March 27, 1997
EPA-SAB-IHEC-LTR-97-005
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
Subject: Science Advisory Board Review of the Office of Research and
Development's (ORD) draft Exposure Factors Handbook
Dear Ms. Browner:
The current Exposure Factors Handbook (EPA/600/8-89/043) was first published
in July 1989 by the ORD's Office of Health and Environmental Assessment. This
Handbook was prepared in response to requests from many EPA regulatory program
offices for additional guidance on selecting values for exposure factors when perform-
ing risk assessments. The Handbook was intended to encourage consistency in
exposure assessments, while allowing risk assessors the flexibility to tailor assessment
approaches to specific situations. Since its publication, new data on exposure factors
have become available, and revisions were necessary to update the Handbook. The
current revised draft Handbook updates information on the consumption of drinking
water, fruits, vegetables, beef and dairy products, and fish. It also addresses factors
such as soil ingestion, inhalation rates, skin surface area, length of life, activity
patterns, and body weight. In addition, the current draft adds new chapters dealing
with grain consumption, consumer products use, and the reference residence.
Following a request from the ORD, the SAB's Integrated Human Exposure
Committee (IHEC) met on December 19-20, 1996 in Arlington, VA to review the draft
revised Handbook. The Charge for the meeting called for the Committee to focus on
the following questions:
-------
a) Is the Handbook consistent with the EPA Exposure Guidelines?
b) Are the data presented in a way that is useful to exposure assessors?
c) Are groupings of background studies into "Key" and "Relevant" categories
done in an appropriate manner?
d) Do the Handbook's recommendations reflect the proper interpretation of
the data and the appropriate characterization of the data's limitations and
uncertainties?
In terms of an overall assessment, the Committee finds that the revised edition
of the Exposure Factors Handbook has been substantially expanded relative to the first
edition, and that it provides a great deal of useful data for exposure assessment. The
Committee commends the Agency on its efforts to assemble, evaluate and present
up-to-date data for exposure assessors in this revision to the Handbook.
Addressing item a) of the Charge, we note and commend the Handbook's
emphasis on presenting distributions of data wherever possible. This approach, in
particular, helps harmonize the Handbook with the Exposure Assessment Guidelines.
The Committee noted, however, that the Handbook must be made more consistent with
the Exposure Assessment Guidelines with respect to the definitions and usage of the
terms "exposure" and "dose." In addition, revisions (detailed below and in the Techni-
cal Appendix to this letter) to the summary tables at the end of the Handbook's various
chapters would also serve to increase the document's consistency with the Exposure
Guidelines.
In the Exposure Assessment Guidelines, the Agency did a very good job of
clearly defining many of the terms used in exposure assessment and of showing their
relationships (In fact, these definitions have been adopted by the Journal of Exposure
Analysis and Environmental Epidemiology as Key Words and definitions). The Factors
Handbook should use these definitions. With respect to the tables summarizing the
recommended data for use, the Committee recommends that these tables should
present the 50%, 90% and the 99.9% (bounding estimate) wherever possible to
emphasize the relations to key distributional percentiles in the Exposure Assessment
Guidelines. Because of the importance and usefulness of these tables, the Committee
also recommends that some way be found to make it easier to find these key tables in
each chapter, e.g., by changing the print size or through some other formatting
changes.
Vis-a-vis Charge element b), the IHEC found that the data in the Handbook are
generally presented in a way that is useful to exposure analysts. In order to make it
-------
even more useful and user-friendly, the Committee recommends that the Handbook be
carefully edited to achieve consistent reporting of summary statistics in the text and
tables throughout the document and consistency with respect to regional and ethnic
groupings in the tables, wherever possible.
Addressing item c) of the Charge, the Committee found that the groupings of
background studies into "Key" and "Relevant" categories were, in most all cases, done
in an appropriate manner. The only significant exceptions to this finding were the "Key"
and "Relevant" ratings given in those chapters utilizing obsolescent data on food
consumption (see below and the Technical Appendix).
The final element of the Charge, item d), asked the Committee to determine if
the Handbook's positions reflected a proper interpretation of the data and characterized
appropriately the presented data's limitations and uncertainties. The Committee has
identified some problems in this area. Summary tables at the end of each chapter of
the Handbook present an assessment of the confidence in the recommended data, and
provide information with respect to the criteria used to make this judgment. The
Committee noted, however, that it is not clear how the overall rating in each table was
derived. Presumably some criteria were more heavily weighted than others. This
needs to be more clearly defined. The Committee also noted that the precision and
accuracy of the measurements appeared to be a heavily weighted (but implicit) criterion
used in arriving at the overall rating, and suggests that these criteria be decoupled from
the others and made more explicit.
Addressing other issues arising during the review, the Committee recommends
that the Chapter on variability and uncertainty be revised to emphasize variability,
which is clearly related to the distributional nature of the data presented in the Hand-
book. The material on the distinction between natural variability and uncertainty should
be retained, but much of the material on uncertainty should be edited to shorten and
de-emphasize it, since the purpose of the Handbook is to provide data for exposure
analysis, not to provide a treatise on the complex subject of uncertainty and how to
treat uncertainty in exposure analysis. In particular, sections on both uncertainty and
variability which are not clearly related to other materials in the Handbook should be
reduced or removed.
Although not discussed by the IHEC as part of the public review, one Member of
the SAB Executive Committee identified an additional issue concerning uncertainty.
The Handbook (page 1-10) discusses cases in which only a ranges of values is known
for an exposure factor, and offers, as an option, use of a mid-point value. This is not
-------
considered to be a good option unless there is reason to believe that distribution within
the range is uniform, and/or a sensitivity analysis has been performed.
Finally, as mentioned above, the Committee found that some of the tables on
food intake (meat products in particular) were not as up-to-date as they should (and
could) be. We urge the Agency to obtain more current data available from the depart-
ment of Agriculture and revise the relevant sections of the Handbook as quickly as
possible. Food production, the dynamic change in consumer food practices and
consumption, and a vast number of new food technologies and products, etc., require
data that is contemporary for appropriate use in risk assessment processes. However,
we also recognize that dietary intake is a "moving target" and that the risk analysis
community cannot wait another year or more for the revised Handbook. One possibility
would be to issue Volumes 1 and 3 immediately, and delay issuing Volume 2 until the
tables are brought up to date. Another, less desirable option, would be to issue
Volume 2 with the other volumes, with the out of date food consumption tables, but with
an alert to the user, and a statement that replacement tables with more up-to-date data
will be issued as soon as possible. Vis-a-vis the issue of timely dissemination of the
Handbook, the Committee suggests that the document might be issued in a binder-type
format so that relevant updated sections could be incorporated as they are produced.
