
version. I have a few minor comments:
   1.  Although some limitations of CAPs exposures described in this Chapter (such as
       ineffective concentrations of particles smaller than 0.1 microns in diameter and
       elimination of gases by the concentrators used at that time) are pointed out on page 7-24,
       these limitations are not reflected in the Summary section.
   2.  Section, page 47-49 discusses the Utah Valley study; see my comments No. 3 and
       4 above.
   3.  Page 7-78, line 19: term HULIS includes not only bioaerosol, but also most probably
       some combustion generated and secondary organic compounds.
   4.  Page 7-159: Section is really
   5.  Page 7-180, lines 1-8. It should be benzo[a]pyrene, not  benzo[z]pyrene. Polycyclic
       aromatic compounds (PAC) are ubiquitous in the environment and their sources include
       not only noncatalyst-equipped gasoline vehicles (not very common today or even in 1993
       when the ambient samples were collected) but also catalyst equipped gasoline vehicles
       (especially older and higher PM emitters), diesel vehicles,  natural gas combustion, and
       other combustion sources. 2-Nitrofluoranthene is the secondary product of atmospheric
       reactions worldwide (see,  for example, Ramdahl et al., 1986) not only in the Los Angeles
   6.  Page 7-181, line 22: benzo[a]pyrene, not benzo-a-pyrene
   7.  Page 7-182: what are highway-emitted PAH-like compounds?
   8.  Page 7-197, line 16-18: decreased mutagenicity suggested  nitroarenes? This doesn't
       make sense.
   9.  Page 7-198, line 25-26: organic fraction is more mutagenic than the PM component?
       Which PM component?  Isn't organic fraction a part of PM?
   10. Page 7-199, line 12: ethylene
   11. Page 7-210, line 32 and 7-211, line 1: again, strange statement: organic fraction appears
       to be more mutagenic than the PM component (see 9 above)

    12. Page 7-212, line 1-3: It is difficult to rank mutagenicity according to the fuel type; rather,
       the efficiency of combustion processes may play more important role.

Ramdahl,  T.,  B.  Zielinska,  J.  Arey,  R. Atkinson, A.M. Winer and J.N. Pitts, Jr.  (1986).
Ubiquitous Occurrence of 2-Nitrofluoranthene and 2-Nitropyrene in Air. Nature, 321, 425-427.
Comments on revised Chapter 8: Epidemiology of Human Health Effects Associated with
Ambient Particulate Matter

       The revised Chapter 8 represents a significant improvement in comparison with previous
versions. I'm not an expert in epidemiology, but I found Section 8.4 and 8.5 readable and very
useful. In my opinion, this chapter is ready to be accepted as final with some minor revisions. I
have only  a few comments:
    1.  The paper by Hoek et al. (2002), if left in the final version in Section (p. 8-124 -
       125), should be appropriately qualified. NO2 is not a good tracer of motor vehicle
       emissions; in fact it is mostly a photochemical reaction product of NO, which is emitted
       by motor vehicles (primarily diesel vehicles). There are also other sources of NOx (for
       example, power plants) and the assumption that 50% of NO2 is coming from motor
       vehicles may or may not be correct, depending on the specific area. In addition, the
       exposure data in this paper were estimated, not measured.
    2.  The statement that Coefficient of Haze (CoH) is a good PM index for motor vehicle
       sources  (Section, p. 8-66) is odd. CoH is not a PM component (p.8-66, line 25-
       26), but it is related to the amount of fine PM in the atmosphere. It reflects the
       contribution of all sources, not only motor vehicles.
    3.  Page 8-72, line 12. Pb is no longer a motor vehicle tracer in the U.S., since unleaded
       gasoline has been used since over 20 years.

                                 Dr. Jane Q. Koenig
Comments of Ch 9
Jane Koenig
July 10, 2004

In my judgment the Integrative Synthesis is well written. However I continue to believe,
strongly, that these documents should be much more concise than they are. I don't see the need
to repeat data in this chapter. Supposedly the description of the data was accomplished in earlier
chapters. Just as a matter of interest, and to reinforce my point, I refer EPA to a recent American
Heart Association Scientific Statement on associations between heart disease and air pollution
(Brook et al. Circulation 2004; 2109: 2655-2571. In 17 journal pages, there is an excellent
description of the deleterious effects of ambient air pollution on health and its relation to heart
disease and stroke. I believe that journal review would convince most people of the association
between PM and cardiovascular disease.

Page 9.5  Should there be some mention  of the fact that some individuals voted for a PM1.0

9-10  Table 9-1 is a good addition as is Figure 9-3.

9-18  I don't know if the quote from the 1996 CD makes the point desired. It seems a true
      statement today.

9-35   Selection of lag certainly can obscure associations. But what is the alternative?

9-38   The findings from intervention studies may make the strongest case for a more stringent

9-39   to 9-71.  This seems to be mainly a rehash of text from  Ch 8. (and 7)

9-72.  The format seems to change at this point. What is the significance of the Bold heading?

9-80.  introducing the concept of coherence is good. Might use some sort of pyramid of effects
       from changes in pulmonary function to inflammation -to hospitalization- to death?

9-85 In the second full paragraph the authors compare exposures to doses.  This text should be
cleaned up.

9-95.  Attempts have been made to rate the public health impacts of air pollution.  Eg. Report
       of an expert panel to review the socio-economic models and related components
       supporting the development of Canada-wide standards for PM and Ozone (J toxicol
       Environ health, 2004; 7 (3).

 This section brings up the association between PM and lung cancer and also between PM and
diseases of infancy but doesn't seem to work these data into the general conclusions.

One section  that may be lacking, is a description of short term effects based on PM metrics of
less than 24  hours. These data are needed for consideration of a short term averaging time for

In general with a little editing I judge Ch 9 to be sufficient for its purpose and recommend that
we not attempt to make it reflect individual tastes but keep an eye on the big picture. Does this
chapter adequately reflect the material presented in the CD and does it give EPA an adequate
synthesis for writing the staff paper?  I think it does.

I have no additional comments on Chapters 7 and 8 beyond what I said in the past.

                                Dr. Petros Koutrakis
Chapter 7

Page: 7-4, Line 8:  The reproducibility of particle composition from day to day may be a
problem; however, one could take advantage of this and investigate associations between health
effects and composition. The Godleski CAPs studies have been very successful in doing this.

Page: 7-4, Line 10: To  date there are coarse and ultrafine concentrators which can make it
possible to expose subjects to coarse and ultrafine CAPs.

Page: 7-6, Line 2: We are now conducting a study where we investigate the toxicity of primary
and secondary particles  from coal power plants, the TERESA study. The toxicity of primary
particles may differ from that of secondary particles due to the photochemical reactions that can
result in many reactive compounds.

Page: 7-50, Line 15:  I am not sure that there are many atmospheric studies that have examined
the relative concentrations of water soluble and non soluble metals in PM. If such a strong
statement is made then it is necessary to back it up with some data. My guess is that PM metals
are mostly in the form of sulfates and nitrates and are quite soluble.

Page: 7-148, Line 14: Godleski and his group used a similar exposure scenario for MI rats.
CAPs were found to reduce HRV. Ozone was not found to have a synergistic effect. The
experiment needs to be replicated before publication.

Page: 7-155, Line 18: Mass is not the  only important parameter. Composition may also play an
important role.

Page: 7-204, Line 16: Include reference

Page: 7-204, Line 20: Include reference
Chapter 8

Page 8-1, Line 5:  I congratulate the authors for their hard work. I am sure that it was quite
difficult considering the plethora of information that was reviewed. Although the text was very
long, it was nicely structured so it was relatively easy and enjoyable to read.

I have made many individual comments but my two major ones are the following:

1) Although I think we need to set up a coarse particle standard, I am not sure that the available
information is sufficient or convincing. One can argue that for the previous PM standards we did
not have enough information  and that the decision was based on our scientific judgment.  For
example  for the fine particle  standard we  had only a handful of papers but it made sense to

prepare a new standard and we did so. In the case of coarse particles  we  do have  some
information but is not coherent. I hope that it will be kept in mind when final recommendations
are made about the levels of the new standard.

2) My second major comment is about the interpretation approach. In the beginning the chapter
claims that no calls will be made about which methodology is right or wrong and I absolutely
agree with this. As the authors suggested, the conclusions  will be based on the collective
available information. However, I found  that the rule was not followed consistently. For
instance, some individual papers  received too much attention in spite of what relatively little they
had to say, while  the chapter did not try hard to  synthesize information  from large  studies.
However, this was not done intentionally to prove a particular point.

Below please find my individual  comments:

Page 8-13, Line  12:  This  is  an important point  and represents the cornerstone of the
epidemiology chapter. I agree that this is the only way one can approach such a complex issue.
However, the evaluation process  could become subjective and one should be cautious.

Page 8-19, Line 15:  I assume that the CD refers  to exposure error. If this is the case I suggest
calling it as such or exposure misclassification.

Page 8-19, Line 16:  "Functional" refers to dose-response relationship?

Page 8-19, Line 19:  Of course, a major criticism of the epidemiological studies during this time
was  the biological plausibility. One remembers  that very little toxicological information was

Page 8-  49, Line  7:  It is apparent from both the  multi-city and single-city studies that risk
factors are higher for the Northeast.  This is quite an  interesting finding and is in agreement with
the Boston  and New York CAPs  studies.

Page: 8- 54, Line 2: Attributing effects  to the difference of sampler sharpness is pushing it and
is scary!

Page: 8- 54, Line 26: Measurement error is so vague. Can you please specify?

Page: 8-57, Line 1: Although I believe that we need a coarse particle  standard, I think that cited
literature on coarse particle health effects is not compelling at all. If Table 8-2 represents the
state of knowledge about coarse particles then we've  got a problem.

Page: 8-61, Line 28:  Most of the lower risk estimates are not statistically significant.

Page: 8-63, Line 17:  This statement was made before  and I had commented on it.  The way it is
presented here is confusing. It is not clear that it is the Mar analysis and not the Laden.

Page: 8-66, Line 11: The literature for ultrafines is extremely thin and this is reflected by the
above discussion which concerns one group and one study.

Page: 8-116, Line  12: From the acute studies one can see that Northeastern and Midwestern
studies show higher PM relative risks. Therefore, the higher risks found by the Harvard Six
Cities study are in agreement  with  the acute studies, since the six cities  are located in the
Midwest and Northeast.

Page: 8-128, Line 2: It is unrealistic to compare Dublin, APHEA, and Utah while expecting to
make perfect  sense.  We talk  about different populations,  climates and,  more importantly,
composition. So it is silly to draw important conclusions about acute and chronic effects.

Page: 8-128, Line  21: Is life  expectancy  per year the correct term? If so,  then the effect is
unbelievably huge.

Page: 8-129, Line  7: Maybe there are no effects associated with sulfur dioxide. Note  that the
other pollutants stayed more or less the same.

Page: 8-130, Line 3: I think we have sufficient findings which guide us towards the importance
of acute  effects on  mortality. Putting together the results  from the intervention studies, where
immediate benefits  can  be seen,  one  should start to think that acute effects/exposures are
substantially more important than the chronic ones. Furthermore, I was impressed by the second
follow-up of the Six Cities (not published yet) where concentration decrease in Steubenville
meant decrease in mortality risk. In fact when both the first and second follow-up data were put
together, Steubenville 1  and 2 looked  like two different cities.  Actually, the whole  analysis
looked like a twelve city  study.  The time between the two follow-ups was about 10 years so one
can conclude that the population managed to  recover in that time  span. Some other preliminary
analysis  by our EPA Center suggests that for chronic effects only the last year of exposure
matters!  I am really starting to believe that  maybe there are no chronic effects! We  become
susceptible because of age, genetics, food, smoking, income, and maybe air pollution but a daily
event can kill us. It may sound heretical at this point but the little evidence tells us that we should
start exploring this hypothesis.

Page: 8-131, Line 26: We need to mention that Six Cities and ACS cover different geographic

Page: 8-165, Line 5: How about the Six Cities/Factor analysis by Laden et al?

Page: 8-297, Line  10: Central site  monitoring data  may be inadequate to  assess population
exposures to coarse particles due to the high spatial variability of coarse particles. Also coarse
mass measurement error is higher than that for fine particle mass.  Therefore investigation of the
short-term effects of coarse particles and comparing them to those of fine particles is not a trivial

Page: 8-315, Line 6: Variability of concentrations may also be an important factor.  As bigger
day-to-day variations may result in a stronger health effects signal.

Page: 8-322, Line 22: I am surprised that this is all we get in this important issue. In the future
we need some more thinking in this area because this is the million-dollar question we all want
to answer.

Page: 8-326, Line 2: I do not think that the information provided about ultrafmes supports this

Page: 8-326, Line 5: The evidence provided by this chapter was so inconsistent, Therefore the
statement "may contribute under some circumstances ...."is not substantiated.

Page: 8-326, Line 17:  Road dust is an important type of coarse particle that may be of health
Chapter 9

Page: 9-1, Line 2: Overall the authors did a nice job. I have two main points:

1) The synthesis chapter should be shorter. One hundred plus pages is too long and defeats the
purpose of this chapter. It is not necessary to include  summaries of the different chapters, which
was done in some cases. One  should present the distillation of ideas outlined by the different
chapters. It is a difficult job but can be done.

2) In some cases I found that the synthesis chapter included information (or put emphasis) on
things  which were  not discussed (or  stressed) in  individual  chapters.  One example is the
hypothesis about radicals and peroxides. This is a hypothesis and nothing more and should not be
in the synthesis chapter.

My specific comments follow below:

Page: 9-1, Line 12: This is a good point.

Page:  9-6, Line 4:  The physico-chemical properties  of coarse and fine particles are  very
different. On this basis it makes  sense to  develop separate standards for these particle types.
However, one should not ignore the fact that epidemiological  and  toxicological information
regarding coarse particles is very limited. Similarly, very little is known about human exposures
to coarse particles.  The  situation was  similar when  the PM10 and PM2.5 standards  were
proposed and  I  am  not, therefore, overly concerned. However, I  am  concerned about the
heterogeneity of coarse particles both in terms of their origin and composition. Controlling road
dust and desert dust is not the same. This is an important issue that has not been addressed by the
CD. Maybe chapter 9 will deal with this somehow.

Page: 9-13, Line 6:  This is a  good discussion about PM exposure. It is somewhat speculative
due to  the lack of real  data, but the statements made are  sound. One important issue that was
omitted is the inter-home and  inter-city variability of the PM penetration efficiencies.  Studies

have found that homes behave differently and thus the same outdoor concentration may reflect
different exposures. An important factor affecting indoor/outdoor ratios is home ventilation that
depends on house characteristics, city, season, etc. This issue is not important for acute studies
but it can be for chronic ones.

Page: 9-18, Line 2: Does this include the re-analyses/confirmation of some studies: e.g. Harvard
Six Cities and GAMs?

Page: 9-25, Line 7: This is not clear to me.

Page: 9-27, Line 5: Too speculative to be included in this Chapter.

Page: 9-27, Line 12: I would say that there is no evidence for the ultrafine story.

Page: 9-31, Line 21: Nowadays, where SO2 power plant emissions are lower, SO2 may also be
tracer of diesel emissions. This is a thought!

Page: 9-37, Line 12: I found the lag section superb.

Page: 9-42, Line 23: Not enough information to understand Table 9-3.

Page: 9-43,  Line  1: Meteorology  also adds collinearity between species concentrations which
makes source separation very challenging.

Page:  9-45, Line  12:  This  section should  conclude that the  information  about  toxic
components/sources is non-existent. There are a couple of papers only and this is not enough.

Page: 9-47, Line  8: This is wrong! The majority  of CAPs studies  have done  extensive
characterizations of the particle exposures.

Page: 9-47,  Line 19: Who wrote this?! There are ultrafine and coarse particle concentrators!
Gases can be added as desired to the exposures. I strongly disagree with  the philosophy of the
CAPs section and strongly suggest that it be changed.  More importantly it does not reflect the
opinions written in the preceding chapters.

Page: 9-48,  Line  9: Again ROFA has its pluses and minuses but being completely negative is
not acceptable. Again this does not reflect the spirit of the rest of the CD. Furthermore, this  is the
synthesis chapter and one would like to know if one has learned something from, for example,
the ROFA studies.

Page: 9-48,  Line 23: Again, similar points as for CAPs and ROFA. This section came  from
nowhere. It does not belong here.

Page: 9-53, Line 1: As I mentioned above, I think the CAPs studies are meaningless.
The same comment applies to ROFA. We need to realize that there is no perfect approach, but
we learn little by little from each of them.

Page: 9-62, Line 9: This is not a recent study. In addition, many other epidemiological attempts
to implicate ultrafine particles by the same authors and others have failed.

Page: 9-67, Line 17: There are several papers by Clark et al, Batalha et al, Saldiva et al which
have shown associations between health outcomes and specific particle components, therefore
this statement is not correct.

Batalha, J. R. F., Saldiva, P.  H. N., Clarke, R. W.,  Coull, B. A.,  Stearns,  R. C., Lawrence, I,
Krishna Murthy, G.  G., Koutrakis, P., Godleski, J. Concentrated Ambient  Air Particles Induce
Vasoconstriction  of Small Pulmonary Arteries  in  Rats, Environmental Health Perspectives,
110(12): 1191-1197.
Saldiva, H. N., Clarke, Coull, B., R. W., Stearns, R., Lawrence, J.,  Krishna Murthy, G. G., Diaz,
E.,  Koutrakis,  P.,  Suh, H.,  Tsuda,  A.,   Godleski, J.  G. Lung Inflammation Induced  by
Concentrated Ambient Air Particles is Related to Particle Composition. Journal of
Respiratory and Critical Care Medicine, 165:1610-1617,2002.

Clarke, R.W., Catalano, P.J., Koutrakis, P., Murthy, G.G. Krishna, Sioutas, C., Paulauskis, J.,
Coull,  B.,  Ferguson, S., Godleski, JJ.   Urban Air Particulate Inhalation Alters Pulmonary
Function and Induces Pulmonary Inflammation in  a  Rodent Model  of  Chronic Bronchitis.
Inhalation Toxicology, ll(8):637-656,  1999.

Page: 9-67, Line 23: (See comment above)

                                   Dr. Allan Legge
July 18,2004

TO: Phil Hopke/ Fred Butterfield
FROM: Allan H. Legge

Review Comments: Section 9.3.2 Effects of Ambient PM on Vegetation and Ecosystems from
       Revised Chapter 9, Integrative Synthesis, Fourth External Review Draft, 'Air Quality
       Criteria for Particulate Matter, June, 2004'.

