United States
Environmental Protection
Agency
H U M A
H E A L T
«<^5
RESEARCH STRATEGY
RESEARCH AND DEVELOPMENT
-------
-------
EPA/600/R-02/050
September 2003
Human Health Research Strategy
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
Disclaimer
This document has been subjected to review by the National Health and Environmental
Effects Research Laboratory and the Science Advisory Board and has been approved for
publication. It does not constitute an EPA position or policy concerning human health risk
assessment research. Any mention of trade names does not constitute EPA endorsement.
-------
FOREWORD
The 2003 EPA Human Health Research Strategy identifies and prioritizes research
needed to improve the scientific foundation for health risk assessments. The Human Health
Research Strategy identifies two overall strategic objectives: research to improve the scientific
foundation of human health risk assessment and research to enable evaluation of public health
outcomes from risk management decisions. The Strategy provides a conceptual framework for
future human health research by the Office of Research and Development (ORD) and was
prepared by a team of scientists from ORD with input from EPA's Office of Prevention,
Pesticides, and Toxic Substances, the Office of Water, and the Office of Children's Health
Protection.
The Human Health Research Strategy focuses on developing a multidisciplinary,
integrated program to improve linkages between exposure, dose, effect, and risk assessment
methods. ORD efforts in computational toxicology, predicting aggregate and cumulative risk
(exposure to mixtures of pollutants from multiple sources), protecting susceptible subpopulations
(i.e., children, older adults) and accountability reflect the timeliness of this plan.
This research plan is an important planning and accountability tool because it makes clear
the rationale for and intended products of EPA's human health research program. This research
strategy also provides the basis for the development of a multi-year plan on human health
research, which outlines specific research goals and measures to be accomplished over the next
5-10 years.
Paul Oilman, Ph.D.
Assistant Administrator
in
-------
TABLE OF CONTENTS
Page
FOREWORD iii
LIST OF FIGURES vi
AUTHORS, CONTRIBUTORS, AND REVIEWERS vii
PEER REVIEW HISTORY viii
ABBREVIATIONS AND ACRONYMS ix
GLOSSARY x
EXECUTIVE SUMMARY E-l
1. INTRODUCTION 1-1
1.1 PURPOSE OF THE STRATEGY 1-1
1.2 CURRENT RESEARCH PROGRAM ON HUMAN HEALTH 1-2
1.3 FUTURE RESEARCH PRIORITIES 1-4
1.3.1 Framework for an Integrated Research Program in ORD 1-4
1.3.2 Research Themes 1-8
1.4 STRATEGIC PRINCIPLES 1-9
2. RESEARCH TO IMPROVE THE SCIENTIFIC FOUNDATION OF HUMAN HEALTH
RISK ASSESSMENT 2-1
2.1 RESEARCH ON HARMONIZING RISK ASSESSMENT APPROACHES 2-1
2.1.1. Scientific Uncertainties 2-1
2.1.2 Research Objectives 2-2
2.1.3 Research Approach 2-2
2.2 RESEARCH ON AGGREGATE AND CUMULATIVE RISK 2-6
2.2.1 Scientific Uncertainties 2-6
2.2.2 Research Objectives 2-7
2.2.3 Research Approach 2-7
2.3 RESEARCH ON SUSCEPTIBLE AND HIGHLY-EXPOSED
SUBPOPULATIONS 2-12
2.3.1 Scientific Uncertainties 2-13
2.3.2 Research Objectives 2-14
2.3.3 Research Approach 2-14
3. RESEARCH TO ENABLE EVALUATION OF PUBLIC HEALTH OUTCOMES FROM
RISK MANAGEMENT ACTIONS 3-1
3.1 SCOPE AND DEFINITIONS 3-2
3.2 SCIENTIFIC UNCERTAINTIES 3-4
3.3 SCIENTIFIC OBJECTIVES 3-5
IV
-------
3.4 RESEARCH APPROACH 3-6
3.5 RESEARCH IMPLEMENTATION 3-7
4. REFERENCES 4-1
APPENDICES
APPENDIX A- ORD Research Plans and Strategies A-l
APPENDIX B- EPA's Office of Research and Development-Organizational Chart .... A-3
APPENDIX C- Examples of Mechanistic Data Used in Risk Assessment A-4
APPENDIX D- Agencies Having Research Programs Complementary to ORD A-5
APPENDIX E- Examples of Health and Environmental Databases to Evaluate Public Health
Outcomes From Risk Management Actions A-7
-------
LIST OF FIGURES
Figure Page
1-1 Relationship Between Core and Problem-Driven Research 1-1
1-2 The Risk Assessment-Risk Management Paradigm 1-2
1-3 The Exposure-Dose-Effect Continuum 1-5
1-4 ORD's Approach to Solving Agency Problems 1-6
1-5 ORD's Integrated Research Program 1-7
1-6 Integrated ORD Research Program on Molds and Asthma in Children 1-8
1-7 Role of Research in Assessment of Health Outcomes 1-9
in the Risk Assessment and Risk Management Process
3-1 Role of Analysis of Health Outcomes in the Risk Management Process 3-3
VI
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Executive Lead
Harold Zenick, National Health and Environmental Effects Research Laboratory (NHEERL),
Office of Research and Development (ORD), U.S. Environmental Protection Agency (EPA)
Authors
Hugh Barton, ORD/NHEERL
Jerry Blancato, ORD/National Exposure Research Laboratory (NERL)
Michael Callahan, formerly with ORD/National Center for Environmental Assessment
(NCEA)
Larry Cupitt, ORD/NERL
Judith Graham, formerly with ORD/NERL
Karen Hammerstrom, ORD/NCEA
Jonathan Herrmann, ORD/National Risk Management Research Laboratory (NRMRL)
Robert Kavlock, ORD/NHEERL
Gary Kimmel, ORD/NCEA
Hugh McKinnon, ORD/NRMRL
Hugh Tilson, ORD/NHEERL
Vanessa Vu, formerly with ORD/NCEA
Jennifer Orme-Zavaleta, ORD/NHEERL
Contributors
Linda Birnbaum, ORD/NHEERL
Rebecca Calderon, ORD/NHEERL