The Committee also discussed the possibility, as raised by one of the public
commentors, of posting the document to the World Wide Web in an interactive format.
Given the current "state-of-the-art, however, the Committee has some reservations
about this proposal and does not recommend it at this time.
Beyond dealing with the specific food data problems in the Handbook, the
Committee suggests that EPA consider supporting an Agency initiative, potentially in
collaboration with other Federal agencies, to develop and maintain a contemporary
database on food exposure factors. Two factors support this suggestion. First, as
noted by the Committee, some of the data that EPA must rely on is up to 20 years old
and there have been extensive changes in the production and consumption of
agricultural products including grains, dairy and meat and poultry products.
Secondly, agricultural business has increasingly become an international market
in which American farmers and industry must compete. Food and animal exports
worldwide have increased five-fold in the last 25 years and continue to expand
(Reference: Codex Alimentarius). Many of the issues on harmonization and
standardization of food safety standards are being addressed by U.S. participation in
the Codex Alimentarius Commission. Specific examples include the Codex Committees
on Pesticide Residues, Food Additives, Food Hygiene, Residues of Veterinary Drugs in
-------
Food, and potentially the Codex Committee on Import/Export Inspection and Certifica-
tion. Specific risk assessments and development of international food safety standards
require the best available food consumption data to develop contemporary scientifically
based standards.
We appreciate the opportunity to review this document, and look forward to the
Office of Research and Devlopment's response to the issues we have raised.
Dr. Genevieve Matanoski
Chair, Science Advisory Board
r. Joan M. Daisey
Chair, Integrated Human Exposure Committee
ENCLOSURES
-------
APPENDIX A
DETAILED TECHNICAL COMMENTS
Specific Comments and Corrections for Chapter 1 (Risk Calculations)
p. 1-10, column 2, line 2: IR and ED are not defined; the symbols can be inserted in the
paragraphs below which define intake rate and exposure duration.
p. 1-10, column 2, line 2: "total exposure can be expressed as .... total potential dose..."
- Exposure is defined as the integral of concentration in the exposure medium times
time. The equation which is given in the Handbook defines a dose. The definitions of
exposure and dose should be made consistent with those in the Exposure Assessment
Guidelines. In addition, the text should be carefully reviewed to ensure that the usage
of these terms is consistent with the definitions.
p. 1-11, paragraph 4: "Exposure can be expressed as ..." - This is not consistent with
the Exposure Assessment Guidelines. This sentence addresses dose and it is not
correct to say that "exposure can be expressed as..."
p. 3-10, Table 3-11: Units are not given.
Specific Comments and Corrections for Chapter 2 (Analysis of Uncertainty)
The concepts, sources and treatments of uncertainly presented in the chapter on the
analysis of uncertainty do follow the USEPA's Guidelines for Exposure Assessment.
This chapter provides very important information for exposure assessors on the issues
of uncertainty. However, the way that uncertainty is handled may cause some confu-
sion to readers who attempt to distinguish uncertainty from variability.
The authors of this Handbook have made a clear distinction between uncertainty and
variability, based on the USEPA's Exposure Guidelines, by defining "uncertainty"as a
lack of knowledge about factors affecting exposure, whereas "variability" is true
heterogeneity across people, place or time. In other words, uncertainty can lead to
inaccurate or biased estimates, whereas variability can affect the precision of the
estimates and the degree to which they can be generalized, (p. 2-2) It is also further
stated in Section 2.2, by quoting USEPA's position, that variability should not be
treated as a specific type or component of uncertainty. But inconsistent with the above
statement, variability has been repeatedly treated as a subclass of uncertainty in many
parts of the document. Here are some examples:
A-1
-------
- On the next page (p. 2-3), readers will be puzzled by Table 2-1 where
uncharacterized variability is included as one of four sources of parameter uncertainty.
The related text of this table can be found on the following page (p.2-4) where "Sources
of parameter uncertainty include measurement errors, sampling errors, variability,
and use of generic or surrogate data" is stated.
- In Chapter 1, Section 1.3.1. under General Considerations, "Minimal (or defined)
uncertainty in the data: Studies were sought with minimal uncertainty in the data,
which was judged by evaluating all the considerations listed above" (p. 1 -2). The
considerations include variability in the population. This inclusion is also stated in the
next sentence: "At least, studies were preferred that identified uncertainties, such as
those due to inherent variability in environmental and exposure-related parameters or
possible measurement error."
- Section 2.6. deals with both variability and uncertainty, but its title is "Presenting
Results of Uncertainty Analysis." Chapter 2 has separate sections on variability and
uncertainty, however only uncertainty is identified in the chapter title ("Analysis of
Uncertainty"). The heading of uncertainty alone, instead of referencing both
uncertainty and variability, in the section and chapter titles implies that variability is a
component of uncertainty.
Occasionally, the authors make no distinction between variability and uncertainty, such
as in Section 2.5 ("Methods of Analyzing Uncertainty and Variability"). Although both
terms are defined initially as different entities in this chapter, the whole section treats
them as inter-changeable entities. Based on the methods of quantitative analysis
presented in this section, readers have no guidance on how to deal with them
differently. Four approaches to quantitative analysis of uncertainty are described in
Section 2.5. But readers still do not know how to apply these four approaches to
increase precision and reduce bias/errors quantitatively. Users of this Handbook would
expect to see the exact operative procedures to handle uncertainty and variability
quantitatively during the process of exposure assessment.
For exposure assessment, variability in measurements is important for contaminant
concentration and other exposure factors. In addition to measurement errors
mentioned under parameter uncertainty, measurement variation deserves some
attention also. In this chapter, the measurement variability has not been addressed at
all. The need to address measurement variability lies in the inherent nature of
measurement. It is also a very important consideration by evaluators for their overall
ratings of exposure factors. Although the precision of measurement tools is not listed
A-2
-------
as one of the considerations for overall rating, it is the implicit consideration by
evaluators for their judgement. As the overall ratings in Table 1-2, physical measure-
ments with high precision and accuracy such as inhalation, body surface area, lifetime,
and body weight are all rated high. Although food intake surveys, which are compli-
cated in nature, subject to recall bias and difficult to quantify, are in general rated as
medium to low. Surveys of time-activity patterns, which are less complex, more reliable
and consistent than food consumption surveys, are rated as medium to high. Measure-
ment variability is clearly the most important consideration for the overall rating of
factors.