Overall Comment:
       The authors are to be commended for concisely bringing together the key scientific
information/understanding on vegetation and ecosystems from Chapter 4 as well as identifying
the important data gaps and uncertainties which currently prevent relating ambient
concentrations of PM to ecosystem response. The discussion of the potential application of the
'critical loads concept'  in the U.S. opens a very important philosophical/scientific door for
environmental protection in the future.
       Other than a few minor comments, this section is acceptable.

Specific Comments:
1. Page 9-108.
       i) lines 7-9.
       It is important to note that sulfur is also closely linked with nitrogen.
       ii) lines 11-13.
       It should be noted that plants can also gain nitrogen nutrition from the atmosphere by
taking up NO and/or NO2 .

2. Page 9-110, lines 6-7.
       Point (4) as written is too restrictive. If the authors are referring to HNO3, it can be
formed without the presence of sulfur compounds.

3. Page 9-110, lines 12-14.
       As phrased, this sentence seems 'overstated'. The intent of the authors is unclear. Later
sections of the text, for example, show that Nr can be removed from a terrestrial ecosystem by
NOs- leaching. One can reduce Nr accumulation by reducing the emissions of Nr sources.

4. Page 9-114, line 1.
       Suspect this should read  " Acidic deposition and acidification of soils can lead to high
Al-to-base cation ratios that limit"

5. Page 9-115, lines 14-15.
       What is the intent of saying "- -, rather than on those in sensitive ecosystems." Do the
authors mean to say "- -, rather  than on rural exposures."

                                   Dr. Paul J. Lioy
                            Chapter 9: Integrative Synthesis
                                         in the
             Air Quality Criteria Document for Particulate Matter - Draft #4

                              Comments of Dr. Paul J. Lioy
  Professor and Deputy Director of the Environmental and Occupational Health Sciences
                Institute, UMDNJ- Robert Wood Johnson Medical School

       The EPA has made a good faith attempt to provide an integrative synthesis of the results
presented in the Criteria Document for Particulate Matter and should be commended for their
efforts. I do have specific major concerns, but I believe that the Agency is on the right track for
completing Chapter 9.
       Chapter 9 presents concise summaries of the major findings found in the 1996 document;
however, I do not always find a synthesis or contrast with recent results. There is an exposition
of many findings, but there is  not always a clear message  as to how these have improved our
understanding and filled gaps, or can be used to put to rest specific issues.

       The results from Utah studies are extremely valuable, especially in terms of describing
source to exposure to  response relationships. They do not, however, represent the general or
overall PM2.5 exposure - response relationships seen around the country. Thus, I do not think the
results demonstrate "coherence' with respect to the general composition PM and health response.
The results from Utah do provide very valuable insights into source specific pollution situations,
and how reductions in a complex gas and particle mixture emitted by a particular source type can
improve the health of a local population. A  very important scientific  conclusion, and an
important exposition within  the context of accountability.   Thus, from the vantage point of a
chapter on "integrative  synthesis," the results  provide a  very  useful case study. It can be
developed as a concise stand alone subsection that uses the available data on health outcomes,
toxicology, ambient air quality and exposure, and source reductions

       In contrast, the results from reanalysis of the Six Cities study  on the reduction  of health
effects over time are compelling and must be considered (pg. 9-36)  as part of the  analysis for
national PM2.5 and other air pollutant issues. They show that the character of the air has changed
over entire cities. However,  the questions that require discussion are: What sources have been
eliminated or controlled.? What pollutants have seen significant reductions, and led to reductions
in health  effects,?  Important questions requiring some  analysis in a chapter on integrative
synthesis. To summarize, a concise analysis of the above can tell us something about the sources
which have contributed to PM  and  other pollutant related health effects. Further, the analyses
may provide insights on ambient pollution exposure characterization  needed for prospective
health evaluations among  US cities.

       I am still a bit troubled by the discussion on toxic element linkage to PM health effects.
The toxicological results for metals related health outcomes are for exposures that are orders of

magnitude higher that those found  in the ambient atmosphere. Such  a dichotomy  does not
provide much information for achieving "coherence."

       One issue that was given brief mention in the Chapter, has broad implications: the timing
of PM exposure with acute cardiac related health effects. This appears to be a major  scientific
finding that has been made since the publication of the, 1996 CD and the results strongly suggest
that we may need to re-assess the  form of the  standard,  and consider  an  acute PM2.5 health
standard that is 2 to 8  hr in duration. This can be in addition to 24 hr standard for short term
effects. The issue may need  a separate section.  Promulgation of an acute health standard may
not be feasible at this time, but the results indicate that more attention must be given to this issue
in the future. I hope the Staff Paper that results from a review of the final version of the CD can
at least provide discussion on an acute exposure standard

       Conversely, the GAM issue did in fact appear to lower the exposure - response estimates
for  the chronic  effects. What is  the  implication, if any, in terms  of exposure  response
relationships used in the 1996 standard setting process? Would it change the conclusions about
risk?  I do not think stating that there  is still a relationship is sufficient.

       The chapter on air quality is almost ignored in Chapter 9.  How has  PMio  and, where
comparative data exists, PM2.5 changed since the publication of the 1996 Criteria Document. In
its current form, the reader is left with the impression that we have major exacerbations of the
PM  standards.  That is just not  true.  I  think the issue needs to  be  correctly defined  and
characterized for the reader.  The text discusses the % increase in risk/ 25 or 50 ug/m3 increase in
PM2.5 and PMIO, respectively, but these values need to be placed into a context of the current
situation.  At the same time, the integration and synthesis should point out that controlling PM
and other  pollutants is reducing relative risks and these efforts need to be sustained in order to
deal with current PM health  issues, e.g. cardiac related health outcomes. This gap in the analysis
of information  provided in  the CD  can  be filled using  information  from the chapters on air
quality, and should be added as a separate section on page 9.13 (Short but to the  point),  and
appropriate revisions made  to the conclusions on page 9.16. As I reviewed  the outline for this
chapter, the requested changes fit well within the stated goals for Chapter 9.

Major Specific comments:

Page 9.3 lines 1 through 3. These numbered bullets do not make much sense. Please shorten the
length and strengthen the points.

Page 9.3 lines  4-5, what do you mean by a coherent synthesis? A confusing statement.
Section 9.2.2  Some effort needs to be made to differentiate short term and long term health
effects  with respect to the timing of the onset of an event. A statement on page 9-36 suggests
that we may need to start thinking about three standards acute - less than 24h, short term - 24
hours,  and long  term - annual. In a synthesis such information needs to be fully discussed to
continue the quest "coherence " within  the chapter. I am a bit puzzled by  the fact that that
summary and conclusions for this section are so short and focus almost exclusively on the size

fractions when the results presented in the section discuss lags, and long term and short term
effects. Overall, the section can use editing to reduce its size by about 25%

Figures 9.5 and 9.6 are excellent.

Section 9.2.3 This section is much too long. It can easily be reduced in size by 60% and still
convey the message.

Page 9.45 Again there are statements about acute effects lines 4 to 12 which need to be folded
into a synthesized discussion about the need to consider a shorter term, acute, standard.
Page 9-48 and others. I think that the conclusions need to make the point that metals exposures in
toxicological studies are at much higher levels that ever seen the ambient environment. I do not
think that this message has been clearly articulated on page 9-84 or 9-85

Please eliminate the appendix. These information have been thoroughly discussed in chapter 8.
You can reference chapter 8 for the interested reader.
Dated: July 14, 2004

                               Dr. Morton Lippmann

                                REVIEW COMMENTS

                              Chapter 7 of PM CD, Draft 5

                                     M. Lippmann
General Comments
             Chapter  7 is greatly  improved for the fourth external  review draft,  and now

represents a thorough and well-balanced review and analysis of a large, complex, and less than

definitive literature on the associations between laboratory-based PM exposures and biological

responses  in  humans,  laboratory animals,  and  in vitro preparations.   It  could certainly  be

somewhat condensed to eliminate repetitious text,  but that is not a serious failing at this point.

As it stands,  it can serve well as comprehensive  and unbiased  compendium and resource for

Chapter 9 (Integrative Synthesis) and the OAQPS Staff Paper.

       It now provides a more thoughtful  and appropriate view, than did the prior drafts, of the

nature and significance of the literature on: the short-term CAPs inhalation exposures; exposures

to motor vehicle exhausts and  extracts  thereof;  PM of biological origins; and interspecies

differences in PM dosimetry. Also, it provides a more thorough and thoughtful discussion on the

evidence supporting the plausibility of the epidemiology that points to cardiovascular causes of

excess mortality and morbidity, and to excess lung cancer.

       In my view, Chapter 7 now warrants CASAC endorsement and closure.

Specific Comments

Page         Line(s)        Comment
7-35          14-23        This discussion of a human exposure study seems to be out-of-
                           place in this  section.

7-117         10           What is a "vaporous milieu"?

7-150         17           The concentrations of MLjHSC^ and 63 should be specified.
7-166         15           change "tissue" to "region".  The use of tissue implies that there is
                           a uniform distribution of particles, which is far from the actual
7-192         20           Should a study published in 2004 be cited?
7-203         5            change "co-primary" to "the primary".
7-211         22           change "mutating" to "mutations".
7-198         26           What is meant by "the organic fraction being more mutagenic than
                           the PM component"? Is it organic in the PM or VOCs?
                            Appendix 7A of PM CD, Draft 5

General Comments

              This  revised  Appendix  provides  an interesting and useful  analysis  of the
similarities and differences in particle deposition,  retention, and dose between humans and rats
used in  controlled  exposure studies.   While many  of the complexities  are  presented  and
appropriately discussed, others are either not addressed or fully discussed, such as:

       1)     The substantial differences  in  intra-airway  distribution of instilled and inhaled

       2)     "no net absorption or secretion of mucus during transport" (p. 7A-8, lines 17 and
              18). This statement is clearly incorrect.

       3)     An  inconsistent  use  of ICRP,  1994  assumptions,  e.g., not  using its  two-
              compartment  bronchial  region clearance,  while  using its  three-compartment
              alveolar region clearance rates.

   Despite the somewhat arbitrary decisions incorporated in the MPPD model, the human-rat
   comparisons were made using the same variables for both species, and can be considered to

   be useful and informative in understanding the applicability of published animal experiment

   data to the interpretation of human health risks.
Specific Comments
7 A- 10
7 A- 10
7 A- 11
7 A- 14
I challenge the statement: "For instillations into the lung, the dose
can be characterized fairly well." Does this mean the "delivered
dose"? It is necessary to caveat such a statement, acknowledging
that the dose distribution in an instillation is highly nonuniform
and brings PM into dependent airways without the preferential
deposition at bifurcations and in centrilobular airways that occurs
with inhalation.
change "(RIM)" to "(RIVM)".
insert "most" before "humans".
change "monopodial" to "monopodal" here and in subsequent
insert "related lung ventilation" before "level".
This sentence should be deleted.  There was a typo in the HEI
report that was cited, leading to the misinterpretation here.
change "is effected" to "onset is determined".
                              Chapter 8 of PM CD, Draft 5
General Comments:
       As befits a fifth iteration, this draft provides a thorough and balanced review of a very

large  literature of observational studies  of associations between  ambient particulate matter

concentrations, compositions, and particle size distributions and human health  effects.  It also

provides quite thorough and balanced discussions of the adequacy and interpretability of models

and exposure-response relationships that have been used to describe the associations and their

strengths and limitations.  The section on exposure characterization error is a particular strength

on an issue that really needed a more thorough and dispassionate analysis than was  previously


       In my view, Chapter 8 now warrants CASAC endorsement and closure.

Specific Comments:

1.      page 8-51, line 30:  change "Detroit" to "Detroit Metropolitan Area" or to "Windsor, Ont,
       adjacent to downtown Detroit"
2.      page 8-60, line 19:  change "at" to "below"
3.      page 8-291, line 21:  clarify the meaning of "more sensitive"
4.      page 8-292, line 27:  change "PM2'5" to "PM2.5"
5.      page 8-306, lines 7-15:  This whole paragraph is much too speculative. Delete it.
6.      page 8-315, line 2:  change "merely" to "solely"

                              Chapter 9 of PM CD, Draft 5

General Comments

       This revised chapter now warrants the  title of "Integrative Synthesis".  It is clearly
written and successfully draws forward the most important elements of the preceding chapters
that  should inform subsequent decisions  affecting the selection of NAAQS for PM.  It is,
however, too long, and would benefit from condensation by about one third to one half, with
notes that the details are available in the preceding chapters.

       I agree with Dr. Crapo that Figures 9-1 and 9-2 leave the impression  that 2.Sum is an
inappropriate cut size for fine particulate matter.  It  would  be  much  better  to illustrate the
minimum in the size-mass continuum with data more representative of typical urban and regional
aged air masses in relatively humid climates where aging and hygroscopicity lead to larger
aerodynamic particle size minima.. The discussion  of these issues should also  provide more of
the background that contributed to the selection of 10 and 2.5 um as  cut-sizes, and that it was

recognized that they were selected, at least in part, as conservative choices from a public health

protection standpoint. A somewhat smaller cut-size than lOum would probably better represent

average penetration into the thorax,  and a somewhat smaller cut-size than 2.Sum would better

represent the minimum for dry air conditions and freshly generated PM, but they would be less

public health protective for conditions and people at the upper bounds of normal variations. In

this context, they remain prudent choices for NAAQS criteria.

       I was particularly impressed with the synthesis of the evidence in Sections on

"Inhaled particles  as  potential  carriers  of  toxic  agents," and  on "Coherence of


       The Section text on particle bound water provides a framework for addressing

the apparent paradox of having a substantial epidemiologic database exhibiting significant health

responses to current ambient PM exposures and a rather meager supporting  literature from

toxicology.    While  most of  this  section  draws  its  evidence from  physical  chemistry

considerations, and therefore  not from Chapter 7 on toxicology, which  is appropriate  for an

integrative  synthesis, it does cite one important toxicological study, i.e., Morio et al. (2001), a

study not cited in Chapter 7. I therefore strongly recommend that Chapter 7 be revised to feature

this study and to briefly summarize that part of Chapter 2 that provides a basis for the Morio et

al. (2001) study. I would also like to  see a broader discussion of particle bound water in Chapter

9 and in earlier chapters,  including its influence on positive artifacts in mass concentration

measurements, and on chemical transformations within airborne particles.

       Section on the coherence among the  human population studies, and  of the

epidemiology with  the relevant  toxicological  studies,  provides a  concise  and important

summation of the overall  biological plausibility of adverse human  health effects  caused by

exposures to  ambient  air PM,  and provides important  support for NAAQS that provide

equivalent, or perhaps more  stringent, public  health protection than the current suite  of PM
Specific Comments


"known" is misspelled.
insert "exposures" after "short-term".
There have been no studies in "Central America". The authors
presumably were referring to studies in Mexico, which is in North
change "PM2.5" to "PMio-2.5".
change "Figure 9-5" to "Figure 9-6".
insert ", Ito, 2003" after "Lippmann et al., 2000)".
bifurcation "hot-spots" are not limited to the "TB tree" but also
occur on bifurcations of respiratory bronchioles and alveolar
ducts, as demonstrated by Brody, Warheit and their colleagues.
"inhalation" and "CAPs" should be transposed.
change "larger" to "smaller".
add "Gauderman et al., 2002 and Avol et al., 2001" as support for
the statement.

                                  Dr. Joe Mauderly

Review of Chapter 7, PM CD                                              Mauderly

General Comments

Overall, the chapter is improved from the last draft. Numerous previous content issues have
been resolved, and the text is much cleaner.  The issue of exposure-response could still use
some tuning up.

Specific Comments

   P 7-8, L 9: Is there a reference for a study showing that cardiovascular effects can be due to
           direct PM uptake into the blood, or is this just a recitation of hypotheses?

   P 7-8, L 17:  It's not the oxygen-carrying "capacity" of the blood, in any conventional sense,
           it's the reduced blood flow.

   P 7-17, Table 7-2: Here and in following sections, the city is used as a descriptor of ROFA.
           This isn't useful.  While city might be a somewhat (if questionably) useful descriptor
           of CAPs, ROFA varies in composition from plant to plant and from feedstock to
           feedstock. Our own experience demonstrated that two samples of "Boston" ROFA
           from two units of the same plant at two different times had different compositions
           and biological effects.  I recognize the difficulty of describing ROFA - in fact the
           original papers typically provide very poor descriptions. However, using the city is
           both meaningless and misleading.

   P 7-77, L 21:  By definition (and in the study cited), semi-volatile materials are in both the
           PM and vapor phases.  It should be "—PM and vapor-phase semi-volatile -".

   P 7-153, L 25-28 (and entire section Understanding dose-response, and the extent
           to which we can determine it from present data, is important. This section makes an
           attempt to address previous calls for a summary of the subject, but needs  some tuning

   The Ohio et al., 200a citation is a good example. The present statement does not say, but the
           wording implies,  that increases in fibrinogen were observed at 23 ng/m3,  which is not
           true; however, the study did yield some very useful perspectives on exposure-
           response. Those ought to be described. First, it should be stated that any
           comparisons using exposures to CAPs at different times do not comprise  a true dose-
           response study - CAPs are not identical from day to day. Moreover, any study that
           does not expose the same individuals at different levels (Ohio did not) cannot be a
           true dose-response study. Given that, if you are going to cite the study in this
           perspective (which I recommend that you do, with appropriate caveats), there is more
           to be learned from it than you report. Ohio et al. divided their exposures into  three
           groups (tritiles) with a fourth group as controls. They did not find a progressive

       exposure-response relationship for fibrinogen, but they did find fibrinogen to be
       increased slightly for all tritiles of exposure. They wrested statistical significance
       from the response only by lumping all exposure groups - which is fair game - but the
       slight increase was observed even at the lowest exposure tritile, which was a mean of
       47 ng/m3. Even given the caveats, that's an important perspective. Moreover, they
       found a progressive increase in BAL neutrophils with exposure, and again, the
       increase clearly started at 47 |J,g/m3.  This latter finding isn't showcased in this
       section, but is more important than most of the factoids that are.

The Godleski et al. (2000) paper, on the other hand, doesn't give results in such a manner
       that one can infer exposure-response relationships between total CAPs and
       responses. They show regressions with components of CAPs and eke out some
       significant relationships.  However, stating that they found significant effects at
       CAPs exposures from 100 to 1000 |J,g/m3 is misleading. They found effects from
       exposures that encompassed that range, but the paper gives no indication of the
       lowest exposure level producing significant (or strongly suggestive) effects. The
       study was useful, but not for this purpose.

The point is that when discussing exposure-response relationships and LOELs, it is
       misleading, if not outright disingenuous, to cite significant effects across a range of
       exposure concentrations without citing the concentrations that caused the effects. If
       the cited report doesn't let you do that, then it's not a useful citation for this section.