Robert Chapman, formerly with
ORD/NCEA
Gary Foureman, ORD/NCEA
Herman Gibb, ORD/NCEA
Annie Jarabek, ORD/NCEA
Carole Kimmel, ORD/NCEA
Suzanne McMaster, ORD/NHEERL
Patricia Murphy, ORD/NCEA
Dale Pahl, ORD/NERL
Julian Preston, ORD/NHEERL
Chris Saint, ORD/National Center for
Environmental Research (NCER)
John Schaum, ORD/NCEA
Linda Sheldon, ORD/NERL
William Steen, formerly with ORD/NERL
Michel Stevens, ORD/NCEA
Reviewers
Tom Barnwell, ORD/NERL
William Farland, ORD/NCEA
Elaine Francis, ORD/NCER
Fred Hauchman, ORD/NHEERL
Steven Hedtke, ORD/NHEERL
Robert Menzer, formerly of ORD/NCER
Lee Mulkey, ORD/NRMRL
Kevin Teichman, ORD/Office of Science Policy
(OSP)
vn
-------
PEER REVIEW
Peer review is an important component of research strategy development. The peer review
history for this research strategy follows:
ORD Science Council: March 21, 2001, Elaine Francis and Kevin Teichman, Lead Reviewers
External Peer Review: Science Advisory Board/Human Health Research Strategy (HHRS)
Review Panel, November 20-21, 2002
External Peer Review Panel Members:
Dr. James Klaunig, Panel Chair, Professor of Toxicology and Director of Toxicology,
Department of Pharmacology and Toxicology, Indiana University School of Medicine,
Indianapolis, IN
Dr. Paul Blanc, Professor of Medicine and Endowed Chair in Occupational and Environmental
Medicine, University of California-San Francisco, San Francisco, CA
Dr. James Gibson, Research Professor of Pharmacology and Toxicology, The Brody School of
Medicine, East Carolina University, Greenville, NC
Dr. Michael Jayjock, Senior Research Fellow, Rohm and Haas Co., Springhouse, PA
Dr. George Lambert, Associate Professor of Pediatrics and Associate Director of the Clinical
Research Center, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson
Medical School, Piscataway, NJ
Dr. Joseph Landolph, Associate Professor of Molecular Pharmacology and Toxicology, School
of Pharmacy, University of Southern California in Los Angeles, Los Angeles, CA
Dr. Steve Lewis, Distinguished Scientific Associate, Exxon Mobil Biomedical Sciences, Inc.,
Annandale, NJ
Dr. Randy Maddalena, Indoor Environment Department Lawrence Berkeley National
Laboratories, Berkeley, CA
Dr. Maria Morandi, Assistant Professor, University of Texas Houston Health Sciences Center,
School of Public Health, Houston, TX
Dr. Beate Ritz, Associate Professor, Department of Epidemiology and Center for Occupational
and Environmental Health, School of Public Health, University of California in Los Angeles,
Los Angeles, CA
Dr. Herbert Rosenkrantz, Professor, Environmental and Occupational Health and Pharmacology,
University of Pittsburgh, Pittsburgh, PA
Dr. Robert Spengler, Associate Administrator, Agency for Toxic Substances and Disease
Registry, Atlanta, GA
Dr. Bernard Weiss, Professor, Environmental Medicine and Pediatrics, University of Rochester
School of Medicine and Dentistry, Rochester, NY
Peer Review Coordinator:
Dr. Sue Shallal, Designated Federal Officer, Environmental Health Committee EPA Science
Advisory Board, Environmental Protection Agency, Washington, DC
Vlll
-------
ABBREVIATIONS AND ACRONYMS
Ah Arylhydrocarbon
ATSDR Agency for Toxic Substance and Disease Registry
BBDR Biologically Based Dose Response Modeling
CCL Contaminants Candidate List
CDC Centers for Disease Control and Prevention
EPA U.S. Environmental Protection Agency
FFDCA Federal Food, Drug, and Cosmetic Act
FIFRA Federal Insecticide, Fungicide, and Rodenticide Act
FQPA Food Quality Protection Act
GPRA Government Performance and Results Act
HUD Housing and Urban Development
IPCS International Programme on Chemical Safety
MOE Margin of Exposure
NAS National Academy of Science
NCEA National Center for Environmental Assessment (EPA/ORD)
NCEH National Center for Environmental Health
NCER National Center for Environmental Research (EPA/ORD)
NCHS National Center for Health Statistics
NCI National Cancer Institute
NCT National Center for Toxicogenomics
NCTR National Center for Toxicological Research
NERL National Exposure Research Laboratory (EPA/ORD)
NHANES National Health and Nutrition Examination Survey
NHEERL National Health and Environmental Effects Research Laboratory (EPA/ORD)
NHEXAS National Human Exposure Assessment Survey
NIAID National Institute of Allergy and Infectious Diseases
NICHD National Institute of Child Health and Human Development
NIEHS National Institute of Environmental Health Sciences
NOAEL No-Observed-Adverse-Effect Level
NRC National Research Council (NAS)
NRMRL National Risk Management Research Laboratory (EPA/ORD)
NTP National Toxicology Program
OPP Office of Pesticide Programs
ORD Office of Research and Development (EPA)
OSP Office of Science Policy
PBPK Physiologically Based Pharmacokinetic Modeling
PK Pharmacokinetic
QSAR Quantitative Structure-Activity Relationship
RfC Reference Concentration
RfD Reference Dose
SAB EPA's Science Advisory Board
STAR EPA/ORD Science to Achieve Results Extramural Grants Program
TSCA Toxic Substances Control Act
TEF Toxicity Equivalent Factor
UF Uncertainty Factor
IX
-------
GLOSSARY
Aggregate Exposure: The combined exposure of an individual or defined population to a
specific agent or stressor via relevant routes, pathways, and sources (working definition
developed by EPA Science Policy Council).