To avoid confusion and improve clarity, the following changes are suggested:
a) Shift the emphasis from uncertainty to variability. The term "variability" is
clearly defined and well addressed in the Handbook. For all factors,
variability has been presented in numerous tables of means, standard
deviations and percentiles. While there is a lack of consensus on the
definition of uncertainty as well as its distinction from variability. The
treatment of uncertainty in this Handbook is far less extensive and sys-
tematic than that of variability.
b) Change the title of Chapter 2 to "Variability and Uncertainty" to indicate
the shifting of the weight of treatment.
c) Define and address variability before uncertainty. Alter the sequence of
sections and rename some of the titles to reflect the emphasis on variabil-
ity.
d) Remove or reduce the parts that are not related to the other chapters and
cannot be easily used by exposure assessors.
Specific Comments and Corrections for Chapter 3 (Drinking Water Intake)
The Exposure Guidelines are not cited in the Exposure Factors Handbook, but given
that the guidelines are largely policy and definitions, the Exposure Factors Handbook is
mostly data. Therefore, there is not a big potential for error with regard to inconsis-
tency. In one specific case, however, there is an overlap in the Handbook's description
of percentile of exposure. Fig 5-1 of the Exposure Guidelines has the terms "high end
of exposure" "and bounding estimate" which correspond to particular percentiles. It
would be convenient if Table 3-30 (Summary of Recommended Drinking Water Intakes)
A-3
-------
in the Handbook had 50th, 90th and 99.9th percentile values to match decision points
of the guidelines.
The summary table should be introduced with its own subsection in the Recommenda-
tions section, rather than its current location in the High Activity and Hot Climate
section. The table citations for pregnant and lactating women, high activity and active
adults appear to be wrong
High activity data in the Recommendations does not appear to come from the cited
"Key Studies." This is confusing to the reader because the summary table cites a table
(which seems to be wrong) and not the study.
The Summary table would be improved by:
a) Retaining the age group/population category, eliminating the mean
values, and using exposure guidance percentiles of 50, 90, and 99.9
b) Addressing multiple percentiles and fitted distribution items with foot-
notes, and citing the relevant study directly in its own column instead of
the indirect table reference.
In general, provide table numbers for tables noted in the recommendation section
Specific Comments and Corrections for Chapter 4 (Soil ingestion/Pica)
Overall, the Agency has performed a good evaluation of the available literature on the
highly controversial topic of incidental soil ingestion.
Chapter 4 is consistent with the 1992 Exposure Guidelines. It provides recommenda-
tions for mean and high end soil intake for children (<6 years of age), and mean values
for adults and pica behavior. Given the many limitations of the extant data, detailed
and reliable distributions of intake cannot be derived at this time and this precludes full
consistency with the Exposure Guidelines. The term "upper percentile" exposure
should be more explicitly defined in a manner consistent with the guidelines (i.e.,
should this be considered as a 98th percentile, or a bounding estimate?). The avail-
able information on purposeful ingestion by children, and incidental ingestion by adults
is even more limited. In spite of the limitations of the studies and, as the Chapter
A-4
-------
indicates, the data presented are rather consistent in terms of average intakes across
studies. Since there are a number of on-going investigations trying to characterize
incidental soil intake specially by children (for example, the National Human Exposure
Assessment Survey (NHEXAS)), it may useful for the Agency to coordinate with these
investigators to try to fill the data gaps in this area. As most of the studies reviewed
indicate, there is a large variability in incidental soil ingestion among children and it is
important to know the factors which affect this variability so they can be included in
future revisions of the Handbook. The Agency might consider adding some resources
to on-going studies to investigate these factors.
The data are presented in a manner useful to exposure assessors. A paragraph should
be added, however, to the effect that the recommended values should be used with
caution. Incidental soil intake can occur not only via direct ingestion through hand-to-
mouth or object-to-mouth contact, but also through the contamination of foods inside
the home. There has been some preliminary investigation of micro-environmental
contamination of foodstuffs once brought inside the household (these data were
presented by Linda Sheldon of RTI at the last Exposure Assessment meeting in New
Orleans). This factor, in addition to the large variability in intake estimates among
individual subjects in the studies reviewed, suggests that assessors should consider
each specific application carefully in order to avoid underestimating intake for specific
cases.
The grouping of key and relevant studies is appropriate, although the order in which
they are presented within the key or relevant groupings is unclear, neither following an
alphabetical or time-of-publication pattern (this is also true of the other Chapters), or a
hierarchical one. This may be confusing to the users of the Handbook. A presentation
of the studies by date of publication may be the most appropriate.
The citations to foreign studies (e.g., the Dutch investigations in child care centers)
should be viewed with caution. One of the factors that could be instrumental in
affecting soil intakes is hygiene (e.g., frequent washing of hands). Hygienic practices
can vary across countries and cultures and may be more stringently emphasized in the
highly structured environment of Dutch or other European countries child care centers
than among children in the US (the reverse is true for other countries). Also, note that
the description of the Stanek and Calabrese, 1995a, study on page 4-1 contains an
error: the time period for day 1 of the fecal sample was noon (not midnight) on Monday
to noon on Tuesday.
In general, the document addresses the limitations of the available data. A sentence
A-5
-------
should be added, however, about the limited information available on factors which may
affect individual soil intake and could help explain the observed inter-subject variability
and identify situations that are conducive to higher intakes, indicating that this is an
area where more research is needed.
Specific Comments and Suggestions for Chapter 5 (Inhalation Route)
Chapter 5 provides a comprehensive listing of information on inhalation. The relation-
ship between inhalation and dose of toxic agents to the lung is complex, varying with
route and ventilation rate, and the physical and chemical characteristics of the inhaled
material. The document gives insufficient attention to the need to conceptualize
inhaled dose rather than simply the concentration at which exposure is received. In
this regard, there is a lack of consistency with the Agency's Exposure Guidelines. At a
minimum, the introduction to the chapter should cover the complexity of the relation-
ship between exposure and dose. Example cases might be given, such as a soluble
gas like sulfur dioxide which is removed in the upper airway, and radon progeny,
inhaled in the form of very small particles that are deposited in the airways. Reference
should also be given to standard lung models.
A more focused treatment is needed of the inhalation route. The respiratory tract has a
number of physical zones, the upper airway, the lower airways-subdivided into the
bronchi, and bronchioles, and the gas exchanging region of the lung, the alveoli. There
are diverse target cells within the lung. A short paragraph reminding the reader of the
complexity of the architecture of the lung should be included.