P 7-159, L 1: These are exposures, not doses.

P 7-166, L 6 and Table 7-14: It is true that the Ohio et al. (2000a) study showed that
       exposure to CAPs at a mean of 120 ng/m3 cause more than a two-fold increase in
       BAL PMNs (they called them neutrophils). However, the more meaningful
       information  is that the increase began at a mean of 47 |J,g/m3.

7-186, L 3 - 7-187, L 17: Although the information here isn't wrong, it pertains to coal
       "smokes" that have little, if any, relevance to components of coal emissions to which
       the public is exposed in the U.S.  If the examples are going to be retained, a sentence
       should be added to make this clear.

7-187, L  19:  The section on mutagenicity of mobile source emissions says nothing about
       emissions from natural gas-burning engines. There are now multiple reports
       indicating that natural gas emissions are also mutagenic. Although some are more
       recent that the cut-off date for the CD, an earlier paper that could be used as an
       example is Lapin et al., Mutation Research 519: 205-209, 2002. I'm not suggesting a
       big section on natural gas, but a paragraph indicating that such data exist and such
       emissions have mutagenic activity would be appropriate.

P 7-197, L 15-18: As written, this sentence doesn't make sense. The paper reported that the
       mutagenicity of the "cold diesel" sample was decreased by addition of S-9. The
       sentence left out the S-9 part.

   P 7-201, numbered list:  One of the key contributions of toxicological studies (if they are so
           designed) is information on exposure-response relationships. While it's true that few
           of the studies cited in the chapter were designed to accomplish that, the point
           certainly belongs in a list of what role toxicological studies "can play".

   P 7-210, L 32, and next page:  The organic fraction is part of PM. If the organic fraction
           came originally from collected PM (as is the case for nearly all citations in this
           section) it is not appropriate to speak of the organic fraction as if it were something
           other than PM. You can talk about the extracted organic fraction of PM vs the rest of
           the PM, but they are both "PM".

   P 7'-217, L 14: "Dosimetric" should be "dose metric".

   P 7-218, L 24-25:  You say "several"  controlled human exposure studies provide interesting
           findings, but you cite only two, and neither of them provide much insight into the
           relative roles of PM and co-pollutants. If you include all studies in which humans
           were exposed to mixtures of PM and co-pollutants, but don't test their relative roles,
           then there are several more.

   P 7-219, L 5: Here, six studies are a "few", but on the previous page,  two studies were

   P 7-221, end: You are missing a "bottom line".  You've gone through a recitation of an
           assortment of studies using "susceptibility" models, but are any summary
           conclusions to be drawn from the collection?
Review of Appendix A, Chapter 7, PM CD                                  Mauderly

General Comments

The appendix is somewhat improved, but remains denigrated by containing comparisons that,
while perhaps mathematically correct, are seriously flawed in concept. Calculating that a rodent
exposure to a single-mode aerosol would have to be very high to achieve deposited or retained
doses similar to those achieved by humans exposed to multi-modal aerosols at the roadside is
simple not helpful. No human is exposed to that material in those multiple modes. We do not
need a dosimetric appendix if its conclusion is that, if one only picks the right mathematical
comparison, any  dose can be justified. We already knew that.

The appendix contains information on comparative biological responses that belongs in the main
body of the chapter. It is exactly for the purpose of making sense of comparative responses that
the comparative dose modeling is required.  The purpose of the appendix is to provide a
foundation for judging the value of animal data and comparing animal and human responses in
the body of the chapter.  The health  response comparisons belong there, not here.

Specific Comments

   P 7A-1, L 26:  It should be "—equivalent to those -".

   P 7A-2, L 15:  It should be "—use of these -".

   P 7A-3, L 10:  This is only true if you believe (or hypothesize) that the effect is due to
          retained dose, rather than the most recent dose. Some effects probably are, and some
          probably are not. If the statement is strictly true, then the interpretation of the time
          series studies must be wrong. If so, and because all humans are exposed chronically,
          how could effects be ascribed to only the most recent dose?

   P 7A-3 L 25: It should be "—the total dose—". You can't characterize regional doses (e.g.,
          TB vs A) at all.

   P 7A-6, L 21:  "Disposition" ordinarily also includes translocation within the organ.  Does
          "clearance" in this appendix mean only removal from the organ?

   P 7A-6, L 24:  It should be "RIVM".

   P 7A-8, L 11:  Again, does "clearance" here mean removal from the alveolus, or removal
          from the organ. Material can be removed from the alveolus and still be in the lung
          (interstitum, lymphatics, capillaries, etc.).

   P 7A-8, 20-22:  Of course, it is known that mucus does not flow evenly up all airway
          surfaces. Indeed, "streaming" of mucus and the PM it carries is a well-known and
          almost universal  phenomenon.  Hopefully, the modelers know this, but it should be
          mentioned as a caveat.

   P 7A-9, L 7-8: According to the dissertation on overload, as reflected in Table 7A-13, even
          the lowest of these exposures exceeds "overload" and the higher two are well beyond
          "stasis".  Why would one select for the MPPD model alveolar clearance rates that are
          thought to reflect "overloaded"  kinetics?  Moreover, why should we believe the
          model results for cumulative lung burden if they are all based on "overloaded"
          clearance kinetics?

   P 7A-9, L 21:  The relevance of the point is not clear. Why would one worry about 10-
          minute computations on one computer when many people have spent over a year
          reviewing and revising this chapter?

   P 7A-11, L 26-28:  Why would one introduce a peculiar terminology for PM size in this
          chapter?  Here, we have "coarse" including  everything over 1.0 jim, and
          "accumulation mode" going downward from 2.5 |im, thus making PM2.sboth
          "coarse" and "accumulation mode". That hardly fits with the descriptions in the
          remainder of the Criteria Document.

P 7A-11, L 28: Here we have an inkling of the bias to come - that rats are "normally" only
       exposed in certain ways (limited to PMfme CAPs and resuspended ROFA). That's
       pure bunk. Rats are exposed to most every size distribution imaginable - sometimes
       (unfortunately) including sizes they can't even inhale! Rats are certainly often
       exposed by inhalation to atmospheres simultaneously containing PM ranging from
       well above 1.0 jim (remember, you said this was "coarse") downward to below 10
       nm. The authors can  showcase the EPA ROFA studies if they like, but they
       shouldn't posit that this is how rats are "normally" exposed.

P 7A-19, Fig. 7A-5:  Something is wrong with the figures, the legend, or both. The time
       scale says "post exposure hours".  If this is retained mass after exposure, then why
       isn't the mass at time  zero 1.0?  Only the amount of mass present in the TB region
       (not the "retained" mass) could increase after time zero (the end of exposure) by
       virtue of more entering the region from the A region than was originally deposited
       there. Regardless, the amount at time zero cannot be zero, as the figure shows,
       unless there was no TB deposition at all.

P 7A-25, Table 7A-5:  The sources of the values used for alveolar surface and FRC ought to
       be stated explicitly. These values are critical to the modeling presented later, and the
       explanation needn't be more than a paragraph.  An attempt to trace the sources
       proved difficult and did not engender confidence.  This is particularly important
       because the surface area values here are much lower than those that are frequently
       cited.  If human alveolar surface is really only 57 m2, then why do physiologists
       usually assume 70-100 m2? Rat surface is often cited at  0.4-0.5 m2.  Of course, the
       most correct value is unknown- all values are modeled estimates that involve several
       assumptions.  The point here is  not that the values are wrong -1 can't tell that - the
       point is that there needs to be an explicit statement of how the values were derived
       (the experimental and modeling foundation, not just a combination of primary and
       secondary references).

Yeh and Schum (1980) are cited as the  source of the surface area value for humans, yet I
       can't find a value for  alveolar surface in that paper (airways yes, but not alveoli).
       That paper reports modeling data from one human lung cast at presumed TLC, and
       assumes the values for FRC, TLC, and number of alveoli used by the ICRP (1974)
       for "standard man" (2.2L, 5.6L, and 3xl08, respectively). Table 7A-5 calculates a
       value for surface area from that paper, but uses a different FRC (3.3L) and cites the
       RIVM report as the source. The RIVM report lists that value for FRC, and cites
       ICRP 1994 as the source (apparently, ICRP's standard man grew during the ensuing
       20 years). Interestingly, in the section on age effects, the RIVM report cites an FRC
       of 1.8L for 21 years and cites 6 different references (I don't have time to read all  of
       them to divine where  that particular value came from). That value is lower than the
       one cited for 18 years (I thought you had to be over 30 to begin shrinking).  The
       RIVM report also cites Gehr (1978) for a human surface area of 143 m2. For rat
       surface, the RIVM report cites a Mauderly (1979) review, which cited Johanson &
       Pierce (1973), who reported that in Sprague-Dawley rats, it ran from 0.55 to 0.88 m2,
       depending on age.  That's a lot higher than the value in Table 7A-5, which cites Yeh

       et al. (1979) as the source.  Those authors reported modeling results using data from
       a TLC lung cast of a female Long-Evans rat and, like the paper on the human cast, no
       value for alveolar surface is given. Table 7A-5 calculates a value for surface from
       the modeling results, and used a value of 4.0 ml for FRC, citing the RIVM report for
       that value. The RIVM report does use that value, but doesn't cite the source.

P 7A-26,1st 4 lines: This information is confusing, and only partially matches with the
       comparisons that actually follow.  There ought to be a simple, factual, list (1,2,3,
       etc.) of the comparisons that follow. For example, the first comparison does indeed
       compare rats and humans one size mode at a time (if I understand it correctly).
       These sentences imply that this is not done at all.

P 7A-28, 29:  These two tables, and those that give the assumptions behind them, comprise
       the most useful material in the appendix.  Unfortunately, the comparisons that follow
       are less well-founded, more confusing, and less helpful. One could end the appendix
       here and have provided a service.

P 7A-30, L 1-4: At about this place in the text, a very important concept that is completely
       missing from the appendix should be clearly stated. Of course, one can expose rats
       to PM having a size distribution including the coarse fraction in such a manner that
       the deposited burden is the same as that estimated for humans.  It may take  a huge
       concentration, but you could do it. However, the  rats would only be  inhaling the
       inhalable or respirable portion of the material. What that means in practical terms is
       that unless the material is homogenous in composition through out the size
       distribution one is depositing in rats a different average composition  than would be
       deposited in humans.  Not only that, but the number of particles and  area of particle
       surface would differ between the species at the same deposited mass dose.  The
       bottom line is that no matter what heroics you attempt in order to deposit the same
       mass, you will be depositing different "stuff (on  average) in the two species.

Stating this concept clearly is not only important to the discussion at hand in this section, but
       it is also important because it shines light on a problem that infests the rest of the
       appendix. For example, this concept is totally ignored in the comparison of
       resuspended ROFA in the rat to roadside exposures of humans. Who cares  what
       exposures would produce the same dose? The stuff is not, and cannot be, the same.
       Humans don't breathe ROFA in a tri-modal size distribution at the roadside (or
       anywhere else). It makes no sense to compare any single material  in the laboratory
       to environmental exposures to PM having a distribution of composition and size.
       Dosimetric comparisons in that case become academic exercises that have little, if
       any, basis in real biology. Moreover, the exercise becomes  a distraction at best and
       misleading at worst in an appendix that is supposed to bring clarity to comparisons of
       animal and human study results and potential hazards.

P 7A-30-31:  This and the next sections are troublesome.  It's hard to come up with a
       scientific reason that these comparisons are included.  The authors need to take off
       their ROFA glasses. The majority of experimental PM work over the past 40 years

       did not use resuspended material. As an example, the studies used for the basis of
       the deposition and clearance assumptions in the MPPD model certainly did not -
       they used generated PM having designed size ranges, and other studies have used
       resuspended PM having a monodisperse size distribution.  The statements about
       cyclones are irrelevant to either.

It is stated that it would not be appropriate to use resuspended PM to simulate human
       exposures or doses. That depends on the question you are asking and the design of
       the experiment. Of course it would be appropriate - if you are comparing exposures
       to and effects from the same material (i.e., apples to apples). It only becomes
       inappropriate if you are comparing exposures to different materials (apples to
       oranges, or ROFA to roadside), in which case your conceptual troubles are much
       bigger than your dosimetric troubles. Among the mixtures that people breathe are
       2.0 |im MMAD PM having a certain composition and size distribution - perhaps like
       your resuspended material. In fact, people are breathing many different PMs having
       different compositions and size distributions - not just one material having a trimodal

It's ridiculous at the outset to think that you can do a laboratory study of a single type of PM
       and represent human exposures to a diverse population of PM - matching the mass,
       number, or surface dose is not the issue. On the other hand, if you have PM
       representing some portion of what people breathe, you can study that fraction,  and
       reasonably propose that the results are relevant to that portion of the mixture that
       people are breathing.  Conversely, if you are exposing animals to material that does
       not represent a portion of what people breathe, you shouldn't be worrying about
       doses- you are testing some hypothesis that can't be extrapolated to humans on a
       dose basis anyway.

(Moving on through Section 7A.5.1.4):

It is stated that it is not appropriate to compare a rat dose from one PM size fraction to a
       human dose  from the  same size fraction, because people breathe all size fractions.
       First, on the  basis of the above rationale, that's bunk.  Second,  regardless of whether
       or not that's  bunk, it is certainly not appropriate to compare a rat dose from one PM
       type and size fraction to a human dose from many PM types encompassing multiple
       size fractions. Why would one compare such doses? Implicit in the comparison is
       that neither composition nor size matter to toxicity.  Nobody believes that. The only
       apparent reason to make such a comparison is to justify using extreme doses.

ROFA is a good model  PM for testing certain hypotheses about soluble metals. It is
       ridiculous to propose ROFA as a surrogate for total  environmental PM exposure - or
       even to total combustion-derived environmental material.

P 7A-31, L 32  and P 7A-32, L 1:  The wording used here either states, or strongly implies,
       that human clearance  is never  affected by lung burden and rat clearance is always
       affected by lung burden.  How did you come to that conclusion?  I don't accept that

       as a universal truth, and don't believe that there are data confirming it. Could this
       statement and the modeling results in Figure 7A-9 be influenced by the fact that the
       clearance assumptions for rats were based on source data collected only under
       overload conditions by the chapter's own definitions (point raised above)? Even so,
       the Figure suggests that slowing of clearance is only affecting retained mass
       markedly at exposure concentrations that are extreme (over 10 mg/m3).

P 7A-37, Figure 7A-10:  The discussion of relative exposures to achieve equivalent retained
       lung burdens misses an important point.  The point is illustrated clearly by the figure.
       Yes, it is possible to expose rats to 60 |ig/m3 and hit at 6 months a normalized target
       human lung burden.  However,  you only hit that lung burden on the last exposure
       day.  If the health outcome of concern responds to the lung burden, rather than the
       recent dose, you can probably assume that the effect is not caused by the
       instantaneous lung burden, but rather by a lung burden that exists over some period
       of time.  Comparing the areas under the two curves illustrates clearly that the human
       and rat lung burden x time factors (similar to the concept of CxT) are very different.
       That is an important point that should be made in the section, not just the point that
       you can or cannot hit a target lung burden and how long it takes.  One would have to
       use a constantly,  or frequently, changing rat exposure concentration to mimic the
       time course of lung burden in humans. That could be done and would be an
       interesting strategy, but I'm not aware that it has been done.  This fundamental point
       is missing from the discussion.

P 7A-38, L 6-9: Same comment as above. Again, you are only hitting this burden at the end
       of the exposure period.

P 7A-39, Section 7A5.1.6:  The situation described has a  person exposed one  day to 100
       |ig/m3, PMio after a 10 year exposure to 64 |ig/m3. First, it's not clear why one
       would pick 64 |ig/m3 PMio as the  steady state 10-year level - seems an odd number.
       If one is thinking in a time series mode of course, the excess exposure that day is 46
       |ig/m3, not 100 |ig/m3.  Second, you have a rat exposed for 6 months to a
       resuspended dust that is 50% coarse and 50% accumulation.  May we presume that
       this definition goes back to your calling everything over 1.0 jim coarse?  Third we
       have a human long-term exposure that was 25% highly insoluble. Did the estimate
       of human cumulative lung burden consider deposition and clearance of only 25% of
       the mass, or 100%?  Was the "not highly insoluble" mass considered to  clear like the
       rest, considered to be invisible to the computation, or what? Did the lung burden on
       the last day of human exposure  only consider the  highly insoluble part (that is, did
       the other part instantaneously disappear somehow)? Was the resuspended dust
       considered all highly insoluble? Could we then conclude that it did not  have the
       same composition as the human exposure material (or perhaps any part of it)?  If so,
       why are we doing this?

The authors conclude from the example that it illustrates the "complexity of using a rat
       model to simulate effects of PM in the human lung".  First, the example said nothing
       at all about effects; it focused on dose. Effects are a different ballgame that includes

       a host of additional variables.  Second, and most importantly, the whole example is
       flawed because you weren't dealing with the same exposure materials in the two
       species to begin with. What the example illustrates most clearly is the complexity of
       thinking about the use of rat models, not the complexity of simulating doses.

P 7A-41, "Caveat" section:  The most important caveat is not mentioned.  The quantitative
       accuracy of the results is a minor concern; the mathematics are probably correct and
       the underlying assumptions of the model may not be far off. Ensuring that these
       tools are used to make biologically and environmentally relevant comparisons is the
       much greater concern.

P 7A-41, L 25-28:  This is the first time I've seen anyone propose that overloaded rats might
       be a useful model that represents a "sensitive" population!   Here, we find such rats
       defined as being "highly sensitive" to PM.  The proposition doesn't seem dangerous
       at this point - clearance is slowed in an overloaded rat, so it must be more at risk
       from the next inhaled particle. Later however, that gets translated into a potentially
       useful model of human sensitivity. If we follow that logic, then all of the past
       studies involving overloaded rats were just ahead of their time - they were studies
       using models of increased human susceptibility! EPA should have had the foresight
       to set a unit risk factor for diesel-related lung tumors from the rat data - those were
       studies of a sensitive population!

P 7A-45, L 6:  What do you mean "strictly speaking"? The analysis was flawed - period.

P 7A-45, L 9-10:  By assuming that all PMio is insoluble, this comparison is flawed from the
       get-go. In the last sentence, it's not clear whether you mean "comparable surface
       dose" of the total PM, or just the insoluble part.  This kind of apples & oranges
       comparison isn't useful

P 7A-45, L 26  - P 46, L 21. The presentation of biological results of studies and human-
       animal comparisons of responses is out of place in this appendix.  This is a dosimetry
       appendix. Given all the other variables in the comparison, whether or not you
       happen to get similar responses in the two species doesn't confirm or deny the
       similarity of dosing. The comparative responsiveness of humans and animals is
       another issue, and belongs in the body of the chapter. Moreover, on lines 46, 12-15,
       you decide to ignore PM clearance (which one really can't) and don't say whether
       the comparison is for the extract or the total PM.  The comparison is so cloudy as  to
       have no real value.