Aggregate Risk: The risk resulting from aggregate exposure to a single agent or stressor
(working definition developed by EPA Science Policy Council).
Biological Markers or Biomarkers: Indicator signaling events in biological systems or
samples. There are three classes of biomarkers: exposure, effect, and susceptibility. A marker
of exposure is an exogenous substance or its metabolite(s) or the product of an interaction
between a xenobiotic agent and some target molecule or cell that is measured in a compartment
within an organism. A marker of effect is a measurable biochemical, physiological, or other
alteration within an organism that, depending on magnitude, can be recognized as an established
or potential health impairment or disease. A marker of susceptibility is an indicator of an
inherent or acquired limitation of an organism's ability to respond to the challenge of exposure
to a specific xenobiotic.
Biologically-Based Dose Response (BBDR) Model: A model that describes biological
processes at the cellular and molecular level linking the target organ dose to the adverse effect.
Childhood: Nominally, the period from birth through the onset of puberty. However, the
Human Health Research Strategy addresses adverse effects on the developing organism that may
result from exposure to environmental agents, starting with preconception exposures of the
parents and continuing through gestation and postnatally up to the time of maturation of all
organ systems.
Cumulative Risk: The combined risks from aggregate exposures to multiple agents or stressors
(working definition developed by EPA Science Policy Council).
Dose: The amount of a substance available for interactions with metabolic processes or
biologically significant receptors after crossing the outer boundary of an organism. The
potential dose is the amount ingested, inhaled, or applied to the skin. The applied dose is the
amount of a substance presented to an absorption barrier and available for absorption (although
not necessarily having crossed the outer boundary of the organism). The absorbed dose is the
amount crossing a specific absorption barrier (e.g., the exchange boundaries of the skin, lung,
and digestive tract) through uptake processes. Internal dose is a more general term denoting the
amount absorbed without respect to specific absorption barriers or exchange boundaries. The
amount of the pollutant available for interaction by any particular organ or cell is termed the
biologically effective dose for that organ or cell.
Effectiveness: The improvement in health outcome that a prevention strategy can produce in
typical community-based settings.
-------
GLOSSARY (Continued)
Efficacy: The improvement in health outcome that a prevention strategy can produce in expert
hands under ideal circumstances
Exposure: Contact of a pollutant, physical, or biological agent with the outer boundary of an
organism; exposure is quantified as the concentration of the agent in the medium over time.
Margin of Exposure: The ratio of the critical NOAEL to the expected human exposure level.
Mechanism of Action: The complete sequence of biological events that must occur to produce
a toxic effect.
Mode of Action: A less-detailed description of the mechanism of action in which some, but not
all, of the sequence of biological events leading to a toxic effect is known.
Nonthreshold Effect: An effect for which it is assumed that there is no dose, no matter how
low, for which the probability of an individual's responding is zero.
No-Observed-Adverse-Effect Level (NOAEL): The highest exposure level at which there are
no statistically or biologically significant increases in the frequency or severity of adverse effects
between the exposed population and its appropriate control.
Outcome Measure: The final health consequence (e.g., cases prevented, quality-adjusted life
years) resulting from an intervention.
Pharmacodynamics: The determination and quantification of the sequence of events at the
cellular and molecular levels leading to a toxic response to an environmental agent (also called
toxicodynamics).
Pharmacokinetics: The determination and quantification of the time course of absorption,
distribution, biotransformation, and excretion of pollutants (also called toxicokinetics).
Physiologically-Based Pharmacokinetic (PBPK) Model: A model that estimates the dose to a
target tissue or organ by taking into account the rate of absorption into the body, distribution
between target organs and tissues, metabolism, and excretion.
Program Office: An EPA organizational unit that administers a major EPA program (i.e., Air
and Radiation; Water; Prevention, Pesticides, and Toxic Substances; and Solid Waste and
Emergency Response).
Reference Concentration (RfC): An estimate (with uncertainty spanning perhaps an order of
magnitude) of a continuous inhalation exposure of the human population (including sensitive
subpopulations) that is likely to be without an appreciable risk of deleterious noncancer effects
during a lifetime.
XI
-------
GLOSSARY (continued)
Reference Dose (RfD): An estimate (with uncertainty spanning perhaps an order of magnitude)
of a daily exposure of the human population (including sensitive subpopulations) that is likely to
be without an appreciable risk of deleterious noncancer effects during a lifetime.