The population includes a substantial proportion of persons with abnormal lungs,
including cigarette smokers, persons with asthma, and persons with chronic obstructive
pulmonary disease. These conditions affect the exposure-dose relation and make
some individuals susceptible. The significance of lung disease in conceptualizing lung
dose should be covered.
With regard to the overall approach and the selection of the reference materials, there
are a few gaps. Relevant work by Samet and colleagues in New Mexico and McCool at
Brown University was not covered. A report of the Health Effects Institute provides
these data which address the ventilation-heart rate relationship.
A-6
-------
Page 5-1: The text makes the statement that alveolar ventilation rate is of particular
interest. In fact, the inhaled volume of air (minute ventilation) is the relevant quantity.
Air inhaled into the physiologic dead space of the lung contains contaminants. This
paragraph needs to be corrected.
Page 5-8, paragraph beginning "A limitation of this study...": This sentence may be
incorrect. The wider variety of exercise would simply extend the range over which the
heart rate to ventilation ratio would be observed.
Page 5-11, second paragraph: The Committee suggests not using the term "macho
effect."
p. 5-12, column 2, paragraph 3 - The California Air Resources Board sponsored
research .... GARB did not conduct the research, but rather sponsored the research by
Adams (1993). Also, the reference should be Adams, 1993, NOT GARB, 1993, and the
text in this section should be corrected to reflect this change. This should be: Adams,
William C. (1993) Measurement of Breathing Rate and Volume in Routinely Performed
Daily Activities, Final Report. California Air Resources Board Contract No. A033-205.
Page 5-21, Table 5-22: In this table, and throughout the document, the term 'confi-
dence rating" is used in reference to literature evaluation. The Handbook reads as
though these confidence ratings were an indication of the degree of certainty. The
Committee interprets the ratings in a different fashion; i.e., that they offer a measure of
the quality of the source information. In the introduction and throughout the text, the
authors should revisit the use of this term, particularly as they take on the difficult task
of reassessing the handling of uncertainty throughout the report.
p.5-22, Table 5-23, Summary of Recommended Values for Inhalation: It is not clear
why children from 1-12 years old are combined in this table since there is a substantial
change in inhalation rate as a function of age. A more appropriate grouping might be 1
to 5 and 6 to 12 years.
Specific Comments and Suggestions for Chapter 6 (Dermal Route)
The dosage terms are inconsistent. The terms "dermal dose" and "absorbed dose"
which are used in the Exposure Factors Handbook (eq 6-1 and 6-2) are not found in
the relevant sections of the Exposure Guidelines. The Exposure Guidelines' dose
terms (fig 2-1) for the dermal route are: potential dose, applied dose, internal dose and
biologically effective dose. The term "applied dose" in the Handbook is probably not
A-7
-------
the same as the "applied dose" in the exposure guidance but actually the "potential
dose."
Two formulas (Equations 6-1 and 6-2) are discussed in Section 6.1 (pages 6-1 and 6-
2). Equation 6.1 predicts dermal absorption of contaminants in water; equation 6.2 is
identified as a variant of 6.1 for estimating dermal absorption from soil. The second
paragraph on page 6-2 states that the soil equation is simpler because it does not have
to deal with an infinite contaminant sources, as does the water-borne exposure
scenario. It appears, however, that equations 6.1 and 6.2 are computationally identi-
cal. The term "EV" (events per day) appears in equation 6.1 but not in 6.2, and the
term "EF" has units of "days/years" in 6.1 and "events/year" in 6.2. When the mathe-
matical operations dictated by the two equations are carried out, however, the results
are the same - absorbed dose per event is multiplied by the number of events. It
appears that either equation could be used for both water and soil exposures (with an
explanation of the term "DA" which incorporates the actual difference in soil vs. water-
borne contaminant uptake).
The Handbook's recommendations summary table lists the 50th and 95 percentile, but
should refer to the 50th, 90th and 99.9th percentile to be consistent with the Exposure
Guidelines. Finally, the text refers to these values at one point as a 90th percentile,
and at another point as a 95th percentile.
Specific Comments and Suggestions for Chapter 7 (Body Weight Studies)
Among the many factors reported in the Handbook, body weight is considered one of
the factors with the least uncertainty. However, the data were collected in 1976-1980,
16 to 20 years ago. Only when the average body weight and the demographic compo-
sition of our population have not changed significantly since the data were collected,
can the average body weight be used as an unbiased estimate for the current national
population. Otherwise, the average and distribution of body weight should be adjusted
based on the current population data. Consequently, the Agency should start consider-
ing the data collected during NHANES III, which would have more recent body weight
information, and would also provide data for Hispanics.
Body weight is a function of age, gender, and race. The populations of many geo-
graphic regions are very different from the national population. The users should be
warned that when applying the percentiles to other geographic regions with different
age, gender and race composition, appropriate adjustment should be made.
A-8
-------
Strictly speaking, the rating of "currency" is not consistent with the rules of consider-
ations in Chapter 1. Based on the rules, currency should be rated as low, if data were
collected prior to 1980. But the rating of currency cited the year of publication (1987),
rather then the years of collection, and ranked currency as medium.
Body weight is one of the key factors for the calculation of dose. It is widely used to
calculate the average daily dose for the national population as well as various
subpopulations. In this chapter, values for adults and children are recommended
separately, and no body weight is formally recommended for the entire population that
includes both adults and children. In other parts of the Handbook, however, 60 kg is
recommended for the whole population. This Chapter on body weight should make a
formal recommendation for users. Body weight data for infants younger than 6 months
are not provided. If reliable data can be obtained, body weights of this age category
should be included. Among various ethnic groups, only data for whites and blacks are
given. If it is feasible, data for other ethnic groups should be provided also.
Specific Comments and Suggestions for Chapter 8 (Lifetime)
The chapter provides life expectancy data for the years 1970-1993 by gender and
black/white ethnicity. These data were derived from the Department of Commerce
Statistical Abstracts and should be representative for the populations included.
Although this information dates to 1993, and life expectancy is increasing, this increase
would probably be only a few months, as shown by the 1995 projection included in the
table. As with the body weight data, other ethnic groups are not included and their life
expectancy may be at variance with those presented in these tables. An effort should
be made to include data for these groups in future revisions.
Also, as part of a future revision, the Agency might consider providing information on
life expectancy, not only at birth, but also according to current age. In this manner, an
assessor can derive exposure estimates based on the age distribution of a particular
population.