P 7A-56, L 11: Again, it is not scientifically sound to justify high level exposures to
       resuspended PM on the basis of simulating doses from multi-modal environmental

P 7A-57, L 1-6, and 57,L 22-58, L 8:  Again, presenting health data is out of place here. You
       need to address the similarity or difference in response between species in the body
       of the chapter. Comparative dosimetry is one piece of information needed to do that

          job, thus, the reason for this appendix. However, suggesting here that comparable
          biological responses confirm similar dosimetry is circular reasoning.

   P 7A-58, L 17-20: Overloaded rats may be models for overloaded humans, as long as the
          mechanisms of "sensitivity" are the same. The broad wording here suggesting that
          overloaded rats may be useful models for "decreased pulmonary defenses" is an
          overstatement.  I have a problem with promotion of this concept, unless you are
          limiting it to simulating humans in whom defenses are specifically impaired by PM
Comments on Chapter 9                                            Mauderly

General Comments

For the most part, this integrated synthesis is a reasonable start - but only a start.  The job isn't
done yet. At the Panel's recommendation, staff developed a review of (what the reader must
suppose are) the most important conclusions from foregoing chapters. This is a reasonable
approach, if the rest of the job (a true integrating synthesis) is done. On the other hand, the
chapter reviews could be left out altogether and only a truly "integrative synthesis" presented.
The chapter does not yet present a succinct, clear, synthesis across chapters that shows how the
different types of information can be related to frame the most important findings related to
standard setting.

It is not clear that the most important knowledge is brought forward.  It is clear that some
information is brought forward that seems to be of secondary or tertiary importance.  An
example is the speculation  (without confirmation in that particular study) about the chemical
species that might have been responsible for bacterial mutagenicity of collected engine emission
samples. The point is not that the speculation was wrong - it was a reasonable sidebar
speculation in a discussion section of a recent paper, and one that was a good bet on the basis of
previous confirmations by bio-directed fractionation two decades ago. The point is that it is a
mystery why that particular factoid is "cluttering up" what is supposed to be a focused synthesis
of the most important facts and conclusions bearing on the PM NAAQS. Indeed, the
aforementioned speculation had little to  do with even the most important conclusions of that
particular paper.

I would prefer that the chapter conclude with a section summarizing the most important points
supporting judgments about the PM NAAQS. If not the entire chapter, then as least a final
section, ought to make clear what the science says (or can't yet say) about the indicator, level,
averaging time, and statistical form  of PM standards. The authors need not fear usurping the role
of the Staff Paper. The section wouldn't draw conclusions regarding the components of the
standards, but it could point toward the knowledge having the greatest bearing on each of the

Specific Comments

P 9-7, Figure 9-1: This is a venerable figure, but it needs to be updated to reflect the terminology
       described in the text.  The term "ultrafine" is used in the text, but appears nowhere on the

P 9-8, L 11:  It is stated that progress has been made in measuring diesel particles. Isn't the
       progress in regard to measuring carbonaceous PM per se? Why just diesel particles?

P 9-9, L 13-14:  The wording implies that ultrafmes don't penetrate readily because they deposit
       to surfaces. Ultrafmes are defined in the chapter as equal to or less than lOOnm.  Figure
       9-3 suggests that some PM in this size range should have a high penetration factor.

P 9-17. L 19-21: We still have only a cursory understanding of the potential health importance
       of ultrafmes. We know enough to know that it's plausible that some of them may be
       important health concerns, but we don't know if this is true.  I'd strongly recommend
       waiting until we have more data produced y systematic laboratory studies before
       investing in an ultrafine sampling network. That's a huge investment on the basis of very
       skimpy data. The data at this time are suggestive, but largely anecdotal - few systematic
       studies have been done.

P 9-39, L 7:  "SO2" should be "SO2".

P 9-39, L 24: This is a silly sentence. How could it be possible that effects of PMi0 are not due
       to its constituent components? Could the effects of PMiobe due to something that isn't in

P 9-41, L 12: "Finds" should be "findings".

   P 9-42, L 29: Why emphasize that these are "powerful" techniques? The following line
   emphasizes their weakness. Why are these techniques called powerful, and others
   throughout the chapter are not? What's the point?

P 9-49, L 27-28: One of the main reasons that non-inhalation  dosing is used is that inhalation
       exposures typically require large samples. That ought to be mentioned.

P 9-53, Table 9-4:  This comparison is seriously flawed because the exposure materials were not
       the same.  You might play with such a comparison (with substantial caveats) in an earlier
       chapter, but the comparison is not sufficiently scientific to showcase in a summary

P 9-54, L 27: As noted in the preceding source chapter, this manner of presenting the
       information is misleading.  Without further knowledge, the reader takes the statement
       literally to mean the effects were observed at the lowest concentrations listed. This is not

P 9-55, L 10: It should read "inhalation exposures to ROFA".

P 9-56, L 28: See comment for 9-54, L 27.  Again, the presentation is misleading.

P9-57, L 1: Ditto.

P 9-57, L 11: It should be noted that the  exposures were by instillation. Otherwise, the reader
       will assume that they were by inhalation.

P 9-59, L 10: Isn't the "condensate" fraction part of PM?  If it's something else, make that

P 9-59, L 13: As noted in comments on chapter 7, if you are going to refer to the "coal smoke"
       results, you should put in a few words making clear to the reader that this is something
       different than the material to which people are exposed in the U.S.  The problem is that if
       you don't, most readers will assume that the conclusions pertain to coal-fired power plant
       emissions.  That would be misleading.

P 9-59, L 17-18: The wording suggests that the referenced study determined that the
       mutagenicity was caused by these chemical classes.  That is not true. The authors
       speculated in the discussion on the basis of differences between tester strains and +-S9
       that these classes might have been responsible. The speculation may  or may not have
       been correct, but it certainly isn't something sufficient to bring forward into a Criteria
       document summary.  In fact, the chemical classes responsible for the  mutagenicity of
       such samples was pretty well defined by bio-directed fractionation in the 1980s.

P 9-62, L 3-2: This statement in isolation doesn't help much. Give another sentence or two that
       explains the evidence (e.g., what you mean by "low" concentration).

P 9-72, L 13: Why include  this factoid?  Of what relevance to the integrated synthesis is it that
       you might be able to instill enough ROFA to kill a  rat?

P 9-73, L 11-17: This paragraph probably doesn't belong  in an integrated synthesis.  It isn't an
       update of real knowledge, it is just a recapitulation of an old hypothesis that hasn't been
       tested in any thorough way. We know that some PM can travel through the circulation.
       We still plausibly speculate, but still don't know, if such can affect heart function.

P 9-96, L 1: Being exposed to a higher concentration isn't a "susceptibility"  factor in any
       conventional sense.  Receiving a higher dose because a lung disorder caused enhanced
       deposition is a dose-related susceptibility factor, but just being exposed to high
       concentrations is not increased susceptibility.  This is a new use, or misuse, of the term.

P 9-99, L 22-27: The paragraph is not helpful unless the comparison is completed. You give
       relative  risks and exposures for the London fog, which isn't relevant to the synthesis, but
       you don't bother to give a risk factor or an  exposure level for a current scenario.

P 9-100, L 2: Are this total or non-accidental deaths?

P 9-107, L 22: Define "occult" deposition. Not all readers will have read the relevant preceding
       chapter (assuming that you defined it previously).

                              Dr. Roger O. McClellan
                      Roger O. McClellan, DVM, DABT, DABVT
                  Advisor, Toxicology and Human Health Risk Analysis
                            13701 Quaking Aspen Place N.E.
                             Albuquerque, NM 87111-7168
                                   Tel: 505-296-7083
                                     July 14, 2004

  Comments on Chapter 7, "Toxicology of Particulate Matter in Humans and Laboratory
                  Chapter 9, "Integrative Summary," June 2004 Draft
                                  Roger O. McClellan
       The following comments are provided on the two revised chapters.
A.     Chapter 7, Toxicology
       General  Comments
       1)     In general, this revised chapter represents a substantial improvement over the
earlier drafts. However, there are certain areas in which further improvements are needed to
provide a chapter suitable for closure.
       2)     A major shortcoming of the chapter is the lack of a contextual basis for the total
chapter and the appendices. Specifically, the authors do not provide a basis for the reader
understanding how the findings reported in the chapter will be useful in establishing the four
elements of a National Ambient Air Quality Standard (NAAQS), namely, (a) the indicator(s), (b)
averaging time(s), (c) numerical level(s), and (d) statistical form(s). I would argue that the
information contained in the chapter is probably of limited use for decision making on any of the
elements. However, one will never know unless the authors address the four elements. Perhaps
the strongest case could be made with regard to identifying "indicator(s)." Even with regard to
indicator(s) the chapter falls short.  Thus, the reader is left with a single broad conclusion after
wading through 350 pages of material - "Some kinds of particulate matter (PM) in some particle

sizes and with some chemical characteristics, when inhaled by humans at sufficiently high
concentrations and for sufficiently long periods of time may cause an increase in adverse health
outcomes."  As meager as this conclusion may appear if it were flushed out it could provide
information that would help in setting the PM NAAQS and in identifying future research needs
and the limits of toxicology in addressing these needs.
       3)     Section 7.1.1 - "Methodological Considerations" provide a useful, but limited as
noted above, context for this chapter. I especially appreciated the identification of a number of
important caveats or limitations that must be kept in mind in interpreting the findings reported in
the chapter.
       4)     One important caveat that needs discussion in Section 7.1.1 is the issue of
statistical considerations. The entire Criteria Document, and especially Chapters 7, 8, and 9,
need to be more forthright in describing the very weak signal for increased health effects
statistically associated with small increases in various PM metrics. The increases in health
effects that have been observed  are very small increases above the very low background rate of
adverse health outcomes that are statistically distributed (see Chapter 9, pages 95-100). Indeed,
it is apparent that statistically significant increases can only be found when the number of
observations are very large in terms of the human population studied and the "delta" of the
change in PM metric is large. Based on the above "facts" I would argue that most, if not all, of
the experimental studies lack the statistical power to describe changes that are relevant to human
populations at ambient levels of PM currently encountered in the U.S. The experimental designs
used are largely relevant to  studying deterministic outcomes rather than stochastic outcomes.
       5)     A careful review of the chapter reveals that even with flawed experimental
designs (that usually involve an inadequate number of subjects, a lack of multiple exposure/dose
levels, a lack of long-term observations and a lack of comparison between treatments) it is
apparent that PM is extraordinarily heterogenous as to its size distribution and chemical
composition and, moreover, there are marked  differences in the potency of different kinds of PM
for producing an increase in adverse health effects.  The chapter needs to clearly state this
conclusion which has profound  implications for setting and implementing a NAAQS for various
PM indicators that will yield positive health benefits.  Indeed, the only conclusion one can draw
is that the current approach  using an indicator defined by size, without regard to chemical

composition, is likely flawed and results in substantial societal costs and may not always have
associated health benefits.
       6)     Appendix 7A - "Rat to Human Dose Extrapolation" while improved over the
earlier version is seriously flawed in "over-selling" a product and attempting to justify much of
the animal research reported in the body of the chapter.  The chapter is an interesting exercise in
modeling. However, models are only that - models of reality - and need rigorous validation. I
would characterize the Appendix as providing a sound basis for model validation. The chapter
would be substantially improved by clearly identifying the very few observations that provide
the basis for extensive mathematical extrapolations. I am very aware of how limited the
biological inputs are since many of them originated in research programs I directed.
              All of the calculated values are presented with a high degree of precision (two or
three significant figures) and only rarely with an indication of uncertainty in the results leading
the unwary reader to perhaps over-interpret the results presented. It should be recognized that
the results are based on model calculations and in the absence of further validation should be
used and interpreted with caution. I suggest that all of the table and figure captions be reviewed
and the words "estimated" or "calculated" placed in front of any estimated or calculated
              The Appendix would be substantially strengthened by including "measured"
values for comparison with "calculated" values. For example, "measured" lung burden data
from the study of Ronald Wolff et al (1987), [Alterations in Particle Accumulation and
Clearance in Lungs of Rats Chronically Exposed to Diesel Exhaust, Fund. Appl. Toxicol. 9:  154-
166] of rats clinically exposed to diesel exhaust could be plotted in Figure 7A-8.
       Specific Comments
       Pg 7-69, lines 16-18: This sentence does not need qualification - i.e. "the relatively
recent new studies" - it should simply state - "There is little evidence for acute or chronic
exposures to aqueous acid aerosols contributing to acute respiratory effects on chronic lung
pathology, except a much higher than current ambient levels."
       Pg 7-69, lines 23-25: This sentence is  a gross over-statement that needs qualification,
i.e., non-cancer effects are observed with high levels of exposure.  For example, see pg 7-70, line
       Pg 7-74, lines 12-17: This conclusion  is not supported by the text.

       Pg 7-102, lines 30-31:  This important conclusion stands in sharp contrast to other
comparisons made elsewhere.
       Pg 7-126, lines 5-6:  This short but important conclusion needs to be linked to the
statistical limitations of toxicological studies I noted earlier.
       Pg 7-143, lines 18-22:  The fable of particles serving as effective carriers for
formaldehyde is simply that - a fable. See paper by Simon J. Rothenberg and colleagues who
demonstrated, based on the physical chemical characteristics of formaldehyde, that the particle
associated "dose" is insignificant compared to the vapor associated "dose." . [Simon J.
Rothenberg, Paul A. Nagy, John A. Pickrell and Charles H. Hobbs, "Surface Area, Adsorption
and Desorption Studies on Indoor Dust Particles," Am. J. Hyg. Assoc. J. 50(1):  15-23, 1989].
Yes, formaldehyde absorbs to particles.  However, the quantity of formaldehyde delivered to the
pulmonary compartment by deposited particles was calculated to be less than 1% of the quantity
delivered to the upper respiratory tract.
       Pg 7-152, line 11: Remove the term "greatly," the evidence for modification of effects
indicates it may occur but it is hardly supportive of "greatly modify."
       Pg 7-182, lines 23-30:  In my opinion, the authors over-interpret the evidence.  Yes, some
ambient air particulate matter is mutagenic. However, the levels of mutagenicity measured are
remarkably low.
       Pg 7-192, lines 1-12:  The discussion of the work of Driscoll et al (1997) needs to
indicate that the "oxidative damage" of carbon black particles  appeared to be a threshold
exposure dose-response phenomena.
       Pg 7-196:  The experimental design of the Strandell et  al (1994) and Seagrave et al
(2002) studies should be more fully described. As I recall Strandell et al studied only particulate
emissions, Seagrave et al studied both particle and vapor phase emissions.   The  same
consideration applies to the Rannug (1983) study which, as I recall, focused on particulate
emissions.  It is not possible to compare different emission sources without considering total

       Pg 7-206: This page is loaded with speculation. The discussion of the paper of
Kleinman etal (1998) inappropriately implies a linkage between PM-hypoxygenation-angina
and further adverse cardiac effects. This is totally speculation. If the authors wish to discuss CO
and hypoxygenation and times to onset of angina they should reference the results of the classic
HEI multi-center CO study. The next paragraph describing a study in which blood oxygen levels
were never measured is total speculation and leads one to question the scientific objectivity of
the authors on other matters.
       Pg 7-210, lines 20-24: What a mouthful!  Why not say - The Pope et al (2002) analysis
of the American Cancer Society longer-term database provides evidence for chronic ambient PM
exposure being associated with increased risk of lung cancer.  The Pope et al study is certainly
not strong evidence. Indeed, alternative interpretations related to the weakness of smoke history
data leads to legitimate questions as to whether there is a PM-lung cancer effect.
       Pg 7-212, lines 1-3: I question whether the qualitative analyses of mutagenicity for
different fuel types is warranted without  substantially more details as to what is being ranked
here or elsewhere.  Indeed, I view a ranking based on type of fuel as being very misleading.
These rankings are more a ranking of the inefficiencies of the combustion processes studied in
specific paniculate experiments rather than inherent characteristics of the fuels.  For example, a
majority of the references to coal combustion are based on studies of coal being used for cooking
and heating in poorly ventilated rooms in China.  Yes, this is coal combustion but it is hardly
representative of a modern coal-fired electrical power generating plant.
       Pg 7-216, lines 5-18: This paragraph overstates the strength of the evidence for diesel
PM exacerbating allergic responses to inhaled antigens. The quantities (dose) of diesel PM or
extracts studied, the sources of the test material and the mode of administration all raise serious
questions as to the relevance of the findings to paniculate matter exhaust from modern diesel
engines.  The same statement applies to pg 7-221, lines 20-25.