Susceptibility: Increased likelihood of an adverse effect related to intrinsic (i.e., life stage,
genetic predisposition) or extrinsic determinants (i.e., preexisting disease) unique to the
organism.
Threshold Effect: An effect for which there is some dose below which the probability of an
individual's responding is zero.
Uncertainty Factor (UF): One of several factors used in calculating an exposure level that will
not cause toxicity from experimental data. For example, HP's are used to account for the
variation in susceptibility among humans, the uncertainty in extrapolating from experimental
animal data to humans, and the uncertainty in extrapolating data from studies in which agents are
given for less than a lifetime.
Vulnerability: Synonymous with susceptibility
xn
-------
EXECUTIVE SUMMARY
The mission of the U.S. Environmental
Protection Agency (EPA) is to protect
public health and safeguard the natural
environment. Risk assessment is an integral
part of this mission in that it identifies and
characterizes environmentally related
human health problems. The Human Health
Research Strategy document presents a
conceptual framework for future human
health research by EPA's Office of Research
and Development (ORD). This research
strategy outlines ORD's core research effort
to provide broader, more fundamental
information that will improve understanding
of problem-driven health risk issues
encountered by the EPA's Program and
Regional Offices. The scope of this
research document is strategic in that it
discusses broad themes and general
approaches. Implementation of an inte-
grated research program on human health is
described in greater detail in ORD's
Multiyear Plan on Human Health Research.
The Multiyear Plan identifies specific
performance goals and the measures needed
to achieve those goals over a 5- to 10-year
period. Each Laboratory and Center in
ORD is also developing an approach linking
research at the project level to the goals and
measures in the Multiyear Plan and the
general themes outlined in this research
strategy document.
Based on the needs of the EPA's
Program and Regional offices, recommen-
dations made by external advisory groups,
and goals established by EPA in response to
the Government Performance and Result
Act (GPRA) under Sound Science (Goal 8),
ORD has identified two strategic research
directions that will be pursued over the next
5 to 10 years (see text box).
Strategic Research Directions
Research to Improve the Scientific
Foundation of Human Health Risk
Assessment, including:
• Harmonizing Cancer and
Noncancer Risk Assessments
• Assessing Aggregate and
Cumulative Risk
• Determining Risk to Susceptible
Human Subpopulations
Research to Enable Evaluation of
Public Health Outcomes from Risk
Management Decisions.
Research in these strategic areas will
improve the scientific foundation for EPA's
risk assessments and lead to principles that
can be used to evaluate the effectiveness of
risk management actions aimed at improv-
ing environmental public health. Chapter 1
of the Human Health Research Strategy
document provides background information
regarding the regulatory and scientific basis
for a core research program on human health
risk assessment. Chapter 1 also develops
the need for a multidisciplinary, integrated
research program and describes how ORD
will formulate problems and approaches to
study complex questions related to human
health. Chapter 2 describes the scientific
uncertainties, objectives, and approaches
that ORD will use to harmonize risk assess-
ments, assess aggregate and cumulative risk,
and determine risk to susceptible subpopu-
lations. Chapter 3 describes ORD's public
health outcomes research program that will
work toward providing more scientifically
defensible assessments of actual reduction
in risk.
E-l
-------
ORD will focus on developing a
multidisciplinary, integrated program that
will build linkages between exposure, dose,
effect, and risk assessment methods to pro-
vide the scientific basis for harmonizing risk
assessment approaches, predicting aggregate
and cumulative risk, and protecting suscep-
tible subpopulations. In addition, ORD will
develop an integrated research program
utilizing its intramural scientific capacity in
conjunction with extramural grants, cooper-
ative agreements, and interagency agree-
ments. Efforts have been and will continue
to be made to identify and foster collabora-
tion with other Federal and State agencies,
as well as academic and private organiza-
tions having research programs that
complement ORD's research efforts.
Research to Improve the Scientific
Foundation of Human Health Risk
Assessment
ORD's human health risk assessment
program is based on the assumption that
major uncertainties in risk assessment can
be reduced by understanding and elucidating
the fundamental determinants of exposure
and dose and the basic biological changes
that follow exposure to pollutants leading to
a toxic response. Research in this area will
provide the scientific knowledge and prin-
ciples to improve the risk assessment of all
human health endpoints, aggregate and
cumulative risk, and susceptible and highly
exposed subpopulations.
Harmonizing Risk Assessment Approaches
ORD's research in this area will address
the disparate approach for the risk assess-
ment of cancer and noncancer endpoints.
Research on harmonizing risk assessment
approaches will lead to a common set of
principles and guidelines for drawing
inferences about risk based on mechanistic
information. The overall goal of this
research is that Program and Regional
Office risk assessors will use mechanistic
data in a harmonized manner for risk
assessments for all health endpoints.
Specific research objectives include the
following:
Q Develop emerging technologies or
methods to study mode or mechanism of
action;
Q Develop the biological basis for
understanding mode or mechanism of
action;
Q Develop a basis for comparing risk
across all health endpoints using
mechanistic information;
Q Develop principles for the use of
mechanistic data to select the most
appropriate risk assessment model; and
Q Develop principles for the use of
mechanistic data to reduce or replace
uncertainty factors in risk assessments,
especially for inter- and intraspecies
extrapolation, including approaches for
linking dosimetry models, such as
pharmacokinetic models, with empirical
or pharmacodynamic models for effects
of pollutants with similar or different
modes of action.