Specific Comments and Suggestions for Chapter 9 (Intake of fruits/vegetables)
In this chapter, USEPA has decided to combine three years of relatively current data
(1989-1991 from the Continuing Survey of Food Intakes by Individuals (CSFII) as a
basis for their recommendations on fruits and vegetables. This approach provides a
more stable base upon which short term and long term consumption patterns may be
based. Use of these data are more relevant than either the 1977-78 and 1987-88
A-9
-------
Nationwide Food Consumption Survey (MFCS) because these data are relatively old
and based on a poor response rate (37%).
In spite of using the 1989-1991 data base, it is surprising that in Table 9-5 there is no
specific entry for relatively common fruits (e.g., oranges and bananas) and vegetables
(e.g., green beans) in the tables, while there are some less common ones that do have
specific entries. In Table 9-8 there may be a data entry or format error. For example,
the first column contains a two line entry for "cow peas, field peas, black-eyed peas,"
presumed to be one entry, yet in subsequent columns in this entry there are double
entries. In Table 9-10, there is a possible error in column three, first entry. It states
that 86.8 percent of the population use (consume ?) fruit per day, yet in all subsequent
entries in that column, no demographic unit listed has a consumption rate of 70 percent
and 14 entries are less than 60 percent. In Table 9-18, there may be a registration
error in column 2, non-citrus fruits.
Specific Comments and Suggestions for Chapter 10 (Intake of Fish/shellfish)
There are two summary tables in Recommendations-General Population section (RGP)
and Recommendations-General Population- Fish Serving Size (RFSS).
Section 10.10.1 (Recommendations - General Population) contains a discussion on
pages 49-50, on changes of mean intake of 20.1 g per day based on the 1989 CSFII
study and compares that to the 13.5 g per day in the TRI study. Acknowledging that
the CSFII short term data cannot be used to extrapolate to long term average daily fish
intake, the report states that because the TRI data are over 20 years old, it was "felt"
that these data could be "adjusted" to account for the recent increase in fish consump-
tion. Therefore, the report says, the TRI should be "shifted upward by 50 percent".
Using this approach the authors recommended percentiles of long-term average daily
fish consumption as those of Javitz (1980) adjusted 50 percent upward. This is not an
acceptable rationale. In a technical and science-based document this change should
be supported by relevant data and studies.
In general, the summary tables are not well explained in the recommendations section
of the text. They should be given numbers and referred to directly with an unambigu-
ous explanation of the numbers and where they came from. It is possible in most cases
to reconstruct this information by referring to the numbered tables in the body of the
text, but it is a time-consuming process for the user.
A-10
-------
Acknowledging that EPA assigned the most relevance to the 1989-1991 USDA "Key"
study for its recommendations, the data base, includes a "Key" study conducted by the
Tuna Research Institute (TRI) in 1973-74, reflecting a primary concern on key studies -
it is too old to qualify as a key study considering the changes that have occurred in
production (e.g., aquaculture) and consumption that would tend, at the least, to suggest
this study be a "Relevant" study in spite of its size and one year duration for collection
of the data. Studies such as the CSFII, although much more current, suffer by having
data provided on 3 consecutive days. In addition to requiring respondents to estimate
serving sizes, it is hard to extrapolate the data to "long term" exposure. Somewhat
troubling, though not unexpected, is the variability of results from different studies
reviewed in the early portion of the chapter. As an example, the Javitz (1980) study
estimates high mean consumption rates, exceeding similar estimates for anglers,
possibly reflecting bias due to the 3-day recall period method of data collection. It may
highlight a particular concern with this chapter because it addresses a food commodity
that is, by the vast majority of consumers, eaten only intermittently (although when
consumed, it may be in portions greater than 100 g). This offers great opportunity to
misrepresent consumption by U.S. consumers, and hence, exposure factors. In
addition, there is not a clear distinction, whether it makes a difference or not, of the fish
or shellfish being of domestic or international origin. This may be relevant to this
chapter because of the uniqueness of the fish industry - the U.S. is among the largest
exporters and largest importers of edible fish and shellfish in the world. It would be
constructive to consider some discussion on this point.
The large and complex tables in this chapter should be reviewed for consistency of
units (e.g., see Tables 10-8 through 10-12). Consistency of units may reduce possible
errors of interpretation by risk assessors.
The discussion sections on recreational anglers is extremely extensive and seems out
of proportion to the total estimated contribution of daily consumption. However, the
Committee acknowledges that there is concern for subsistence anglers and susceptible
populations that may not be properly recognized in exposure assessments. EPA may
wish to consider consolidating or streamlining this portion of the report.
In section 10.9 (Other Factors), there is an extensive tabulation of moisture and fat
composition data. With it is the assumption that "the residue levels of contaminants in
fish are reported as the concentration of contaminant per gram of fat." This is true only
for lipophilic contaminants. However, there are other veterinary medical products
relatively often used in aquaculture that are not lipid soluble. Specific examples are
some of the antibiotics such as the tetracyclines. This consideration is not addressed
A-11
-------
in the narrative of this section. The Codex Committee on Residues of Veterinary Drugs
in Food for example, has recommended an maximum residue limit for oxytetracycline in
giant prawns and principles have been elaborated for use of muscle tissue as the target
tissue for other maximum residue limits. The discussion, therefore, needs to be
broadened and more inclusive.
The Tables 10-50 through 10-53 do not seem to reflect the confidence ratings in Table
1-2. An explanation of the ratings in the tables in the recommendation sections of the
individual chapters would be helpful to justify the entries in Table 1-2.
Specific Comments and Suggestions for Chapter 11 (Intake of Meat/dairy Prod-
ucts)
The issue of "Key" and "Relevant" studies is a critical issue in Chapter 11. Extensive
reference is made to the 1977-78 MFCS and to the 1989-91 CSFII study. These data
are insufficient to recommend exposure factors as noted by the following USDA
publication "Meat and Poultry Inspection - 1994 Report of the Secretary of Agriculture
to the U.S. Congress"(USDA, March 1995). Between 1984 and 1994, using 1984 meat
and poultry production (animals presented for inspection) as a reference point, cattle
production increased less than one percent; market hog production increased about 10
percent; chicken production increased approximately 70 percent; turkey production
increased approximately 75 percent; total poultry production increased approximately
67 percent. These data clearly indicate the dynamic changes in consumption patterns
by consumers and the inadequacy of relying on "old" data. In addition, consumption
patterns have changed as has the introduction of a significant number of new food
products, including "ready to eat" entries and several others. Indicative of the changing
technology is the significant number of meat and poultry product labels approved by
USDA in 1994 (about 180,000). Considering that there are many label changes that
are for non-technical reasons such as a label for a different product size, for example,
there are substantial numbers of new food products being introduced in response to
consumer demand and developments in food technology. Though it may be difficult for
EPA to use a reference point for the EFH, a strong recommendation is made to use the
most current published data.