B.     Chapter 9, Integrative Summary
       General Comments
       1)     In general, the draft chapter provides an adequate summary of key "air quality
criteria" for establishing the National Ambient Air Quality Standard (NAAQS) for Particulate
Matter.  However, as detailed below it is my professional opinion the chapter can be improved
with some changes and additions are shortening.
       2)     As I have previously noted, I think it is important for the criteria document, as a
whole and certainly the summary chapter, to clearly indicate the relevance of the science being
reviewed to the four specific elements of a NAAQS; namely, (a) the indicator(s), (b) averaging
time, (c) numerical level, and (d) statistical  form.  The organization and content of the present
chapter does not adequately review the science relevant to decisions on these four elements.
Thus, by default, the Agency (and CASAC) is leaving critical consideration of all  four elements
to the more policy-oriented Staff Position Paper.
       3)     From an organizational standpoint it is my opinion that the chapter would be
improved if very early in the chapter there was a section summarizing knowledge  on historical
and current ambient concentrations of the several PM indicators and summary information on
chemical composition.  This section would  provide useful background  information that would
provide a contextual setting for the rest of the chapter. This section could be followed by the
background health data now presented on pages 9-95 to 9-100.
       4)     Building on the information  presented above the chapter would be improved by
including a brief section describing how the epidemiological evidence is based on the use of
relative risk models and the critical importance of the baseline prevalence of cardio-respiratory
disease in determining estimated absolute risk associated with air pollution.
       5)     It is remarkable that the chapter could be written with only a single sentence
acknowledging the role of cigarette smoking as the single most substantial risk factor
determining the "susceptible" population for PM-associated excess disease. Pg 9-19, lines 17-
19: "The etiology of most air pollution related health outcomes is highly multifactorial and the
important ambient air pollution exposure on these outcomes may be small in comparison to that
of other etiological factors (e.g., smoking).  This could well be the lead sentence in the Summary

       6)     In my opinion, the chapter would be improved by more directly addressing the
issue of establishing PM size, mass-based indicator(s) without consideration of chemical
composition. These mass-based indicators are selected by default, they are simply the best we
have in the absence of science to the contrary. In many places, the summary notes that PM is
heterogenous with regard to chemical composition and there is substantial evidence indicating
that PM is equally heterogenous in potency.  Unfortunately, our ability to link chemical
composition and toxic potency is scientifically inadequate.
       7)     It would be appropriate for the chapter to note that as more PM speciation data
become available and is used in epidemiological studies, it will be possible to evaluate the
relative health protection provided by mass-based versus chemical based indicators. Likewise,
as more continuous monitoring air quality data become available, it will be possible to evaluate
the relative health protection provided by standards for indicators with different averaging times
and associated statistical forms.
       8)     I think the chapter could be substantially reduced in length without any loss in
scientific content. The chapter would benefit from "heavy-handed"  scientific editing by a single
editor to provide a more consistent presentation.  The current uneven nature of the presentation is
apparent in the summaries that range in length from a short paragraph to two pages.
       9)     In performing the final editing, it will be important to strive for a more balanced
scientific presentation that even-handedly describes the evidence for PM-associated health
effects and the associated strengths and weaknesses. It is important that the science-based
criteria document avoids a tone that advocates or justifies a particular standard including past or
anticipated actions of the Agency.
       Specific Comments
       Eliminate Figure 9.2
       Pg 9-11, Discussion of Figure 9.3:  Emphasize that "infiltration" is strongly influenced by
"climate" and, equally important, the nature of building ventilation.
       Pg 9-13, Section, Figure 9-4: Discuss the extent to which this figure is based on
modeling of "monodisperse" particle sizes.  It should note that the real world is "polydisperse"
and,  thus, for any real world situations there is a translational issue.
       Pg 9-23, line 14: This should be PMi0-2.5 rather than PM2.5.

       Pg 9-26, line 12: If the Laden et al (2000) paper is to be cited, it is appropriate to
reference a paper that provides an alternative analysis with different conclusions. (Grahame, T.
and Hidy, G., "Using Factor Analysis to Attribute Health Impacts to Particulate Pollution
Sources," Inhal. Tox. 16 (Suppl. 1): 143-152, 2004).
       Pg 9-38, lines 16-20: I think this statement is much stronger than warranted based on the
strength of the evidence and the challenges faced in evaluating exposure-response relationships.
This represents one of the key conclusions in the CD and, thus, it must be accurate and
appropriate caveats added as to the strength of the evidence.
       Pg 9-49 through 9-53:   This section on dosimetric considerations, including consideration
of Appendix 7A, should be substantially shortened. A shortened version should emphasize the
extent to which the mathematical dosimetric models have only been validated to a very limited
extent. The results of the modeling exercises should be used with more caution than suggested
by the material presented on these four pages.
       Pg 9-95 to 9-100:  This excellent section on baseline health statistics should be moved
forward in the chapter as recommended earlier.  It would be appropriate to indicate, and perhaps
illustrate, the significant regional differences in the various health indices and some of major
underlying demographic factors.  In this section, it would be appropriate to repeat the statement
made on pg 9-19 noting the important role of cigarette smoking as a risk factor in several of the
key cardio-respiratory diseases of concern for being impacted by PM.

                              Dr. Giinter Oberdorster
        This chapter is quite improved compared to the previous draft, the reorganization looks
good and new sections fit well into this  chapter.  I have a few comments that can easily be
considered and should not — in my view — delay coming to closure on this chapter.

       Section 7.1.1 Methodological Considerations is a nice addition as an introduction to this
chapter.  These are very good comments and thoughts on the concepts of PM toxicity.  I would
suggest on page 7-3 to include the concept of hypothesis-generating studies — which could be
many of the high dose studies to identify mechanisms — and contrast these with studies of
relevant low doses to test a specific hypothesis. This will provide some justification for many of
the high dose studies that have been generated.

        Also, I suggest to address the notion that not every small  or  slight response — e.g.,
inflammatory cell influx — to be considered adverse, rather one should differentiate between a
normal adaptive or physiological response as opposed to toxic response.

        Page 7-8, line 11:  I suggest as an introduction to the figure to mention  briefly the two
pathways outlined in the figure: These mechanisms involve either the ANS or direct effects on
the endothelium.
                    Lines 16/17: What is the heart oxygen carrying capacity of the blood? Is
meant here: Decreasing blood flow?

       Page 7-11, line 2:  How is "blood chemistry" related to coagulation? Perhaps it is better
to state: "effects on endothelial function".

       Page 7-14, line 9:  The MCT model is described here as a model of COPD in humans.
However, this is not a chronic but a very acute injury and it is questionable as to whether it
resembles COPD.
                           Lines 24-28: The mostly very high doses of the studies in tables 1
and 2 should be mentioned here.

       Page 7-15, Table 1, first entry: An estimated total dose is given here, is it deposited or
inhaled dose? This is not made clear in the text or here.

       Page 7-33, line 21:  The elevated heart rate was found in which exposure group, the
highest only?

       Page 7-35, line 7:  Additional information to include here and which would be useful is a
study by Terashima et al. (2001; BAL induces an increase in peripheral blood neutrophils and
cytokine levels in healthy volunteers and patients with pneumonia, Chest 119:  1724-1729)
showing that simply BAL performing in humans also induces acute phase responses and release
of bone marrow stimulation, whereas bronchoscopy alone does not.

                    Line 26: Add that carboxylate particles have a negative charge and amine
coated particles have a positive charge.

       Page 7-36, line 9: Emphasize also that all these studies were performed in healthy

       Page 7-51, Table 6:  Since these are all intratracheal instillation studies, there would be
no need to have the column Exposure Technique; also, change "concentration" to "dose" and
delete in the other column "Exposure Duration", which will be only a second or so and is not
needed here for instillations.

       Page 7-55, Table 7:  Add to the column heading "Concentration/Dose".

       Page 7-61, line 14: This summarizes data of Table 7-6, however, where are studies
summarized which are listed in Table 7-7?  They were not described in this section.

       Page 7-62, lines 29/30: The statement that iron may play little role in adverse effects of
PM is not justified based alone on the results of lung function and lung permeability
measurements; certainly there were inflammatory changes in this study by Ohio et al. (2001).

       Page 7-83, lines 16/18: It is not clear what the dose of 15 jig means, is it inhaled or
estimated deposition? Also, what is the particle size of LPS? This would be useful information
to give the reader some understanding of how much of the inhaled LPS was  deposited in the
lower respiratory tract.

       Page 7-85, line 13: Add at the end of the line "at higher concentrations or doses".

       Page 7-107, line 6: Replace "these" with "transition".

       Page 7-108, line 4:  Rather than giving the rpm for the centrifuge, what were the g-
                    Line 11 to
       Page 7-109, line 29: Possible mechanisms/hypotheses of metal induced effects are
described here .  When discussing these mechanisms, it needs to be pointed out that the studies
were done at high doses and that the mechanisms are not necessarily the same as those being
induced at low doses. (This concept of "the dose making the mechanism" was already well
addressed earlier in this document).

       Page 7-113, line 18: A statement is made here that inhaled particles are trapped in the
epithelial lining of the nasal and tracheal airways. I think what is meant is that particles are
deposited, or else what study shows that the particles are "trapped"?

       Page 7-119, line 28: Newer studies are  referred to here which should demonstrate a
plausible neurogenic basis for PM inflammation.  I am not sure how plausible these studies are,
given that large doses were used in  an in vitro assay only. At this point, this is a hypothesis that
requires confirmatory testing.

       Page 7-121, line 10: The study by Gilmour et al. (1996) was with PMio, not ultrafme

       Page 7-124, lines 22-27: it is well known that in addition to particle surface area,
chemical composition impacts on the cellular responses to PM.  However, I am not sure that the
study by Schluter demonstrates this: Was the size of these differently coated silica particles the
same for all coatings? Were these fine particles, ultrafme particles? Was the uptake by
macrophages the same for the differently coated particles? Not knowing the answers, one has to
be careful with the  statement as presented here.

       Page 7-134, line 6:  It should also be pointed out at the end of this paragraph that
mortality is only a very crude endpoint in these genetic studies.

       Page 7-153: This new section on quantitative comparisons of experimental PM has a
very good introductory chapter, although several of the statements in the later sections are
somewhat superficial and need to be rephrased, or explained below:

       Page 7-166, lines 16-20: An effort is made to compare tissue doses and responses
between rats and humans exposed to PM:
                    Line 17:  refers to the surface area dose, please replace the term "tissue
                    Line 19:  suggests that even healthy humans are notably more susceptible
to the inflammatory effects of CAPs than are rats. However, this cannot necessarily be derived
from these studies:  Rats in this comparison were exposed for 3  days in a row, humans only once,
and observed effects are the basis for this conclusion. However, there are data showing
(including our own with the highly toxic PTFE particles) that 3 days in a row exposure can cause
adaptive responses: Rats in our PTFE studies became less or even unresponsive after 3 days of
exposure as opposed to a one-time exposure only.  Thus, one has to be very careful comparing
different exposure scenarios, i.e.,  repeated exposures vs. a one-time only exposure; in addition,
species differences have to be considered in such adaptive responses, so I think that this
comparison and respective  conclusion are not valid.

       Page 7-168, line 14: The  document states: "These in vitro studies of human AM may be
compared to three available studies that investigated animal AM responses	" I don't think
that this can be done so easily, comparing primary human cells to a mouse  cell line, how can that
be compared? Even when comparing primary human AM to primary mouse AM, it will be
difficult: Culture conditions, even using the same culture medium, may be optimal for one or the
other species, but not for both, so there are many caveats that have to be considered when
making  such comparisons of in vitro studies with cells and cell lines from different species.

       Page 7-171, line 5:1 suggest to replace "human AM are at least as sensitive" with
"human AM may be at least as sensitive".
                           I suggest also at the end of this section, line 14, to include a
general concluding sentence pointing out that for making comparative statements it is also
important 1) to compare the same cell types, include a control particle and evaluate relative

differences of human and rodent primary cells to this control particle; 2) compare the same types
of dusts, rather than one with significantly teachable metals vs. another devoid of metals.
Overall, don't mix PM types and cell types (primary/secondary).

       Page 7-185, line 3: Typo, m3.

       Page 7-192. lines 8/9:  In this study by Driscoll et al. (1997), it should be added that the
neutrophils in the BAL are the source of ROS, which in turn result in secondary genotoxicity via
DNA damage; this is quite different from primary genotoxicity. It needs to be also made clear
that the results of this study were definitely a high dose effect and that these effects are not seen
at low doses.

       Page 7-202, lines 1-3:  The faster clearance of PM the lower respiratory tract in rats vs.
humans is most important for long-term exposures and effects; for acute effects differences in
deposition between rats and humans rather than clearance differences are more important.

       Page 7-204, line 25: The statement in this line suggests that in these studies by Veronesi
et al. the proton cloud was measured; however, I assume that this is just a mechanism suggested
as a likely scenario by the authors.

       Page 7-206, lines 23/24: I suggest to again add here that the instilled ROFA involved
high doses.

       Page 7-211, line 22: As mentioned before, it should be added here that the mutations
were due to ROS released from inflammatory cells, resulting in secondary genotoxicity, not
primary genotoxicity which is the case for some PAH.

       Page 7-216, lines 12/13: It also needs to be noted that adherence of allergen-laden pollen
as shown here for DPM has not yet been investigated with other PM; in other studies which were
summarized in this criteria document (Steerenberg study) it was shown that DPM was least
immunogenic when contrasted with other PM.

       Page 7-220, lines 11/12: The MCT model has been repeatedly presented in this criteria
document; as pointed out before, this is very likely not a relevant model for human conditions,
and I suggest it should not be presented here again as the model for a compromised host. Rather,
the difficulty  should be pointed out to establish relevant models for human compromised
conditions in  animals, and that there is clearly a need to search for new and better models to
reflect the underlying human pathophysiology.

       Page 7-221, lines 22/23: Again, this last sentence points to DPM as one particle that is
effective in exacerbating allergic asthma responses; however, so do other particles if they are
tested in the same study together with DPM; it would, therefore, be helpful to point out the
shortcomings in most of the DPM by studies not having included a comparison particle.

                              REVIEW OF APPENDIX 7A

       Overall, this chapter provides good and thoughtful discussions; it would be desirable to
include some concluding comments after individual sections similar to what is  done in the
summary.  Specifically,  it would be useful to emphasize more the complexity of dosimetric
extrapolations,  stressing  that  this is highly dependent on PM parameters, exposure  scenarios,
breathing and activity patterns of different  species  and — not yet achievable by models —
expected differences between responses of a compromised host vs. healthy host.  The summary
does a nice job in this  regard, and in between conclusions after individual sections would
strengthen this.

       Page 3, line 22 and subsequently throughout the chapter and figure legends:  I suggest not
to use the term "highly insoluble" particles, but rather call them "poorly soluble" particles; I have
used this term "highly insoluble particles" myself in earlier publications, but I think the general
consensus is that these should more appropriately be called "poorly soluble particles" because
there is no particle which is highly insoluble (perhaps  iridium is the most insoluble particle so far

       Page 4, line 11: The symbols f and Vt need to  be explained, i.e., frequency and tidal
              Lines 26-29:  These sentences are a bit confusing:  Normalization and
uncertainty factors are introduced here together, they seem to blend into each other without
distinction, and it needs to be clarified.  Also, in line 28, the term "acceptable" human dose is not
clear, probably what is meant is "extrapolated" human dose?  This continues on line 9 of the next
page where the acceptable human dose is defined as 1/3 of the no-effect level for the animals: so
the acceptable dose is a no-effect level?

       Page?, line 31:  It is stated that the alveolar surface area is not included in the MPPD
model, however, the version that has been distributed allows to model alveolar surface area as
well (and is also used later in this chapter as a result).

       Page 9, line 9: Alveolar rate constants for the rat equivalent to retention halftimes of 100
days, 323 days and 835 days are given here to be used in the retention modeling exercises.
These values are from a recent article by Bermudez et a/., however, there are numerous earlier
studies which show that normal retention halftimes in  rats are between 60-80 days, in fact in the
article by Bermudez et a/., a similar study with TiO2 by Warheit et al. is mentioned which gave
retention halftimes of 68 (normal, un-impaired), 110 and 330 days for the  same TiO2 exposures.
These values are more in  line with the normal rat clearance values observed in earlier studies
using radioactive test materials. Use of the loer T 1/2  may change the predictive modeling
results, and it should be indicated here that the modeling exercises are performed only to show
principles of extrapolation models and that actual results will vary depending on model inputs, as
       For humans there  was a three-phase alveolar clearance assumed, which is well
demonstrated by earlier studies by Bailey et a/., was that also assumed to be the case in rats?
Kreyling and Scheuch report a time-dependent clearance rate for the rat as well.

       Page 9, line 31: As in my earlier comments, I suggest again that the term "clearance
halftime" should be changed to "retention halftime" throughout this section.

       Page 10, upper para.: For the human clearance rates in the alveolar region, was there also
assumed to be a load dependent retardation as is the case in rats?  Although there is no direct
evidence in humans for this phenomenon due to the lack of measurements, there is some
evidence from coal miners (Stober; Freedman) that such retardation with increasing lung load
does happen. Obviously, such high lung burdens will never be achieved at levels of PM in the
ambient air.

       Page 11, line 19: Add the term "(mass)" after "|ig" and the term "(surface area)" after
       Page 14, footnote B of table legend: The reported data for the resuspended PM
distributions were done with the use of a cyclone? Give references.

       Page 18, lines 1-5:  Summary data of my old 1988 retention review are given here;
although the general principles of differences among species in terms of particle retention are
still the same, I suggest to use newer data, specifically a review by Kreyling and Scheuch (2000,
In:  Particle-Lung Interactions, Chapter 7: Clearance of particles deposited in the lungs, pgs.
323-376) as the most up-to-date  reference. Data in this summary also show that rats as well as
other species show a change in the alveolar clearance rate over time,  similar to what has been
reported by Bailey et al. in the 80' s for humans.

       Page 22, line 27: Replace "mass" with "dose" in both places.

       Page 24, Table 7a-3 : Another parameter useful to compare and add here would be the
deposited dose per unit surface area of the alveolar region.

       Page 26, Table 7a-6: Add a footnote to the slow clearance for humans and add this  as b
on the table indicating that the alveolar clearance rate is a function of time. Also, stating in the
PM size distribution column for  humans "exposed to all three atmospheric modes" is not quite
correct in case of exposures to CAPs.

       Page 28, Table 7a-7: This is an interesting table showing the large variations of EqER
and its dependence on selected parameters.  For clarification I suggest to add to the title, first line
"deposited dose after" before "a  6-hr,  exposure

       Page 31, line 2: With respect to surface area of particle aggregates, one would not
require very high doses since the aggregated surface area may still  be the  same as that of
singlets; what changes in that case is the deposition fraction and the site of deposition.

       Page 31, line 13 : The large ranges of EqER values require  a concluding comment, rather
than just reporting the model values. For example, that this shows how questionable such
comparison may  be and that there are still a number of uncertainties due to the complexity of the

       Page 36, lines 1-6:  Under realistic ambient exposures, rats should not enter the overload
state, thus, there is no need to consider such scenario. Neither does it for humans as is pointed
out here, however, the human model does not include a change of clearance rate with time. That
the rat will not reach overload state under realistic exposures is also shown in Figure 7-9, so
there really is no need to dwell on the overload situation, but what should be indicated is that in
rats,  like in humans, the clearance rate changes (becomes slower) with time which is not
modeled in these exercises.
       Figure 7-9, the labeling on the ordinate is not clear, retained mass as a function of
deposited mass?

       Page 41, section on "caveats": It should also be emphasized here that only the dose is
considered in these modeling exercises nothing else; other modifiers are:  cellular responses; the
fact that affected humans are compromised; that there are changes in deposition, retention and
response in compromised hosts; that this dosimetry modeling is for healthy rat to healthy human
and vice versa; for compromised organisms many parameters will change.
             In line 11, add at the end of the sentence something about the usefulness  of these
modeling exercises for study design and avoiding nonsensical rat exposure studies as was
discussed on the previous page (i.e., exposure of rats for 6 months to resuspended dust).

       Page 43, line 11:  It would be helpful to also indicate how much PM material was needed
to achieve 500 jig of the extract.

       Page 45, line 22:  The differences between a subchronic exposure and the highly acute
manner of a delivery by instillation is very important and should be emphasized.  Obviously,  the
dose rate is very different, and the dose rate is most important for acute effects. This issue of
dose rate should be discussed here, the concept and principle are already addressed in this and
the following page.