Aggregate and Cumulative Risk
ORD's research program on aggregate
and cumulative risk will address the fact that
humans are exposed to mixtures of pollu-
tants from multiple sources. Research will
provide the scientific support for decisions
concerning exposure to a pollutant by
multiple routes of exposure or to multiple
pollutants having a similar mode of action.
ORD will also develop approaches to study
how people and communities are affected
following exposure to multiple pollutants
that may interact with other environmental
stressors. Specific research objectives
include the following:
E-2
-------
Q Determine the best and most cost-
effective ways to measure human
exposures in all relevant media, includ-
ing pathway-specific measures of multi-
media human exposures to environ-
mental contaminants across a variety of
relevant microenvironments, exposure
durations, and conditions;
Q Develop exposure models and methods
suitable for EPA and the public to assess
aggregate and cumulative risk, including
mathematical and statistical relation-
ships among sources of environmental
contaminants, their environmental fate,
and pathway-specific concentrations;
models linking dose and exposure from
biomarker data; and approaches to assess
population-based cumulative risk,
including those involving exposure to
stressors other than pollutants; and
Q Provide the scientific basis to predict the
interactive effects of pollutants in mix-
tures and the most appropriate
approaches for combining effects and
risks from pollutant mixtures.
Susceptible and Highly-Exposed
Subpopulations
ORD research on susceptible subpopu-
lations will focus on developing a scientific
understanding of the biological basis for
differing responsiveness of Subpopulations
within the general population, including
factors associated with their differential
exposure. Research on biological suscepti-
bility will focus on the role of intrinsic
factors, such as life stage and genetic
background, and extrinsic factors, such as
preexisting disease, on responsiveness to
environmental pollutants. Specific research
objectives include the following:
G Identify the key factors that contribute to
variability in human exposure, including
the distribution of human exposures and
behavior associated with exposure to
pollutants;
Q Improve the accuracy of dose estimation
in the general population;
Q Identify the biological basis underlying
differential responsiveness of sensitive
Subpopulations of humans to pollutant
exposure; and
Q Determine how exposure, dose and
effect information can be incorporated
into risk assessment methods to account
for interindividual variability.
Research to Enable Evaluation of Public
Health Outcomes from Risk Management
Actions
Generally, EPA has not prepared retro-
spective evaluations to determine if the
intended benefits in protecting public health
were realized once an EPA decision had
been in place for a period of time. With the
advent of GPRA and calls for EPA to stress
and demonstrate outcome-oriented goals and
measures of success, research is needed to
enable evaluation of actual public health
outcomes from risk management actions.
Estimating public health benefits of EPA
regulatory decisions and rule making or, in a
more general sense evaluating public health
outcomes from risk management actions
will be a challenging undertaking. It will
involve a number of disciplines grounded in
both the physical and social sciences and
increasingly must take into account the
economic and behavioral aspects of human
decision-making.
The long-term goal of ORD's research
on public health outcomes will be to provide
the scientific understanding and tools to
EPA and others for use in evaluating the
effectiveness of public health outcomes
resulting from risk management actions.
Research will focus on identifying, dis-
covering, or developing the most effective
methods and models; determining how they
can be integrated into a decision-making
framework to assist Federal, State, and local
decision-makers in evaluating changes in
E-3
-------
public health as a result of risk management
actions; and developing a framework to
quantify such changes accurately. Specific
research objectives include the following:
Q Establish the linkage between sources,
environmental concentrations, exposure,
adverse effects or disease, and effective-
ness such that a change in a human
health outcome consequent to a risk
management action can be determined
by measuring or modeling any one of
these linked steps; and
Q Improve methods and models by which
others can measure or model changes in
public health outcomes following
various risk management actions.
Because of the novelty of the long- term
goal and research objectives and the require-
ment for an unusually high degree of
interdisciplinary coordination, ORD will
develop a multiyear implementation plan for
the public health outcomes research
program. This plan will provide
considerable details on the development,
investigation, and delivery phases of the
research.
E-4
-------
1. INTRODUCTION
The mission of the U.S. Environmental
Protection Agency (EPA) is to protect
public health and safeguard the natural
environment (i.e., air, water, and land) upon
which life depends. Risk assessment is an
integral part of this mission in that it identi-
fies and characterizes environmentally
related health problems. EPA's Office of
Research and Development (ORD) conducts
research that contributes to the scientific
foundation for risk assessment and risk
management decisions in EPA's regulatory
programs. Since 1996, ORD has used a
risk-based strategic planning process in
consultation with EPA's Program and
Regional Offices and the external scientific
community to set research priorities. From
this process, research to improve human
health risk assessment was identified as one
of six priority research areas in the 7997
Update to ORD's Strategic Plan (U.S. EPA,
1997a) and ORD Strategic Plan (U.S. EPA,
200 Ib). As such, fundamental human health
research is also part of the ORD Sound
Science Program under Goal 8, which is one
of EPA's 10 strategic environmental goals
in accordance with the requirements of
GPRA (see text box).
Goal 8: Sound Science, Improved
Understanding of Environmental Risk,
and Greater Innovation to Address
Environmental Problems - EPA will
develop and apply the best available
science for addressing current and future
environmental hazards as well as new
approaches toward improving
environmental protection.