EPA should reconsider the discussion in Section 11.2 (Fat Content of Meat and Dairy
Products) and limit the discussion accordingly. The section begins by stating that "in
some cases, the residue level of contaminants in meat and dairy products are reported
as the concentration of contaminant per gram of fat." The entire section is written on
the assumption that all chemical contaminants in meat and dairy products are in fat or
A-12
-------
lipid tissue. This is not a correct assumption for many food animal health products.
This is valid for the fat and lipid soluble pesticides, but is not applicable to most
veterinary drugs that are approved for use in food producing animals. With these
products, organ tissue (liver and kidney) or muscle tissue may be the target tissue. Of
a secondary nature, the discussion on poultry fat does not indicate whether it does or
does not include adhering skin.
EPA should reconsider the discussion in Section 11.3 (Conversion Between As
Consumed and Dry Weight Intake Rates). Although this is a short section, it has little
value for exposure assessment considerations of chemical contaminants, pesticides or
veterinary drug residues in meat products because none of the EPA pesticide or FDA
veterinary drug residues are reported on a dry weight basis for determining compliance
with a U.S. food safety tolerance.
The data in Table 1-2 need some clarification. Whereas the data from Chapter 11
(Total meat intake rate) recommend an exposure factor of 2.1 g per kilogram-day, while
the data from Home Produced Food Intake suggests total meat of 2.2 g per kilogram-
day. The latter figure reflects only those that consume home produced foods.
EPA should also consider food exposure consumption factors based on the individual
animal and bird slaughter classes identified by USDA. The reason for the recommen-
dation is that historical residue data and knowledge of current production practices
indicate significant differences in residue contamination profiles. For exposure/risk
assessments, this may be pertinent information.
Specific Comments and Suggestions for Chapter 12 (Intake Rates for Home-
produced Food Products)
This chapter relied on the NCFS 1977-78 and 1987-88 studies for its recommenda-
tions, which limits its value, as noted previously. It is not clear that the equations 12-2
and 12-3 achieve their intended objective and there do not appear to be entries in the
tables that include the data identified in components of the above noted equations.
Perhaps some additional review and clarification is needed on this section. The 50
percentile summaries in Tables 12-8, 13, 18, 23, and 28 do not seem to agree with the
entries in Table 1-2. Some explanation may be necessary or a footnote in the summary
table. In some cases, the entries for home grown products in Table 1-2 are higher than
the exposure factor for total fruits, vegetables and meat listed from assessments of
these respective commodity chapters.
A-13
-------
Specific Comments and Suggestions for Chapter 13 (Breast Milk Intake)
In general, the Chapter is consistent with the guidelines in that it provides average and
upper-percentile estimates of breast milk intake for children from 1 month up to 1-year
of age, and it recognizes that there is socio-demographic and ethnic variability in the
proportion of and extent to which children are breast fed. The chapter, however, does
not issue any specific recommendation for infants under 1 month of age (when a larger
number of infants are totally breast fed and the intake, on a body weight basis, may be
also larger than for older infants, although the net volume ingested may be smaller).
With the exception of the DARLING study (Dewey, K. G., et. al, in Pediatrics 87:828-
837, and in J. Pediatrics (119:538-547), which has a strong population sample bias,
most of the studies were published almost 10 years ago or earlier (with the data having
been collected even earlier), and have other limitations such as sample size. There
may be some more recent data available from the FDA or other state agencies, that
should be included in a follow-up revision of the Handbook. As with other exposures,
there may be special groups with potentially high exposures that are not captured in
these data. In this case, these may be groups of women highly committed to breast
feeding exclusively for periods of time longer than one year; these groups should be
identified. This may also be true of subgroups on the other end of the distribution, i.e.,
women who may not breast-feed or do so minimally.
The section on breast milk intake as a function of energy needs/consumption could be
improved by adding a sentence on the uses of this information to estimate exposures of
breast milk intake (this is not included in the recommendations). The practical rele-
vance of this information to the exposure assessor is not clearly stated in the chapter.
Specific Comments and Suggestions for Chapter 14 (Activity Factors)
Many of the tables are not clearly labeled; they could also benefit from some editing to
remove unnecessary symbols for groups. In addition, since there is blank space on
many of these pages, the print on some of these tables could be enlarged to make
them more easily readable.
p. 14-1, column 1, paragraph 2, line 6 - should be "micro"environments; also, the term
"micro environments" should be explicitly defined.
p. 14-2, column 1, last paragraph - the reference to GARB is incorrect. This should be
Wiley, J. A., J. P. Robinson, T. Piazza, K. Garrett, K. , Cirksena, Y.T. Chang and G.
Martin. (1991) Activity Patterns of California Residents. Final Report, Survey Research
A-14
-------
Center, University of California, Berkeley, CA. California Air Resources Board Contract
No. A6-177-33.
p. 14-3, column 1, paragraph 3 - Should the mean or geometric mean be presented for
modeling purposes?
p. 14-3, column 1, paragraph 3 - Were these statistically significant differences?
p. 14-3, column 1, last paragraph, last sentence - Suggest that these activities be listed
in order of time.
p. 14-3, second column, paragraph 2, on limitations." This may somewhat bias the
GARB data set" - presumably what is meant here is "..with respect to its application to
other populations."
p. 14-4, column 2, paragraph 2, line 6 - suggest that some indication of what short-term
recall is be provided, e.g., 24-hour?
p. 14-9, column 2, paragraph 2, lines 13-16 - Playground/park data are also available
in:
Phillips, T. J., P. L. Jenkins, and E. J. Mulberg. (1991) Children in California:
Activity patterns and presence of pollutant sources. Paper No. 91 -172.5,
Proceedings of the 84th Annual Meeting of the Air and Waste Management
Association, Vancouver, B.C.
Timmer, S.G., J. Eccles and K. O'Brien. (1985) How children use time, pp.
353-382, In: Time, Goods and Well-Being, F. T. Juster and F. P. Stafford, Eds.,
Institute for Social Research, University of Michigan, Ann Arbor, Ml.
Medrich, E. A., J. Roizen, V. Rubin, and S. Buckley. The Serious Business of
Growing Up: A Study of Children's Lives Outside School, University of California
Press, Berkeley, CA, 1982.