       Page 48, table 7-12:  In this table and in the two tables on the next page, the use  of the
term "equivalent" is not quite correct based on how this is defined in this chapter; it should be
rather "predicted".

       Page 50, line 7:  The definition of dusts as 'nuisance" or "inert" should be avoided rather
they should be called "low toxicity dusts".
             Line  21: Reference is made to a discrepancy between the volume hypothesis and
surface area hypothesis in overload and other situations: However, this is not really a
discrepancy, but it shows our continuing understanding and developing of our understanding
over time; with the increasing awareness of ultrafme particles  it turned out that the volume
concept does no longer hold anymore, but the surface area concept does for poorly soluble
particles of low cytotoxicity. That is, surface area appears to be a more appropriate dosemetric
to model effects caused by different loads of particles of sizes going down to the ultrafme size
range, which case is not well modeled by the volume overload concept.

       Page 53, line 24:  Reference is made here to the importance of particle size on alveolar
macrophage-mediated clearance, justifying using MMADs only between 1 and 4 jim. However,

this covers not the whole problem, aggregated ultrafine particles can also have a MMAD of 1
|im, yet still they elicit responses due to their unchanged large surface area rather than their
volume (see studies with instilled and inhaled aggregated ultrafine vs. fine TiO2).

       Page 55, lines 2 and 3: The statement that high concentrations given to rats may also
better simulate high deposition at "hotspots", or in active portions, of the diseased human lungs
does not necessarily hold; there are no data to suggest this; moreover, if higher concentrations
are used for rats, this will create even greater hotspots in that species, not just the normal
hotspots seen anyway.

       Page 58, lines 19/20:  Again, it is suggested here  that there may be occasions where some
extent of overload could be needed to mimic certain human conditions:  One has to be careful
with this suggestion:  Which decreased pulmonary defenses in humans should be modeled by a
particle overloaded rat? If there is a specific case state it here or else delete this sentence.

       Page 59, Conclusions:
       • - First bulletpoint, I suggest to add in line 2 "in  certain conditions" after "would be
       • -  Second bulletpoint, line 3, this sentence needs to be turned around and read like this:
Given that rats clear PM much faster than humans, the MPPD model results show that much
higher	 Also, in the last line of this bulletpoint - replace "highly insoluble" with "poorly
soluble" particles.
       • - Bulletpoint 3, line 4:  Resuspended PM does contain the smaller particles, but they are
aggregated onto the larger ones.
       • - Bulletpoint 4, line 5, 4th and 3rd lines from bottom: Insert the word "healthy" before
"rats clear PM	" and before "humans".

                                Dr. Robert D. Rowe
To:          Fred Butterfield
From:       Bob Rowe, Stratus Consulting Inc.
Date:        7/9/2004
Subject:     Review comments on PM CD Chapter 9, June 2004 Draft

Chapter is much improved and close to closure, subject to clean-up. Detailed comments follow.

Section 9.3.1 Visibility

The text places emphasis on the term "public values," which conjures up a focus on monetary
valuation (see page 9-102 line 25, page 9-105 lines 15-23 and page 9-106 line 3). A broader
phrase, such as "public perceptions, attitudes, and values," may be preferred to blend the relevant
information and messages from the newer studies (e.g., the cited Denver/Phoenix studies that are
about perceptions and attitudes about what constitutes adverse conditions) and the older
economic studies (where monetary valuation is emphasized in as an indicator of adversity).

The CD does little in Chapter 4 to meaningfully combine information from the old studies and
the newer studies identified to address relevant questions such as "how much or what
characteristics of impairment are adverse?" and "how adverse is it?". This limitation carries on in
Chapter 9 on page 105. On lines 15 through 23 the newer studies are set aside as something
different than the older studies, but being only two newer studies the text identifies there may be
too few to base policy on. However, both sets of studies suggest, through different metrics (with
different measurement issues), that current levels of anthropogenic caused visibility impairment
are viewed as adverse by a share of, if not a majority of, the public.   In the two newer studies,
adversity ratings clearly indicate that levels that are experienced routinely in the two study cities
and elsewhere are adverse. In the economic valuation studies the estimated dollar values can be
considered an indicator that current levels of impairment are adverse, even allowing for
uncertainty in the precise dollar values assigned.

Given the Criteria Document's focus on the limited new literature, minor wording improvements
may be sufficient for closure. However, for the Staff Paper EPA should consider more serious
efforts to see if and how all the literature can be used to address key questions relevant to
selecting a secondary standards.

Page 9-102 Line 24. For consistency, consider replacing "affects" with "impairs."

Section 9.3.2: Vegetation and Ecosystems.

This section is better than prior iterations; particularly the introduction and summary, and is
much closer to acceptance.  The section can use some tightening and clean-up.

>  In places the text tends to provide discussion that may be longer than necessary for the
   synthesis of key points for evaluating potential standards. Page 9-108  seems to retreat to
   laying groundwork rather than staying focused on key implications for impacts to
   individuals, populations and the ecosystem. Page 9-110 lines 3-7 are unnecessary.

>  While many of the key points from Chapter 4 relevant to the state of the science and its
   implications to standard setting are carried forward here, I wondered about the omission of
   others,  including:

   -   Experimental applications of PM constituents to foliage typically elicit little response at
       the more common ambient concentrations (June 2003 draft page 4-60 line 24);

       Although forest ecosystems other than the high-level spruce-fir forests are not currently
       manifesting symptoms of injury directly attributable to acid deposition, less sensitive
       forests throughout the United States are experiencing gradual losses of base cation
       nutrients, which in many cases is expected to reduce the quality of forest nutrition over
       the long term (page 4-115).  Acidic deposition is having a significant affect on nutrient
       cycling in most of the forest ecosystems studied in the US project (page 4-136).

>  Page 9-106, line 24. Sources include mining emissions beyond just iron and lead smelting.

>  Page 9-106, lines 25-27 is it correct to say "these effects tend to be limited in scope" as the
   reason for dismissal here of these situations or is it that these cases are generally addressed
   through regulatory strategies other than the PM SNAAQS.

>  Page 9-114, line 27. The line is confusing "nitrogen or acidic deposition its role..."

Section 9.3.3 Climate
>  Page 9-116, line 26. Although "degradation" is fine, "impairment" may be preferred for
   consistency with the visibility section terminology.

Section 9.3.4 Material
>  Page 9-119, second to last sentence  can be a bit broader. "Available data indicates that
   airborne particles can result in increases in the frequency of cleaning,  maintenance, or
   replacement of exposed surfaces and materials, as well as reduced usefulness and enjoyment
   of injured materials (as is the case with stone monuments or dirty buildings).

>  Page 9-110, last sentence. The focus on perception  and perception thresholds seems limiting.
   Attempts have been made to both address the perception threshold of  soiling and the welfare
   consequences. Also, while there has been little new on traditional soiling costs, some new
   work on public values with respect to cultural monuments/buildings has occurred and
   continues to indicate that ongoing degradation of cultural sites is viewed by the public as
   adverse. (S. Navrud and R.C. Ready (eds.) 2002. Valuing Cultural Resources.  Edward  Elgar,
   Cheltenham, UK, ISBN 1-84064-079-0.)

                               Dr. Jonathan M. Samet
                              Jonathan Samet, M.D., M.S.
                                Professor and Chairman
           Jacob I and Irene B. Fabrikant Professor in Health, Risk, and Society
                    Johns Hopkins Bloomberg School of Public Health
Comments on Chapter 8: Epidemiology of Human Health Effects Associated with Ambient
Particulate Matter


This revised chapter continues to improve; of course, given its length and scope, additional
points for revision can always be identified.  However, I find that the chapter has been improved
sufficiently to warrant its being approved as final, with some minor modifications. The chapter
appropriately sets out how the evidence was gathered and evaluated, and provides clear
summaries of the relevant data.  Consideration might be given to incorporating some of the
material in section 8.5 into Chapter 9, as this material reflects an integration of the
epidemiological findings with other information, largely related to causal inference. I note that
the chapter begins with a description of the "Hill" criteria for causality, but these criteria are in
fact not uniformly applied, either in this chapter or in Chapter 9.

The initial text, offering a primer in epidemiology, has sharpened.  I find some of its statements
too sweeping and have highlighted these below in my specific comments. The chapter offers the
view that "correct" models can never be identified and that there is always a potential for
residual confounding.  This proposition is hardly unique to studies of particulate matter and
health, and has not been a barrier to the use of observational evidence in other contexts.
Confounding is of greater concern when effect sizes are small, as in this case. However, the
sweeping generalizations need to be toned down.


Page 8-10, Lines 3-15: this paragraph is not well written and might benefit from inclusion of a

Page 8-12, Lines 6-7: this statement is far too general.  What is meant by "none of them being
completely satisfactory"?

Page 8-13, Lines 8-17: most important, but not mentioned, is the sensitivity of findings to the
details of model specification. The finding of an association that is robust should be taken as
evidence that substantial confounding "or model misspecification" is not strongly affecting the
effect estimate.

Page 8-327, Lines 27-31: why do the authors of the Criteria Document look for "consensus"?
Again, sensitivity of findings to alternative specifications needs emphasis.

Page 8-327, Lines 39-40: again, a similar sweeping statement.
Chapter Nine: Integrative Synthesis


The Environmental Protection Agency is deserving of congratulations for taking a first step
towards developing a satisfactory "integrative synthesis" of the Criteria Document.  In
comparison with the prior draft, the present chapter is more than simply a culling of the
summaries of the individual chapters. There is an attempt to draw the evidence together around
a set of key uncertainties in any further evolution of the NAAQS.  Nonetheless, I have some
major concerns with regard to Chapter 9, which should still undergo substantial revision:

•  The chapter remains too long, comprising 119 pages of text, with many additional pages of
tables and figures.  Too much detail is included, and I suggest a careful editing in this regard.
For example, some text needlessly repeats the results of individual studies.  I suggest not giving
details of any individual study, unless the evidence is regarded as pivotal.

•  My greatest concern lies in the approach to developing the "Summary and Conclusions"
sections. These are the critical elements of the synthesis, and should clearly specify the
additional knowledge gained since the 1996 Criteria Document, the starting point for this review.
Each section related to the major questions carefully sets out where the knowledge stood in the
last document.  However, statements as to the extent of knowledge gains are variable in length
and in the approach taken to characterizing the gain in evidence.

Perhaps, these sections are generally weak because clear rules were never set out for evidence
interpretation.  Additionally, there is a disturbing difference in approach as the section beginning
on page 9-80 considers the "coherence of evidence" at length, while similar sections are not
provided for the other major issues. This section, extending from pages 9-80 through 9-84,
offers a potential model for how other sections might systematically evaluate evidence.

However, the section suffers from an overuse and excessive reliance on the term "coherence"
which is variably used. I caution EPA against developing another approach to evidence
evaluation, (i.e., are based on "coherence") which may differ from approaches taken in other
Agency reviews and which diverges from general practice in public health. Models for evidence
evaluation are  available from reports of the Surgeon General and the National Research Council,
for example.

•  Additionally, the material on health effects might benefit from a several page summary,
which is truly "integrative" across the full body of the chapter. The degree of certainty overall
might be addressed in such a section.

Specific comments:

Page 9-1, Lines 27-29: Research recommendations might be usefully placed in a single section.
At present, I found them scattered somewhat haphazardly throughout the text.

Page 9-6, Lines 11-14: There is inherent inevitability in the choice of PM metrics used for health
research; EPA has shaped research in its choices and by the size fractions monitored.

Page 9-17, Lines 12-21: These statements are sweeping and far too general. I caution against
such broad calls for research. Additionally, the authors should decide where research
recommendations might best be placed in the text, and certainly they should be separately
identified as such.

Page 9-19, Line 17: What does "highly multifactorial" mean?

Page 9-19, Line 23: The statement "frequently being statistically significant or nearly so" should
be removed.

Page 9-20, Lines 4-8: While the multi-city studies have the mentioned strengths, their results
receive little emphasis in this section.

Page 9-31, Line 9-32, Lines 30-2:  What evidence can be cited to support the concluding phrase
of the sentence "in conjunction with covarying gaseous pollutants."  Much of the cited evidence
weighs against effect modification and changes in effect estimates with inclusion of gaseous
pollutants in models may  represent the contribution of the gaseous pollutants to secondary

Page 9-33, Line 28: Would not a greater proportion of susceptible people imply a higher overall
mortality rate?

Page 9-35, Lines 21-23:1 am not sure I agree with this comment concerning selection of single-
day lags.

Page 9-36, Lines 9-14:1 am concerned that the authors are overinterpreting the study of Mar et
al., and perhaps giving undue weight to possible chance variation.

Page 9-39, Lines 18-27: This section is particularly brief and weak and needs substantial

Page 9-41, Line 28: Perhaps a missing phrase?

Page 9-42, Lines 9-11-13: This sentence is too ambiguous; what is meant by "at least under some

Page 9-45, Lines 3-12: This material is too vague and needs  sharpening.

Page 9-49, Lines 26-27: What is meant by "the most applicable to risk assessment"?

Page 9-52, Lines 11-15: Another unclear sentence.

Page 9-57, Lines 20-25: Research recommendations again offered haphazardly. I also think that
the recommendation is far too general to be useful, and the comment concerning mixtures offers
far too simplistic a suggestion concerning a difficult problem.

Page 9-70, Lines 19-26: Quite confusing discussion and use of the term "causality" in a fashion
that I cannot readily interpret.

Page 9-80, 9-85:1 have previously commented on the use of the term "coherence".  In this
section, the phrase is used relatively often, appearing to refer in some instances to consistency of
epidemiological findings and in others to parallel and complementary findings in different lines
of investigation, e.g. epidemiology and toxicology. I would urge that the authors not "abuse" the
concept of coherence,  which has proved too convenient in these pages. As I noted in my general
comments, the other major issues do not have a similar discussion, which might be helpful.

Page 9-101, 9-102: This section is also weak. For example, Line 22, Page 9-101 comments that
"the public health impact of exposures to ambient PM can be quite large"; this is far too general
a description of public health impact.

                                   Dr. Sverre Vedal

Critique of 2004 revisions of draft PM Criteria Document (chapters 8 & 9)
Sverre Vedal
July 20, 2004

Chapter 8:
1. General.
       This revised draft has dealt with nearly all of the issues I had raised about the previous
version, and nearly always successfully.  Disagreements that I have with the CD at this point are
largely due to differences in interpretation of the evidence rather than concerns about fair
representation of the findings  or about factual errors. Rather than detail the many points where
the CD has improved, I will touch largely on those where the authors elected either to leave the
CD unchanged, or where the changes are not adequate.

2. Cohort studies.
       The VA cohort study is now treated more fairly than before.
       On a seemingly small issue regarding the description of the Hoek report from the
Netherlands on the association between residence in proximity to large roadways and mortality, I
still maintain that the unadjusted effect estimate for black smoke (1.34) reported here in the
summary of that report (p. 125, line 9) should not be reported. This is similar to reporting Six
Cities and ACS cohort findings unadjusted for individual risk factors.  It was the ability to
perform this type of adjustment for individual-level risk factors that set these two studies apart
from the previous generation of cross-sectional ecologic studies.  This unadjusted estimate in the
Hoek study is decreased after  appropriate adjustment for covariates, and approaches the null
value when the analysis is limited to subjects who resided in the same area for a given number of
years. While these effects do  not negate findings from the other cohort studies, they do
nevertheless add fuel to an argument that findings from the spectrum  of cohort studies are not
necessarily in agreement (consistent).

3. Natural experiments.
       There is again reference to both respiratory hospitalizations and mortality (p.9-126, L8)
in the description of the Utah Valley steel mill closure. The formal study that directly used the
steel mill closure in the design only involved respiratory hospitalizations.  Mortality was only
analyzed using a traditional time series design (Pope 1992).  In that paper it was stated that
average deaths per day were 3.2% higher when the steel mill was open than when it was closed.
The baseline daily mortality was 2.7 deaths/day, which translates to less than a 0.1 death per day
increase with the steel mill open. The absence of statistical power here explains why this
"finding" was never emphasized, quite correctly, except in this CD.  The continued reference to
mortality here in the CD is not justifiable, at least without appropriate qualifiers.
       A small issue regarding the "natural" experiments in Dublin and Hong Kong:  it is
claimed that these are not useful for quantitative risk assessment (p. 130, LI).  On the contrary,
given that one can be more  confident about control of confounding in these studies, I would
maintain that they are preferred over either the cohort or time series studies for this purpose. It
seems that the only rationale for not including considering them for this purpose is that the time

course of effects is not clearly the same as that in either the cohort or time series studies.  This is
a weak rationale.

4. Co-pollutants.
       The discussion on gaseous pollutant variables as possibly acting as surrogate measures of
some features of PM composition continues to be illuminating.  However, in my opinion there is
too much made of this point (pp. 9.229-231, and elsewhere) since it remains highly speculative.
As I have noted before, a more cogent argument can be made that daily variations in the
concentrations of these pollutants serve as measures of unmeasured features of meteorology,
given that meteorology is the primary determinant of daily concentration changes and, arguably,
a more plausible cause of these health effects than the small daily concentration changes in either
PM or the gaseous pollutants.

Specific &/or editorial comments:
8-5, L21, etc.: In the description of study designs, although panel studies are now no longer
linked  with cohort studies (an improvement), they have now unfortunately been dropped
altogether.  I propose including one group of studies under the "longitudinal" rubric to include:
cohort (semi-ecologic), panel, intervention, time-series (and its variant, the case-crossover
study). The "non-longitudinal"  study types are ecologic, cross-sectional and case-control.
8-48, L2:  "that" = "than"
8-68, L6-7 & 8-77, L12: The Goldberg study did not investigate deaths due to CHF, but instead
looked at total mortality in the stratum of subjects with pre-existing CHF. This needs correcting.
8-145, L9-11: This work was referenced in press in the last version, and is now referenced from
2004.  I protest.  Many equally important studies have not been  referenced in the CD because of
the date cut-off.  The same should apply to this study.
8-153, Figure 8-10 &  8-187, Figure 8-12: Again, what is the basis  for selecting the specific
studies included in these plots?  This is relevant because they are used to generate a range of
effect estimates for this outcome.
8-262, L8-10: Again, the fact that various mortality outcomes might have different lag structures
has nothing to do with the lag structure for a given outcome across  cities.  Clarify this sentence.
Chapter 9:
Overall, I found the organization around the 5 questions to be helpful. This was more of an
integrative synthesis, as intended, although I have criticisms.