1.1 PURPOSE OF THE STRATEGY
The Human Health Research Strategy
presents a conceptual framework of ORD's
future research directions in human health
risk assessment and risk management. This
Human Health
Research Program
Figure 1-1 Relationship Between Core and
Problem-Driven Research
strategy identifies the broad, overarching
questions that will guide ORD's core human
health research program over the next 5 to 10
years. Core research aims to provide broad,
fundamental scientific information that will
improve understanding of problem-driven
human health issues arising from risk
assessment in EPA's Program and Regional
Offices. Core research consists of under-
standing the fundamental processes that
underlie environmentally related health
problems; the development of broadly appli-
cable research and risk assessment tools and
approaches; and the design, implementation,
and maintenance of appropriate measures of
environmental exposure (NRC, 1997).
Approximately 40% of ORD's research
program has been defined as core research.
Problem-driven human health issues
associated with specific contaminants,
media, or issues (e.g., particulate matter,
arsenic in drinking water, disinfection
byproducts, endocrine disrupters) are
addressed in separate ORD Research
Strategies and Plans (see Appendix A).
Fundamental research issues that cut across
those research strategies must often be
addressed before more problem-driven
1-1
-------
questions can be studied. There will
be a constant need to integrate
problem-driven and core research as
illustrated in Figure 1-1. For
example, problem-driven research is
being conducted to study the
interaction of pesticides in mixtures
because the Food Quality Protection
Act (FQPA, 1996) indicates that
EPA should consider the risk
associated with cumulative
exposures of pesticides having a
common mechanism. However,
core or basic research on the mode
or mechanism of action of these
pollutants will have to be done
before addressing more problem-
driven questions concerning the
interaction of pesticides based on
their mechanism or mode of action.
Risk
Formulate the Problem
Risk
Hazard ID Characterization
Develop Compliance
Assurance Models
and Methods
Develop Measures of
Environmental and Publi
Health Improvement
Problem Identification
Define Risk Management
Objectives i
Identify/Evaluate Risk
Management Options
Risk Management Decision
Implement Option(s)
X
Monitor Environmental,
and Public Health
Improvement
-Public Health and
Legal
Considerations
-Social and
Economic Factors
-Political
Considerations
Reduced Environmental
and/or Public Health Risk
Figure 1-2 The Risk Assessment-Risk Management Paradigm
The Human Health Research
Strategy is not intended to be a
technical document. The target audience
includes EPA and other federal agency
scientists, managers, and policymakers, as
well as the scientific community at large. It
describes the scientific uncertainties,
objectives and general approaches that will
be taken by ORD's core research program
on human health. ORD has developed a
Multiyear Plan for Human Health Research
that describes anticipated goals and
measures over a 5- to 10-year period. In
addition, each Laboratory and Center within
ORD is developing its own approach to link
specific projects and tasks to the ORD
Multiyear Plan and the themes described in
this research strategy document.
1.2 CURRENT RESEARCH PROGRAM
ON HUMAN HEALTH
Human health risk assessment provides a
qualitative and quantitative characterization
of the relationship between environmental
exposures and effects observed in exposed
individuals. In 1983, the National Research
Council (NRC) described four primary steps
in the process of risk assessment, i.e., hazard
identification, dose-response assessment,
exposure assessment, and risk characteriza-
tion (Figure 1-2). Risk assessment is the
primary scientific input to the risk manage-
ment process, which involves the recogni-
tion of a potential new risk and develop-
ment, selection and implementation of EPA
actions to address the risk. Risk manage-
ment often considers a wide variety of other
factors. The overall process of risk assess-
ment and risk management is often called
the Risk Assessment-Risk Management
Paradigm.
Over the last several years, ORD has
aligned its organizational structure and
research program to be consistent with the
Risk Assessment-Risk Management
Paradigm (Appendix B)(see text box on next
page). ORD is organized into three national
Laboratories and two Centers. The National
Exposure Research Laboratory (NERL)
focuses on measuring exposures and
1-2
-------
Laboratories and Centers in OKD
Major Focus
Exposure and Dose
Dose and Effects
Risk Assessment
Risk Management
Extramural Research
Lab/Center
NERL
NHEERL
NCEA
NRMRL
NCER
producing scientifically defensible exposure
models that reduce the gaps in scientific
knowledge related to actual human exposure
to pollutants. In 1995, EPA's Science
Advisory Board (SAB) (U.S. EPA, 1995)
reviewed the state of exposure assessment
science and reported that this area was
hampered by a variety of technical
limitations, including lack of exposure
measurement techniques, a paucity of
exposure databases and other exposure-
relevant data, and reliance on numerous
default assumptions with little justification
for their selection. The SAB also found that
available exposure models had rarely been
evaluated against actual human exposure
measurements. In addition, there were no
comprehensive human exposure models that
could describe the complex relationships
between pollutant sources, environmental
concentrations, exposure pathways, actual
human exposures, and the dose that results
from exposure to pollutants by multiple
pathways. The SAB also found that the
methods available for both human exposure
measurements and exposure modeling were
too intrusive or costly to implement
routinely. Much of the work conducted by
NERL over the last several years has been
directed at these data and methodological
gaps.
In the Risk Assessment-Risk Manage-
ment paradigm, dose-response assessment is
the process for determining the likelihood of
an adverse effect at a particular exposure or
dose. A primary concern for dose-response
assessment is an understanding of the dose
of the pollutant that reaches its target organ,
tissue, cell, or biomolecule. Research on
issues related to dose is largely conducted at
NERL and the National Health and Environ-
mental Effects Research Laboratory
(NHEERL). Research at NERL focuses on
pharmacokinetic (PK) modeling to estimate
internal dose metrics for multiroute aggre-
gate exposure. Research at NHEERL
focuses on determining the basis for meta-
bolic differences between species. This
information is crucial for extrapolating
toxicological data from animals to humans
in risk assessment and determining the
biologically effective dose of the parent
compound or metabolite(s) of the pollutant.