Michelson, W. M. From Sun to Sun: Daily Obligations and Community Structure
in the Lives of Employed Women and Their Families. Rowman and Allanheld,
Totowa, N J, 1985.
p. 14-14, column 2, paragraph 2 (Section 14.4.1) - Cooking should also be included
A-15
-------
here based on the results of the PTEAM study and results reported by Brauer,
M.(1995) /Assessment of indoor aerosols with an integrating nephelometer. Journal of
Exposure Analysis and Environmental Epidemiology, 5: 45-56.) on exposures to
particles from cooking. More recent results on exposures during cooking were also
reported by Brauer at the recent ISEA meeting in New Orleans. There are likely to be
exposures to other pollutants during cooking. Even if the specific pollutants (other than
particles) have not yet been determined, it is worth anticipating the need for this data.
p. 14-15, column 1, paragraph 2 - "Time spent gardening." This may be retrievable
from the GARB 1991 study by using a crosstab of "plant care" with location codes for
outside.
p. 14-15, last two paragraphs and p. 14-16, column 1, paragraph 1 - Suggest including
95th and 99.9th percentiles for consistency with the Exposure Assessment Guidelines.
Specific Comments and Suggestions for Chapter 15 (Consumer Products)
It should be noted at the outset that this information is a very useful addition to the
Handbook!
In many of the tables at the end of the chapter, however, some clarification of the time
period needs to be provided. For example, Table 14-35 provides data on the number
of time an automobile or motor vehicle was started, but there is no indication of whether
this is per day or per week. Also the print in many of these tables is very, very small
and difficult to read. In most cases, there is sufficient space on the page so that the
type could be made larger and more readable.
Specific Comments and Suggestions for Chapter 16 (Reference Residence)
This chapter is a very useful addition to the Exposure Factors Handbook and makes
data on volumes of residential buildings much more readily accessible for exposure
assessment than it is in the REGS database. Figure 16-1 captures many of the key
features of residential buildings which must be considered in modeling indoor air
exposures, and in general, the background discussions on the material in this chapter
are well written and provide a good explanation of the nature and use of the data in the
chapter.
Overall, the chapter is generally consistent with the Exposure Guidelines and provides
valuable distributional data for estimation of exposures. The title of the chapter,
A-16
-------
however, is not consistent with the distributional nature of the data presented in the
Handbook. "Reference Residence" implies that a single house is representative of U.S.
housing. A more appropriate title would be "Reference Residences" or, more simply
and directly, "Residential Building Characteristics."
There are two key sets of data in this chapter - distributions of residential building
volumes and distributions of air exchange rates. Although the presentation of volume
data by geographical regions is appropriate, the air exchange rates should not be
grouped by zones that are geographical, since these are more non-homogeneous
statistically than is necessary or appropriate. Air exchange rates (infiltration and
natural ventilation by window opening) as well as building characteristics related to air
exchange rates, are determined by climate zones and meteorology. It would be more
appropriate to present the air exchange rates grouped by climate zones such as those
defined by EPA's BASE study. This would make the data easier to use and would
emphasize the physical determinants of air exchange rates, i.e., meteorology and
housing type. It should also be noted that the age of housing can be a significant
determinant of air exchange rates. Newer, more energy-efficient homes are much
"tighter" and therefore have lower infiltration rates (see Sherman and Dickerhoff, 1996,
enclosed).
There should also be some consideration of the addition of a table presenting house
volumes grouped by climate zone, which could be used with a climate-zone based
table on air exchange rates.
It may not be appropriate to classify any of the studies on air exchange rates as "key
studies" since this classification seems to imply a representativeness and accuracy that
these data do not have. If the sample size for any given climate zone or season is
small, it is much less likely to be representative and weighting does not correct for this.
The subsets of data on air exchange rates should be checked to see how many
measurements there actually are for a subset.
With respect to section 16.3.5, (Interzonal Airflows), it is strongly recommended that
section 16.3.5.2 be completely eliminated. It is highly phenomenological rather than
scientific and it is very likely to give incorrect answers because of the lack of represen-
tativeness of U.S. residences and conditions. More scientifically accurate interzonal
airflow models have been developed and these should be used for situations in which
such modeling is required.
A-17
-------
The tables which presents the limitations and uncertainties of the data in each Hand-
book chapter are well designed and useful, although there needs to be some clarifica-
tion on how things were weighted to arrive at the overall rating. In the chapter on
residential building characteristics, however, the table on air exchange rates (Table
16-31) does not capture all of the uncertainties associated with air exchange rate
measurements and also has some inappropriate rationales listed.
The PFT measurements of air exchange rates assume uniform mixing of the tracer
within the building. This is not always so easily achieved. Furthermore, the degree of
mixing can vary from day to day and house to house because of the nature of the
factors controlling mixing, e.g., convective air motion, driven by weather, the type and
operation of the heating system. The relative placement of the PFT source and the
sampler can also cause variability and uncertainty. It should be noted that sampling is
typically done in a single location in a house which may not represent the average for
that house. In addition, very high and very low values of air exchange rates based on
PFT measurements have greater uncertainties than those in the middle of the distribu-
tion. The low ones are more uncertain because of analysis uncertainties, while the
high air exchange rates can be very strongly biased by imperfect mixing that is more
common under such conditions. The limitations in the representativeness of the data
with respect to climate zones are also reason for concern. Given the uncertainties in
the PFT (tracer) measurement method for air exchange rates and, more significantly,
the lack of representativeness for the U.S. housing database, it is not clear why the
overall rating is medium. It would probably be more appropriate to give these data a
"low" overall rating and to provide some information to the Handbook user on the
specific limitations of the PFT ventilation measurements. This doesn't mean that they
are not useful. It does mean that the user should be cautious about interpretation of
results based on such data, particularly if data from under-represented climate zones
are used.
With respect to the rationales for ratings for "Reproducibility," the rationale is that the
methodology was clearly presented. This is not the same as reproducibility. If there is
no information on reproducibility, e.g., duplicate or triplicate measurements made in a
house over the same time period, then the table should so indicate. Similarly, it is not
appropriate to characterize "Validity of Approach" by saying that the method is practi-
cal. Validity of approach must address the underlying physics of the situation and the
suitability of the measurement with respect to the underlying physics. In the case of the
PFT measurements, the assumption of perfect mixing and the placement of the PFT
source and the sampler, as well as the sampling and analysis uncertainties, need to be
considered.