Question 1.  How does newly available information continue to support consideration of fine and
coarse particles as separate subclasses ofPM?
I found this discussion credible and useful.

Question 2.  How does newly available information inform our judgments about the strengths
and limitations of epidemiologic evidence for health effects related to ambient fine and coarse
thoracic PM, acting alone and/or in combination with other pollutants?
       My comments here are quite specific, and often reiterate objections that I have raised in
earlier critiques of chapters 8 and 9.

1. Statistical model specification.
       It should be noted that most of the re-analyses included in the HEI report did not report
findings using "alternative modeling strategies" (p.9-3, L3-6).  In fact, the few that took other
approaches found that impacts on effect estimates could be large and often changed the
qualitative conclusions, as opposed to the conclusion presented here.

2. Co-pollutants.
       See my comments on chapter 8 on the topic of gaseous pollutants as PM surrogates.
These are equally applicable to similar statements here (p. 9-30, L27-). Regarding the relatively
strong effect of SO2 in the ACS cohort study, to conclude that it is difficult to interpret because
SO2 is a precursor of sulfate (p.9-31, L20-21) is too dismissive and not very thoughtful.

3. Heterogeneity of effects.
       While chapter 8 of the CD qualifies the lack of statistical evidence for heterogeneity of
effects in NMMAPS by noting the limited statistical power of this test in this setting, no
qualification of this conclusion is included here (p.9-33, L14-15).

4. Consistency of findings.
       The use of "consistency" (section, p. 9-33 to 35) to describe estimates of effect
from the entire body  of studies (particularly the large body of time series studies) has, to my
mind, become nearly meaningless (note that my take on the use of "coherence" below [question
#3] is similar).  This descriptor is repeated so often that it is now accepted without any thoughtful
assessment as to what it really means, or at least what it should mean.  The way it tends to be
used in this setting is that if a study reports  an effect at any lag, and for some model
specification, it is regarded as a "positive" study, and hence consistent with all others in which an
effect is reported. It is very difficult for a study to be "negative", given all of the  associations
that are inspected in these studies, so consistency, by this definition, is almost guaranteed.
Statement of specific hypotheses at the outset of a study would help to partially circumvent this
unfortunate trend, but hypotheses are typically general enough that it is difficult for them to be

5. Lagged effects.
       Justification for the use  of "best" lag is presented here again  (p. 9-35). Given that I have
tried to argue against this, the authors of the CD must disagree with  me and seem committed to
coming up with a justification for using "best" lag. I repeat again, this selection process
guarantees bias.
       I am unclear how the pattern of persistence of pollution in a city (p. 9-35,  L 26-31 and p.
9-37, L6) affects the lag structure.  This may influence how one specifies a pollutant measure in
a city (e.g., number of days to average), but the link with lags is unclear.

6. Found (quasi-experimental)  studies.
       It is again claimed that the Utah Valley steel mill closure study reported effects on
mortality as well as on respiratory hospitalizations (p. 9-38, L 31).  See my Chapter 8 comments
on "Natural  experiments" above.

9-19, L2: "know" should be "known"
Table 9-6 (p.9-22). Note under PM2.5, "dysrhythmias" are listed under "Resp Diseases" and
claim reference #12, which is the Seattle study of asthma.
9-23, L14 (first line): "PM2.5"  should be "PMi0-2.5"
9-33, L28:  Why would a lower mortality rate suggest that "more susceptible people" would be
present?  I would have thought that a lower mortality rate indicated a healthier population (i.e.,
less susceptible).

Question 3.  How does newly available information inform assessment of biological plausibility
and coherence of health effects attributed to ambient fine and coarse thoaracic PM and/or their
       In general, this part provides an evenhanded presentation of findings. Identification of
inconsistencies among studies was refreshing and useful.
       The section summarizing the epidemiological findings (section, pp.9-41 to 44)
is unnecessary. Further, the attempt to identify a role for regional sulfate in this section based on
the studies that employ factor analysis (p.9-44, L3, etc.) to identify sources (i.e., indirectly) is
poorly justified, especially in light of studies that employ a direct measure of sulfate
concentration. Further yet in this section, what does lag have to do with confounding (p.9-44,
L13-15)? In short, this entire section is unnecessary and does a poor job of fairly reflecting the
       Three general categories of effects (cardiovascular, respiratory and mutagenic/genotoxic)
are reviewed (pp.9-54 to 59). Are systemic inflammatory  effects intended to be included under
cardiovascular heading? In addition, it would seem that vasoconstrictive effects (Brook human
exposure study) should also be included.
       A great deal of space is devoted to a  discussion of coherence (p.9-80, etc.). Like
"consistency", "coherence" has been repeated so often it is now accepted as a given.  In this
setting, coherence is claimed when effects on two or more outcomes have been observed in a
city.  Hence, reference is made (pp.9-81 to 82) to figures 8-24 to 8-28 in chapter 8 (pp.8-263 to
267) for Los Angeles, Chicago, Detroit, Pittsburgh and Seattle.  However, no quantitative
assessment of coherence wherein the correlation between effect estimates is analyzed for more
than  one outcome across cities  has yet been published (although NMMAPS IV concerns itself
with this issue).  Further, coherence is moot  in the absence of consistency, yet we seemingly
have rushed forward and claimed coherence when consistency remains at issue.
       Once again there is reference to the Utah Valley steel mill closure study (p.9-83, L4-12)
and mortality effects. Again, see my Chapter 8 comments on "Natural experiments" above.

Question 4.  . How does newly available information inform our understanding of
subpopulationspotentially susceptible to PM-r-elated'health effects?
       It is unclear whether it is being argued that the elderly are more susceptible by virtue of
chronic diseases being more  prevalent in the elderly, or whether higher age alone increases
       Regarding gender effects, there was no mention of the important gender effect in the
cohort mortality studies: in the ACS study, the lung cancer effect was only  seen in males, and in
AHSMOG, the mortality effect was only apparent in males.

Question 5.  . What does the newly available information imply with regard to potential public
health impacts of human exposures to ambient PM in the United States?
       Much of section is a general review of observational findings and is
unnecessary; the points relating to the size of the susceptible pool and on the size of effect
estimates are valuable, however.

                                Mr. Ronald H. White
                         Comments of Ronald White, M.S.T. on
    EPA Criteria Document for Particulate Matter, Chapter 8: Epidemiology of Human
         Health Effects Associated With Ambient Particulate Matter (June 2004)

The sixth iteration of this chapter of the PM Criteria Document provides further refinement of
the PM epidemiologic literature review that was already quite comprehensive in the previous
version, and addresses the major concerns raised by the CAS AC review of the previous draft
document as well as many of the issues raised by other credible scientific reviewers.

Issues related to overinterpretation in the previous draft of this chapter regarding the biological
significance of cardiovascular endpoints such as changes in heart rate and heart rate variability
have been appropriately revised, as have the more speculative statements regarding the possible
mechanisms for cardiovascular-related mortality outcomes.

The revised chapter provides a more balanced discussion of several  of the studies finding mixed
or negative results (e.g. Lipfert et al. 2000 VA cohort study; Abbey  et al.  1999, Beeson et al
1998 ASHMOG studies), as well as expanding the discussion of the cautions and limitations in
interpreting the results of many of the time series and cohort studies included in the chapter. Key
intervention studies (e.g. Clancy et al. 2002; Hedley et al. 2002) that CASAC  suggested for
inclusion in the chapter have been added  and, in general, appropriately interpreted. The
discussion of the results in Hong Kong from the Hedley et al. study  would, however, benefit
from an improved discussion of the potential role of reductions in sulfur dioxide on mortality and
the implications of this finding for other studies where a mortality effect of 862 was found.

The revised chapter also expands and improves the discussion of the potential methodological
uncertainties related to measurement error and model specification in evaluating the
epidemiologic results.

In conclusion, this revised draft of the PM Criteria Document chapter on the epidemiological
evidence of the health  effects of paniculate matter improves on an already encyclopedic and
generally well written  review of the scientific literature published since 1996 on this topic. The
authors have adequately addressed the vast majority of CAS AC's criticisms and suggestions for
improvements of the previous draft document. As such, I find the chapter sufficient for meeting
the Clean Air Act statutory requirement to "accurately reflect the latest scientific knowledge"
and recommend closure for this chapter of the document.

                           Comments of Ronald White, M.S.T.
           PM Criteria Document Chapter 9: Integrative Synthesis (June 2004)

General Comments

This revised draft of the PM Criteria Document Chapter 9: Integrative Synthesis represents a
substantial advancement over the previous drafts of this chapter, and provides a much improved
integration of the key material from the previous chapters of the Criteria Document. The
organization of the material around the five key questions provides an appropriate structure for
an integration of the information contained in the previous chapters of the CD. In general, the
discussion of the scientific information from the previous CD chapters provides a reasonable
balance of indicating the overall results of the scientific information while acknowledging the
limitations of the available information.

However, some sections of the chapter (e.g. section still contain a substantial amount
of detail related to specific study results that could benefit from judicious editing to enhance the
readability of the chapter. The amount of detailed material tends to obscure the key points being
made in several of the sections to address  the related question. This could be accomplished by
reducing the number of specific study citations (as is the case in some sections), eliminating or
moving tables summarizing study results to an appendix, as well as referring the reader to
material contained in the earlier chapters of the Criteria Document where appropriate (as is done
in section

There is extensive reference to "consistency" and "coherence" of study results throughout this
chapter. A clearer definition of what is meant by these terms (and how they differ) in the context
of this chapter would be a useful addition  to the chapter introductory material that discusses how
EPA evaluates scientific evidence (pg. 9-3, lines 8-18).

In conclusion, while this draft of Chapter 9: Integrative Synthesis requires further editing to pare
down the amount of material and improve the consistency of the tone and amount of specific
study detail in the different topics covered, it represents a reasonably well done and
comprehensive synthesis of the key new scientific findings contained in the previous chapters of
the PM  Criteria Document.

Specific Comments

Pg. 9-2, line 10: Emerging evidence from  human clinical PM exposure studies related to
cardiovascular endpoints should be specifically noted here as well in the context of evaluating
the epidemiological evidence. This point is also appropriate for pg. 9-3, line 3.

Pg. 9-19, line 23: "Nearly" statistically significant is a vague term, and doesn't add to the
argument for the strength of the epidemiologic evidence. Statistical significance is only one,
albeit important, measure of the strength of epidemiologic study results.

Pg. 9-23, line 12: Given the much larger prevalence of cardiovascular deaths in comparison to
respiratory deaths, the reduction in precision of the risk estimates for the later in comparison to
the former is not unexpected and should be explained in that context.

Pg. 9-23, line 14: This should read PMi0-2.5, notPM2.5.

Pg. 9-45, lines 6-9: The implications of the sub-daily time lags from the results of the
cardiovascular studies discussed in this section for the PM NAAQS averaging time and
protection from CV health outcomes deserves more discussion.

Pg. 9-49, line 20: A brief (1-2 sentence) summary of the conclusions of the EPA dosimetric
modeling would be useful here.

Pg. 9-89, lines 15-16: The emerging (though limited) evidence of the effects of PM on neonatal
mortality and birth outcomes should be mentioned here as well, and given the birth outcomes
data pregnant women also identified as a potentially susceptible population.

Pg. 9-92, lines 16-18: The ASHMOG and ACS studies (Beeson et al., 1998; Abbey et al, 1999;
Pope et al. 2002) indicate a lung cancer mortality association with long-term PM10 exposure for
males only, and should be noted here.

Pg. 9-94 line 29 - Pg. 9-95, line 2: Given the limitations of the exposure assessment
methodology in the Hoek et al.  study, substitute another example from the traffic proximity
health effects literature.

Pg. 9-95, lines 5-14: Section summarizing the discussion of susceptible subpopulations
should be expanded to more directly emphasize the previous discussion of enhanced
susceptibility due to age (elderly and children) and disease status. There should be discussion
here on the identification of potentially susceptible populations that have emerged or for which
data has expanded since the 1996 CD (e.g. diabetics, neonates).

                                Dr. Warren H. White

Chapter 9 comments by Warren White, 7/29/04

Conceptualization of PM

This chapter is a thoughtful and well-developed introduction to the subject of the CD. There is
one area in which it leaves me a bit unsatisfied, though, and that is its easy acceptance of
aerodynamic size as the "fundamental" discriminator between particle types, overriding other
measurable properties such as chemical composition, water uptake, optical indices, or size
distribution moments. This approach accurately reflects current consensus in the aerosol
community, but I want to keep the door open to alternatives in future reviews.

We would like our PM categories to "carve nature at its joints," and we all recognize that the
atmospheric aerosol includes fundamentally different particle types. My point is that these
fundamental particle types fit comfortably into a variety of different taxonomic schemes,  and that
privileging one particular scheme as normative can inadvertently constrain our subsequent
thinking about atmospheric and biological mechanisms and relationships.

A former Chair of CASAC wrote a book some years ago with the title "Smoke, Dust, and Haze"
(Friedlander, 1977), and we can do worse than to take these labels as the beginning of a PM
*      At the upper end of the particle-size spectrum are the primary emissions from mechanical
processes operating at ordinary temperatures:  "dusts" and, as the API public commenter
reminded us, "sprays." "Dusts and sprays" correspond fairly exactly to what EPA means by
the "coarse mode."
*      At the small end of the size spectrum are the primary particles that nucleate in the
effluents of high-temperature combustion processes.  These "combustion nuclei" are pretty well
what health scientists have in mind when they speak of "ultra-fines".  They are a part, but not the
whole, of "smoke".
*      In the middle is the secondary material that accumulates in 0.2-2 |j,m particles through
nucleation or condensation from the gas phase, evaporation of aqueous droplets, and coagulation
of nuclei with other particles. This hygroscopic "haze" formed in the atmosphere is the essential
component of the "accumulation mode", and in many areas contributes most of the mass in the
"fine mode".

The components "dusts and sprays," "combustion nuclei," and "haze" map fairly neatly into
distinct ranges of particle size, as just noted, but they map equally well into other observable
Chemistry: Combustion nuclei are enriched in heavy metals and elemental carbon soot;  haze is
dominated by ammonium, nitrate, sulfate, and organic material; dusts and sprays are
distinguished by identifying species such as silicates and salts.
Water uptake: Dusts and sprays are of course respectively insoluble and soluble.  Combustion
nuclei are typically insoluble. Haze aerosols are not just soluble, but actively hygroscopic.

Optics: Combustion nuclei are the main absorbers of light, and this absorption can be monitored
in real time. The haze species and sprays are essentially non-absorbing in the visible, and most
soil dusts absorb only weakly.
Integral moments: Nuclei usually dominate the number concentration, haze usually dominates
the surface area and light-scattering cross-section concentrations, and dusts and sprays supply
much of the mass concentration when they are significant components. Number concentrations
and light-scattering coefficients are particularly suited to in-situ and real-time monitoring.

Is a particle's aerodynamic diameter in ambient air somehow more "fundamental" than its other
properties? No - once a particle is inspired into the presumably saturated air of the thorax, even
its aerodynamic behavior is determined as much by its chemical composition (whether
hygroscopic or hydrophobic) as by its diameter in dry ambient air. Even after overlooking this
real complication, dosimetric modeling finds "no sharp cut points that clearly distinguish
between particle size ranges with relatively high versus relatively low fractional deposition
rates." (9-15/9-11).  Moreover the separate biological effects of insoluble nuclei - e.g., demands
on clearance mechanisms - presumably don't simply vanish when are engulfed within an
accumulation-mode droplet, even though their inclusion has negligible impact on that droplet's

I am not arguing that real ambient aerosols are more complex than the current picture of a coarse
mode and a fine mode, or the emerging refinement of coarse mode, accumulation mode, and
ultra-fines. I am not trying to make the point that "everything is complicated." I am only trying
to say that there are other useful descriptive frameworks that are just as simple as particle size.
As an example, one could define smoke,  dust, and haze in terms of black carbon, silicon, and
sulfur as the respective indicators, all of them measured by light absorption and XRF on the
same Teflon PMi0 filter. There would then be no need for the CD's subtle expositions of the
distinctions between fine particles, the fine mode, and PM2.5, or discussions of 1 |j,m vs. 2.5 |j,m
as cut-points.   There would, of course, be a different set of angels-on-pinheads debates.

I don't expect the foregoing to  convince anyone that EPA should now switch from particle size
to chemical composition or some other basis for its PM standards, and I'm not sure I would want
it to do so.  All I'm really seeking is a clear and continuing acknowledgment, in this CD and in
subsequent documents, that particle size is one of several  indicators that are useful in
distinguishing particles from different sources having different  effects, and not uniquely the
determinant of particle properties. I suppose we can't put new words into Ken Whitby's mouth
at 9-4/25 (I wish he had said "The distinction between 'fine particles' and 'coarse  particles' is an
indicator of fundamental differences" there), but perhaps at least 9-15/23-25 could be changed to
    The distinction articulated in the last review, between fine  and coarse ambient
    particles as indicators of fundamentally  different sources and composition, formation
    mechanisms, transport, and fate, remains generally unchanged.

The coarse particle mode

The following three statements are potentially misleading, as they may be read to suggest that the
fine tail contributes an unusually large fraction of coarse-mode PM in dry dusty areas and dust
    9-5/19    .. .in dry dusty areas, resuspended coarse-mode particles may extend down
    to about 1 |j,m; ...
    9-7/1     .. .in dry dusty areas, coarse PM (e.g., resuspended soil) may have a tail
    reaching to 1 |j,m or below.
    9-10 table  bottom,  [coarse travel distance is]
       <1 to 10s of km, (100s to 1000s in dust storms)
Fine coarse-mode PM is indeed elevated in dry dusty areas and dust storms, but coarse and giant
particles are proportionately even more elevated. The coarse and giant particles are
preferentially attenuated as they move away from these sources, so the fine tail is actually a
larger fraction of the coarse mode away from dry dusty areas, and the characteristic lifetime of
the coarse mode in dust storms is shorter than it  is at similar altitudes under fair conditions.  The
fine tail of the  coarse mode is always there, in other words. The point you mean to make, I think,
is that it is most important as a component of PM^ when the coarse mode dominates the fine
mode - this is  the significance of dry dusty conditions and dust storms.


Rich Poirot and I have been pushing throughout  this  review cycle for a clearer acknowledgment
of PM2.5's dominant role, as modulated by relative humidity, in determining visibility impacts.
The very first substantive statement here aims explicitly to obscure this point:
    9-103/6-8:      More  specifically, the efficiency ...  depends on not just the mass
    of fine particles, but also on ... [emphases added]
As a more positive substitute that says the same  thing, I suggest substituting the following
    9-103/6-9:      More  specifically, light scattering and absorption by particles of
    given composition and size distribution are  strictly proportional to their mass
    concentration in the air. These optical interactions, by which airborne particles
    degrade visibility, are well characterized in  terms of a light extinction ...