The goal of hazard identification is to
describe and ultimately predict in humans
the toxicological effects of pollutants that
might occur due to exposure to environ-
mental agents. Research related to hazard
identification is largely conducted at
NHEERL and focuses on test methods
development and characterization of hazard
potential in animal models. Clinical or
epidemiological studies are also used to
identify potential risks in the human
population and generate testable hypothesis
for future studies in animal or in vitro
models. Risk assessment is often confoun-
ded by a number of uncertainties related to
the risk assessment methodology, including
extrapolation across species, extrapolation
from short-term to lifetime exposures, and
variability of response within the human
population. A significant component of
research at NHEERL focuses on reducing or
eliminating uncertainties in the risk assess-
ment process. Research at NHEERL also
seeks to understand the cascade of events
between the presence of a pollutant at a
target site and the ultimate manifestation of
toxicity. Knowledge of the sequence of
biological events that must occur to produce
an adverse effect [i.e., the mechanism of
action, or an understanding of some, but not
1-2
-------
all, of the key biological steps leading to
toxicity, i.e., the mode of action (U.S. EPA,
1996; U.S. EPA, 1999a; IPCS, 1999;
Schlosser and Bodganffy, 1999)] is being
used with increasing frequency in risk
assessment (see Appendix C). Procedures
for the use of mechanistic data are defined
in the EPA's draft Guidelines for
Carcinogen Risk Assessment (U.S. EPA,
1999a).
The National Center for Environmental
Assessment (NCEA) performs complex risk
assessments of national interest and
develops risk assessment methods, data-
bases, and tools based on results produced
by ORD and others. NCEA also serves an
integrating function within ORD, bringing
together results from hazard identification,
dose-response assessment, and exposure
assessment on issues related to the risk
assessment process. The risk assessment
program includes development of dose-
response and exposure models, factors,
databases, and guidance for conducting risk
assessment. Issues confronting the risk
assessment program include how to use
exposure, pharmacokinetic, and mechanistic
data in risk assessment; harmonizing cancer
and noncancer risk assessment methods, and
conducting cumulative risk assessments of
multiple pollutants.
The National Risk Management
Research Laboratory (NRMRL) focuses on
providing the most effective and useful risk
management options and improving the
linkage between risk assessment and risk
management efforts.
Intramural research conducted by
NERL, NHEERL, NCEA, and NRMRL is
complemented by extramural research
sponsored by ORD's National Center for
Environmental Research (NCER). Through
the Science to Achieve Results (STAR)
Program, NCER supports grants that focus
on specific research needs consistent with
the mission of the EPA. For example, the
STAR Program provides support to extra-
mural scientists to develop statistical and
predictive approaches for assessing risks
from pollutant mixtures. Other examples of
STAR research include 12 EPA/National
Institute of Environmental Health Sciences
(NIEHS)-supported Centers for Children's
Health and Disease Prevention Research and
individual studies, such as the development
of biomarkers for risk assessment in
children.
1.3 FUTURE RESEARCH PRIORITIES
1.3.1 Framework for an Integrated
Research Program in ORD
ORD will develop a multidisciplinary
research program that addresses linkages
lying along a continuum from the source of
an agent through exposure and dose to
adverse outcome such as disease (Figure 1-
3). One example of the need for an
integrated research program arises from the
opportunities and challenges associated with
data in recent reports by the Centers for
Disease Control and Prevention (CDC,
2001, 2003). The first publication, the
National Report on Human Exposure to
Environmental Chemicals., provided an
ongoing assessment of the U.S. population's
exposure to environmental chemicals using
biomonitoring. This report contains blood
and urinary values on 27 pollutants collected
as part of the National Health and Nutrition
Examination Survey (NHANES). The
second report, which was released in
January 2003, presents biomonitoring
exposure data for 116 environmental chemi-
cals for the U.S. population divided into age,
gender, and race/ethnicity groups. The
exposure information should help prioritize
research on the relation between exposure
and health effects and help identify popula-
tion groups with unusually high exposure
for health effects evaluation. Efforts will be
made to link the biomonitoring data back to
1-4
-------
Scientific Elements of Human Health Risk Assessment
SOURCE/STRESSOR
FORMATION
Chemical
Physical
Microbial
Magnitude TRANSPORT /
Duration TRANSFORMATION
DISEASE
T
t
ALTERED STRUCTURE /
FUNCTION
Cancer
Asthma
Infertility
etc.
Timing
Dispersion
Kinetics
Thermodynamics
Distributions
Meteorology
T
t
ENVIRONMENTAL
CHARACTERIZATION
EARLY BIOLOGICAL
EFFECT
Edema
Arrhythmia
Enzymuria
Necrosis
etc.
Air
Water
Diet
Soil & dust
T
DOSE
Molecular
Biochemical
Cellular
Organ
Organism
Activity
Patterns
EXPOSURE
Pathway
Route
Duration
Frequency
Magnitude
>
• Individual
• Community
• Population
> Absorbed
Target
Internal
Biologically Effective
Statistical Profile
Reference Population
Susceptible Individual
Susceptible Subpopulations
Population Distributions
Figure 1-3 The Exposure-Dose-Effect Continuum
pathway and source to guide risk manage-
ment interventions.