A-18
-------
Other, more specific comments and corrections noted for Chapter 16:
p. 16-1, column 2, line 1 - "deposition to and re-emissions from ..."
p. 16-2, column 2, paragraph 3, line 8, "The REGS also provides... volumes .."
should be "floor areas" rather than "volumes"
p. 16-6, Table 16-5 - It is not clear how the surface covered would be useful for
the first 6 items; shouldn't these be given as mass applied per m2?
p. 16-6, The discussion on mechanical system configurations is well done and
useful.
p. 16-8, 16-9 - It would be appropriate to point out that the EPA regions and
census regions are not appropriate for statistical sampling to characterize air
exchange rates or housing characteristics and that the climate zones designated
in the EPA BASE study are more appropriate for such purposes.
p. 16-8, in column 2, under "Local circulation" , it should read "Convective and
advective ..."
p. 16-10, column 2 line 1 - delete "because outdoor levels are generally as-
sumed to be zero." This assumption is not necessary and is frequently not
made. Air exchange still provides the major mechanism for removal of most
indoor air pollutants (via both dilution from infiltration and by exfiltration), so the
remainder of the sentence is still correct.
p. 16-12, column 2, paragraph 1 - The statement that the highest air exchange
rates occur in the warmest climates region during the summer should be
checked to see if this result is simply a limitation in the number of measurements
in the subset of data. In the very warm regions, homeowners tend to use their
air-conditioners in the summer months and to keep homes closed up for energy
efficient use of air-conditioners.
p. 16-13, column 1, end of paragraph 2 - the last sentence is not exactly correct.
Leakage areas are inferred from an infiltration model developed by Sherman and
Grimsrud ("The measurement of infiltration using fan pressurization and weather
data, " Proceedings of the First International Air Infiltration Centre Conference,
London, England. Also, Lawrence Berkeley Laboratory Report LBL-10852,
A-19
-------
October, 1980). Blower door measurements made when houses are
weatherized can be used with a model to infer the leakage areas and, these can
be used with meteorological data to estimate infiltration rates for residences.
See Sherman and Dickerhoff, ("Air-tightness of U.S. dwellings," In: The Role of
Ventilation, 15th AIVC Conference, Buxton, Great Britain, 27-30 September
1994; and "Air-tightness of U.S. dwellings," Lawrence Berkeley Laboratory
Report No. LBL-35700, 1996; both enclosed). The latter provides a more
detailed explanation of how to estimate infiltration rates from blower door
measurements. This sentence should be modified to indicate how infiltration
models can be used with blower door measurements to estimate infiltration rates
since there is a large body of blower door measurements available from state
weatherization programs; many of these data have been assembled into a
database by M. Sherman of the Lawrence Berkeley National Laboratory..
p. 16-13, column 2 - This deals with the very large impact of window opening on
ventilation rates. This is an important point. We suggest giving a numerical
example here, e.g., 0.2 h"1 to 10 h"1, and pointing out that measured values of
very high air exchange rates are considerably less reliable and meaningful
because mixing is typically much less complete under high ventilation
conditions.
p. 16-15 - Interzonal Flows - See comments above recommended deletion of
section 16.3.5.2.
p. 16-21, Table 16-26 -What is "totalized?" I think this should be "Total." Also,
the units for dust load should be clarified with a hyphen between g and m-2 so
that this doesn't look like grams -2.
p. 16-22, column 1, last paragraph - "Emanation sources..." Also, line 3, second
column. This term is used for radon emanation from radium-bearing soils and
rocks, but not for solid sources of vapor-phase organics. It may confuse
readers. We suggest using the term "solid sources" and including examples,
e.g., carpet backing, furniture, flooring, dried paint.
p. 16-22, column 2, line 4 - "latch" should be "relate" or "link."
p. 16-23, Equation 16-6 seems out of place. It probably was meant to be placed
somewhere near the end of column 2, p. 16-22. It will have to be re-numbered if
placed before 16-5.
A-20
-------
p. 16-24, Equation 16-9 - the symbol "A" for area has been omitted.
Specific Comments and Suggestions for Chapter 17 (Grain Consumption)
The same general comments apply to this food group on the timeliness of the
data. The data used in this chapter for exposure factors are, to a great extent, too old
for this assessment. More current data are necessary for determining exposure factors.
Data from more contemporary studies are available from USDA as noted above.
A-21
-------
NOTICE
This report has been written as a part of the activities of the Science Advisory
Board, a public advisory group providing extramural scientific information and advice to
the Administrator and other officials of the Environmental Protection Agency. The
Board is structured to provide balanced, expert assessment of scientific matters
relating to problems facing the Agency. This report has not been reviewed for approval
by the Agency and, therefore, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency, 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.
-------
U.S. Environmental Protection Agency
Science Advisory Board
Integrated Human Exposure Committee
December 19-20, 1996
Chair
Dr. Joan Daisey, Lawrence Berkeley Laboratory, Berkeley, CA
Members
Dr. Paul Bailey, Mobil Business Resources Corporation, Paulsboro, NJ
Dr. Robert Hazen, State of New Jersey Department of Environmental Protection and
Energy, Trenton, NJ
Dr. Timothy Larson, Department of Civil Engineering, University of Washington,
Seattle, WA
Dr. Kai-Shen Liu, California Department of Health Services, Berkeley, CA
Dr. Paul Lioy, Environmental and Occupational Health Sciences Institute, Piscataway,
NJ
Dr. Maria Morandi, University of Texas School of Public Health, Houston, TX
Dr. Jonathan M. Samet, The Johns Hopkins University, Baltimore, MD
Mr. Ron White, American Lung Association, Washington, DC
Dr. Lauren Zeise, California Environmental Protection Agency, Berkeley, CA
Federal Experts
Dr. Richard Ellis, US Department of Agriculture, Washington, DC
Ms. Alanna J. Moshfegh, U.S. Department of Agriculture, Riverdale, MD
Science Advisory Board Staff
Mr. Samuel Rondberg, U.S. Environmental Protection Agency, Science Advisory Board,
Washington, DC 20460
Staff Secretary
Mrs. Dorothy M. Clark, U.S. Environmental Protection Agency, Science Advisory Board,
Washington, DC 20460
-------
Distribution List
Administrator
Deputy Administrator
Assistant Administrators
Deputy Assistant Administrator for Pesticides and Toxic Substances
Deputy Assistant Administrator for Research and Development
Deputy Assistant Administrator for Water
Deputy Assistant Administrator for Air and Radiation
EPA Regional Administrators
EPA Laboratory Directors
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
Staff Director, Scientific Advisory Panel
------- |