On another point, I disagree with the claim that
    9-104/12:        Our understanding of how ambient PM affects visibility has
    historically focused on visibility impairment in rural areas, particularly in national
    parks and  wilderness areas  ...
    9-104/24:         ... historically the relationship between ambient PM and visibility
    has been less well studied in such [urban] areas.
I would argue that the majority of our current understanding and theory - for fine particles in
general, and not just visibility - actually derives  historically from urban studies, particularly in
early-1970's Los Angeles (the concepts of bi-modal  size distribution and extinction budget,
estimation of species' extinction efficiencies by  multiple regression, accounting conventions
such as the value 1.4 for organic compound/carbon mass ratio) and in early-1980's Detroit
(measurement  of particle-bound water contribution, sophisticated aerosol optics modeling (Chris
Sloane), accurate characterization of absorption).

                                 Dr. George T. Wolff
                    Comments on Chapters 7, 8 and 9 of the June 2004
                        Criteria Document for Particulate Matter

                                     George T. Wolff
Chapter 8
    1.  p 8-2, lines 30-31 to p. 8-3, lines 1-2 - It is stated that papers published after April, 2002
       will be included if they provide "particularly important information helpful in addressing
       key scientific issues."  On p.  8-3, line 14, "model specification" is listed as one of the
       important issues that this chapter addresses. A paper, brought to the Agency's attention
       during our February 2004 teleconference and in written comments, that appears to be
       central to this issue is not discussed. The paper by Koop and Tole (J. Envir. Econ &
       Mgt, 47:30-54, 2004) and several  other references cited within question whether time-
       series studies are a valid way to study air pollution-mortality relationship. Since almost
       the entire chapter and its conclusions are based on time-series studies, these papers
       cannot be ignored.

    2.  p. 8-20, line 14 - "interpreted with caution" is misleading. Any results from uncorrected
       GAM studies should be ignored.

    3.  p. 8-22, lines 19-20 - I don't see how this follows from the discussion on pages 8-19 to 8-

    4.  p 8-22, lines 20 - 22 - How can this be stated in light of the "interpreted with caution"
       statement on p. 8-20.

    5.  p 8-35, line 4 - "weakly associated" - This is very misleading. If you examine figures 12
       and 14-16 in the revised NMMAPS, you will see that there is one lag for each of the
       gases that show a significant correlation and an effect equal to  or greater than PM.

    6.  p. 8-70 to 8-74 - Somewhere in this section, it should be pointed out that Pb was used as
       a tracer for vehicle exhaust in these studies that utilized relatively old air quality data.
       Presently, and for the last 10 years or so, we can no longer use lead as a tracer of
       vehicles, because highway vehicles no longer burn leaded gasoline.  In those intervening
       years, vehicle PM emissions have  declined significantly as the composition has changed
       as well. Consequently any inferences drawn for these studies, may have no applicability
       to the situation today.

    7.  p. 8-73, line 13 - After "vehicle" insert "/road dust."

8.  p. 8-80, line 5 - The "more than 80 new time series studies" makes a great sound bite, but
   will be used out of context without the accompanying caveats. I suggest you delete it and
   just talk about the non-contaminated studies.

9.  p. 8-81, lines 12-18 - If you look at figures 12 and 14-16 in the NMMAPS reanalysis,
   you will see that the inclusion of PM did not affect the gaseous effect size estimates

10. p. 8-82, lines 11-20 - It would be appropriate to add an abridged caveat discussed in my
   comment 6 here.

11. p. 8-84, lines 2-3 - This conclusion is out of place before the VA and AHSMOG are even

12. p. 8-88, line 29, p. 8-99, lines 27-28, and p. 8-100, line 7 - These are misleading
   understatements that obscure the fact that those with a high school education or more
   showed no statistically significant response.

13. p. 8-101 to p. 8-106 - The discussion of the AHSMOG study is not very balanced. The
   bottom line of this study is that it shows very little indication of a PM/mortality
   relationship.  However, the Agency focuses on the few subsets of analyses that show a
   positive relationship and tries to rationalize why the rest of the analyses fail to show a
   positive response.

14. p. 8-111, line 29 - How does subject size account for  different results?

15. p. 8-113, lines 13-16 - For these same reasons, this make the VA cohort a susceptible
   population which should give special insights into cardiac responses.

16. p. 8-114, lines 22-23 -  Why would cohort depletion be an issue only in the VA study?

17. p. 8-116, lines 1-12 - This clearly comes off as blatant case of cherry-picking the results
   that support the Agency's position.

18. p. 8-118, lines 14-17 - The reasons for not including the VA results in Tables 8-15 and 8-
   16 defy logic. This is another example of selective use (or in this case omission) of the
   data to support the Agency's position.

19. p. 8-121, lines 1-4 - This is only true if the VA and AHSMOG are excluded.

20. p. 8-124, line 20 to p. 8-125, line 24 - There are so many flaws in the Hoek at al study
   that it should not be used to support any arguments. First of all, NO2 cannot be used as a
   tracer of fresh vehicle exhaust since it is a secondary product.  As a result, its
   concentrations tend to be more homogeneous than a primary emission.  NO or CO would
   have been a much more reliable tracer.  Second, and more importantly, Hoek et al
   manufactured the concentration fields for BS and NO2 by using unvalidated statistical

   methods and data from two short sampling studies in 1995. They then extrapolated 1987-
   1990 monitoring data to represent the ambient air quality during his 1986-1994 health
   study. This is so far removed from actual exposure data, that the study should just be

21. p. 8-130, line 25-1 think a few more words on the education effect are in order
   explaining that the only significant effect is seen with the 20% of the cohort who did not
   finish high school.

22. p. 8-130, line 7 to p. 8-133, line 9 - It needs to be mentioned that most of the AHSMOG
   results disagree with the other two studies. In addition, the VA's no effects results need
   to be included.

23. p. 8-227, line 1 to p. 8-229, line 6 - This section does a good job summarizing the HEI
   Commentary. However, it leaves the reader hanging. The last bullet on page 8-229 is
   powerful and has potentially wide-ranging implications concerning the validity of time-
   series studies results.  This bullet plus the results presented in the Koop and Tole paper
   cited above must be addressed because they potentially undermine the validity of the
   time-series results.

24. p. 8-237, lines 17-25 and Figure 8-16 - Again, I refer the authors to figures 12 and 14-16
   in the revised NMMAPS report.  These figures show that there is one lag for each of the
   gases that show  a significant correlation and an effect equal to or greater than PM. As
   co-pollutants are added, the effect size estimates change little. Consequently, using
   figures 8-16 to 8-19 to make decisions on confounders, is not valid unless the co-
   pollutants are treated in the same way.

25. p. 8-245, section -1 am happy to see this section added, bur it leaves me hanging.
   Should we be concerned about bioaerosol confounding or not?

26. p. 8-246, section -1 am ecstatic to see this discussion  of meteorological variables
   and time trend model specifications. However, after reading it, I felt I was left hanging
   again. What are the implications?

27. p. 8-256, line 6 - Do the Smoyer et al papers employ GAM?

28. p. 294, lines 24-28 - Again for the reasons stated above, this sounds like a recipe for
   cherry picking.

29. p. 8-295, section - My above comment 6 applies to this  section as well.

30. p. 8-310, line 28 - The heterogeneity shows itself in these studies by the fact that the
   specifications of the models for each city are very different.  This should be pointed out.

31. p. 8-311, lines 7-16 -It should be pointed out that the results  of most of these studies are
   unreliable because of the GAM issue.

   32. p. 8-311, lines 19-20 - Again, "newly apparent heterogeneity" is a misnomer because the
       older studies displayed model specification heterogeneity.

   33. p. 8-312, lines 7-27 - This explanation of heterogeneity needs some graphs to back it up.

   34. p. 8-323, line 11 - Given the model specificity issue, how can one claim that the reported
       multi-city estimates are more precise?

   35. p. 8-332, lines 19-29 - It should be pointed out that these are the Agencies conclusions
       because they ignore 2  of the long-term studies.
Chapter 9

    1.  p. 9-20, lines 29-31 - By using only single pollutant studies, some of the heterogeneity of
       results is masked.

    2.  figures 9-5 and 9-6 - The figure caption needs to explain what the lengths of the vertical
       lines mean.

    3.  p. 9-26, lines 1-15 - Somewhere in this section, it should be pointed out that Pb was used
       as a tracer for vehicle exhaust in these studies that utilized relatively old air quality data.
       Presently, and for the last 10 years or so, we can no longer use lead as a tracer of
       vehicles, because highway vehicles no longer burn leaded gasoline.  In those intervening
       years, vehicle PM emissions have declined significantly as the composition has changed
       as well. Consequently any inferences drawn for these studies, may have no applicability
       to situation today.

    4.  p 9-28, lines 1 and 2 - The VA results were  consistently non-significant.

    5.  p. 9-28, lines 6-6 - If you are going to ignore the VA and AHSMOG studies, they at least
       deserve a short summary of their results.

    6.  p. 9-29, line  9 to p. 9-30, line 8 - This section is very inadequate. It fails to convey the
       importance of the model specificity issue and the concerns raised by the HEI Special
       committee.  It needs to be expanded to convey these concerns.

    7.  p. 9-31, lines 8-11 -1 refer the authors to figures 12 and 14-16 in the revised NMMAPS
       report. These figures show that there is one  lag for each of the gases that show a
       significant correlation and an effect equal to or greater than PM. As co-pollutants are
       added, the effect size estimates change little. Consequently, the argument made by the
       Agency here is not valid unless the gases are examined in the same way as PM.

    8.  p. 9-32, lines 25-29 - If the PM2.5 is dominated by NO3, the infiltration factor would be

9.  p. 9-33, lines 14-15 and lines 30-31 - This statement cannot be made without the
   appropriate caveat.  The test that was used had low statistical power so it was not

10. p. 9-58, line 14- Is this referring to the Hoek et al study?  If so, see my comment 20 on
   Chapter 8.

11. p. 9-62, lines 23-24 - What was the concentration of the sulfate-coated carbon black?

12. p. 9-65, line 5 - Metals do not make up a substantial part of the mass.

13. p. 9-81, line 24 to p. 9-82, line 3 - On Figures 8-24 to 8-28,1 do not understand where
   the Dominici (2003) estimates came from.  Shouldn't they be the same as the ones shown
   in Figure 8-1?

14. p. 9-85, line 4 to 9-86, line 4 - This section is not very convincing.

15. p. 9-93, lines 22-23 - The Gwynn and Thurston reference should not be used because it is
   based on the flawed GAM analysis.

16. p. 9-94, lines 25-27 - This is pure speculation.

17. p. 9-94, line 28 to p. 9-95, line 2 - As I said before, the Hoek et al study is so poorly
   done, it should not be relied on.  See my comment 20 on chapter 8.

18. p. 9-95, lines 5-14 - This summary and conclusions section is very weak as evidenced by
   the number of qualifiers used. It is not at all convincing.

19. p. 9-100, lines 22-25 - How can this be said in light of the previous sentence? How can
   this be said in light of the model specificity issues?

20. p. 9-100, lines 25-27 - How can this be said with such certainty when two of the
   prospective cohort studies were ignored?

21. p. 9-116, line 30 - Delete the rest of the sentence after the word "gases." The rest is
   speculation and this document does not need to go there.

                   Additional Comments on Chapter 9 of the June 2004
                        Criteria Document for Particulate Matter

                                     George T. Wolff
Based on the discussions at the July 20-21 CAS AC meeting and the written comments of other
panelists, I would like to add a few other points.

Section 9.2.2

Although I disagree with the Panel's decision not to include the Koop and Tole (2004) paper in
the CD, the issue of model uncertainty/model selection should still be rigorously discussed in
both chapter 8 and chapter 9.  The idea that the results of a single model based on sequential
hypothesis testing will under-represent the overall uncertainty is not new to the Koop and Tole
paper.  Similarly, if important explanatory variables are omitted either because of model
selection decisions based on imperfect knowledge or a lack of appropriate measurements, the
results of the reported model will under-represent the overall uncertainty.

As Jon Samet indicated, there is not any one "true" model. The discussion of the decision to
emphasize the results of selected single-pollutant models needs to acknowledge the potential bias
and uncertainty as a result of that decision.

The presentation of results in this section in the text and  in figures 9.5 and 9.6 hinges on how one
considers the multi-city studies together with the individual city studies.  For example, if the
NMMAPS mortality results are viewed as 88 or 90 point estimates of which only a couple are
statistically significant, one gets one impression of the strength and consistency of the data.  If
one focuses on the pooled estimate, one gets a different impression. I would argue that both
views are important and should be presented. One way to do this is to include the range of
individual point estimates whenever a pooled multi-city  result is given.

Section 9.2.3

The idea of adding a table or tables to summarize the toxicological results by endpoint has merit.
However,  as Fred Miller indicated, the tables need to provide a perspective on the doses used in
the relevant tox studies as compared to ambient doses. The method of exposure also matters so
whether the evidence is from inhalation, instillation or in vitro studies needs to be listed too.

The material in Appendix 7A responds to repeated requests by the Panel on this subject.
However,  it would be particularly helpful if additional findings from the calculations can be
brought out in the Appendix and then brought into the discussion in Chapter 9.  For example,
representative  doses for comparison with the acute and chronic studies can be derived from the
material. For comparison with acute epi studies, we are interested in the deposited and retained
doses in humans in one day or up to several days for total fine PM and PM components.  The
amounts per lung surface area and per kg body weight should be reported. The caveat that

localized doses may exceed the average should also be included.   For comparison with chronic
epi results, typical deposited and retained doses will be much greater, (see Figures 7A-8 and 7A-
10, for example.)

In both cases, the dose from ambient PM is added to the dose already in the lung from earlier
exposures. For acute exposures, since the chapter argues that exposures to ambient PM are
generally independent of exposures to indoor-source PM, the small increment from ambient PM
is the relevant potential cause of the health outcomes implicated by the epi studies. However, for
chronic studies, the total dose of particles from outdoor, indoor, active smoking,  passive
smoking, occupational sources, and the personal cloud all contribute to the chronic burden.   As
Dr. Vedal pointed out, the use of central monitor measurements to characterize chronic
exposures is a significant issue.  Whether we can  assume that the chronic burden from all non-
ambient sources is the same from city to city is an important research question. Perhaps
differences in the overall chronic burden within the cohorts can explain some of the differences
among the chronic studies.

To put the in vitro results in perspective, some comparisons in terms of mass or number of
particles per cell should be added to give the reader information on the general order of
magnitude difference in doses between the in vitro literature and typical acute human exposures.

The point made on page 9-60 that, based on tox  studies, some types of particles are more toxic
than others, needs elaboration.  It fits in with statements in chapter 8 that some components may
be benign at current ambient concentrations.  The discussion last week by various panelists that
we should be thinking  about  or  moving toward regional standards or standards  for specific
components as indicators of different sources, all fits in with  the general understanding that all
particles are not alike in terms of their toxicity and effects. It is also relevant to general point 5
raised in Roger McClellan's written comments.   This is an important finding from the science
that needs to be clearly stated in the CD.

It is also a major factor behind the reluctance of some panelists to push for stringent coarse
standards. Crustal components and bioaerosols are two major categories of coarse particles but
the evidence for direct crustal effects is limited and bioaerosols, even though they may vary
substantially in toxicity,  are generally thought to be not conducive or subject to control.

Section on coherence on pages 9-81 to 9-85

This is an important section that needs to pull together and compare the results from epi, tox, and
dosimetry studies. The dosimetry examples can be particularly useful in this context. The
discussion of the Utah Valley results as compelling needs to be re-considered in terms of the new
dosimetric comparisons in 7A that show how large the instilled doses are compare to daily or
even weekly inhalation doses. Dr. Oberdorster's comment on the particularly high dose rate for
the instillation compared to Utah Valley ambient PM levels during the period should also be

There is only one paragraph on page 9-84  discussing coherence in relation to chronic effects.
Although the ACS and Six-City Studies reported associations with cardiopulmonary mortality,

the HEI re-analysis pointed out that the association was with cardiovascular and not respiratory
mortality.  This distinction is important. The text argues that there have not been tox studies of
chronic PM exposure or epi studies investigating chronic cardiovascular disease. However, there
is a substantial body of occupational PM information and what is says or does not say about
chronic cardiovascular health outcomes would be enlightening.  This section should also discuss
the variation in the chronic mortality studies in relation to the dosimetric considerations noted
above. The gender difference in the AHSMOG cohort (with a tendency for positive results in
males and negative or protective effects in females) is too large to be accounted for by a small
difference in time spent outdoors between the genders.  The VA cohort is a susceptible sub-
population so its negative association with PM is puzzling if the ACS result is due to ambient
PM differences. While, the text acknowledges that that evidence of coherence in the chronic
studies is "somewhat limited," consideration of the factors I have noted renders the evidence  as
extremely limited.

For the acute studies, the text should include and discuss the dosimetric comparisons noted

Finally, the question of consistency of results needs to be addressed.  For example, the mixed,
inconsistent, and sometimes conflicting results in cardiovascular studies needs to be
acknowledged and discussed in chapter 9 both earlier in Section 9.2.3 and in the coherence

Also, the EMA comments prepared by Dr. Vendetti should be accommodated.

A final thought on biological plausibility

As the Agency goes through the various CD's for different pollutants, the question of what to do
with a growing number of pollutant/mortality and morbidity associations for each of the criteria
pollutants will come to center stage. While it is critical that EPA and the scientific community
rigorously evaluate biological plausibility in terms of concentrations present and nature of effects
for PM, perhaps such an effort should be extended to include all the criteria pollutants.  I do not
know what form this might take or what forum might be used to carry it out, but leaving the issue
to consideration in separate CDs leaves too many degrees of freedom, from missing effects
because they are blamed on another possible causal pollutant to double or triple counting of
mortality from air pollution and blaming it on each pollutant separately.


       This report has been written as part of the activities of the Environmental
Protection Agency's (EPA) Clean Air Scientific Advisory Committee (CASAC), a
Federal advisory committee administratively located under the EPA Science Advisory
Board Staff that is chartered to provide extramural scientific information and advice to
the Administrator and other officials of the EPA.  The CAS AC is structured to provide
balanced, expert assessment of scientific matters related to issue and problems facing the
Agency. This report has not been reviewed for approval by the Agency and, hence, the
contents of this report do not necessarily represent the views and policies of the EPA, nor
of other agencies in the Executive Branch of the Federal government, nor does mention
of trade names or commercial products constitute a recommendation for use. CAS AC
reports are posted on the SAB Web site at: htt^//