ORD's evolving integrated approach to
problem formulation and research planning
is illustrated in Figure 1-4. Risk assessment
issues arising from Regional or Program
Offices, through legislative or regulatory
mandates or ORD research results will be
evaluated to determine the scientific ques-
tions (Figure 1-4, Box A). This evaluation
will lead to the design of studies to address
those uncertainties (Box B). Results from
these studies (Box C) will be used to refine
additional studies and/or generate models
(Box D) to inform the development of better
risk assessment methods. Efforts to
construct modules or compartments of
models (Box D) will feed back onto the
design and execution of additional experi-
ments. Ultimately, results from all experi-
ments and models will be used to develop
risk assessment methods (Box F) and
develop an integrated framework (Box E)
that will form the scientific basis for risk
assessment guidance and risk management
decisions. Consolidated information
resulting from the integrated framework
may also be used to iform or redefine the
original risk assessment issue.
A conceptual model illustrating a com-
pletely integrated research program is
illustrated in Figure 1-5. As this figure
shows, analysis of risk assessment issues
gives rise to scientific questions concerning
exposure, dose, effects, and risk assessment
methods. For example, risk assessment
questions related to exposure might require
studies involving the development of
analytic methods and the execution of pilot-
scale laboratory or field exposure research
followed by larger scale population or
epidemiological studies to gain important
1-5
-------
BoxB
Scientific
Uncertainties-
Study Design
BoxE
Integrated
Framework
BoxF
Risk Assessment,
Data, Methods,
Guidance
Risk Management
Figure 1-4 ORD's Approach to Solving Agency Problems
exposure and/or exposure factor data. The
results of this research could be used to help
develop exposure assessment models.
Research questions related to dose might
involve experiments to develop analytical
methods, obtain pharmacokinetic data, or
identify biomarkers. The results of these
experiments would be used to develop
physiologically based, pharmacokinetic
models for estimating internal dose. Effects
research may require the development of
sensitive and specific methods to help
understand the biological substrates under-
lying the mode or mechanism of action of
environmentally relevant pollutants. Epi-
demiological studies may provide the basis
for confirming possible health-related
adverse effects in the human population and
generate testable hypotheses for subsequent
confirmation in animal or in vitro models.
The results of effects research would be
used to develop biologically based dose-
response models linking effects observed at
the cellular or molecular level to adverse
health effects. Assessment of data gene-
rated from exposure, dose and effects
research would be used to formulate better
risk assessment methods. All of the data
generated from research on exposure, dose,
effects, and risk assessment methods would
be used to help develop an integrated
framework for the development of guidance
for risk assessment and scientific support for
risk management options.
Figure 1-5 also shows that results from
various experiments and models may feed
back at any time through an iterative process
to improve the design of future experiments.
Results from experiments and outputs from
models in any area of analysis (i.e., expos-
ure, dose, effect, risk assessment) may
influence the design of studies and the
generation of data in other areas. For
example, the results of field studies concern-
ing exposure of children to pesticides might
1-6
-------
Risk Assessment
Methods-Study
Design
Analytic methods
Field and lab studies
Population studies
Epidemiological studies
Analytic methods
PK/PD measures
Biomarkers
Methods
Biological substrates
Mechanistic studies
Epidemiological studies
_C
Analyze data bases
Evaluate biological
data
Guidance/
Risk Management
Figure 1-5 ORD's Integrated Research Program
influence the choice of dose or concentra-
tion of pollutants for future research.
ORD ' s ongoing research on asthma and
exposure of children to fungi and molds
serves as a specific example of a multidisci-
plinary, integrated research program that
uses the scientific expertise and resources
from the various ORD Laboratories and
Centers to address a high priority research
issue (Figure 1-6). In 1998, a team of
researchers from NERL, NHEERL, NCEA
and NRMRL was organized to address the
effects of the S. chartarum fungus, a
common indoor contaminant, on children's
health. The first objective of this program
was to determine, before and after remedia-
tion, the quantities of S. chartarum spores in
dust from homes of children with asthma or
pulmonary hemosiderosis and assess
specific antibodies to mold proteins in these
children. A second objective was to
establish a mouse model of allergic lung
disease to characterize IgE-inducing
proteins from three fungi, including S.
chartarum, immunologically, and identify
any common characteristics using advanced
proteomics. This research addresses the
hypothesis that differences in protein
constituents of mold
are associated with
allergenicity. The third
objective of this
research program is to
demonstrate parallels
between human and
rodent responses to the
mold in order to
facilitate interspecies
extrapolation.
Epidemiological and
clinical studies evaluate
the exposures of
children to fungi that
might lead to asthma
using a cohort of
children to addresses
the hypothesis that
participants in the fungal-exposed cohort
will have significantly more asthma than
control participants. Other objectives of this
research are to test methods to reduce spore
release and growth of fungus and to begin to
develop a risk assessment model. The
ultimate goal is develop a model than can
be used to address risk assessment and risk
management approaches for indoor molds
associated with asthma and other health
conditions.
Figure 1-6 illustrates the integrated
multidisciplinary approach that has been
developed to address this high priority need
of EPA. Exposure data from field studies
identify and characterize exposures to fungi
that might be associated with childhood
asthma. These studies also help define the
relationship between exposure and effects
and provide important exposure information
for the design of effects studies and risk
assessment approaches. Research on effects
focuses on developing animal models of
allergic lung disease that can be extrapolated
to humans and on studies providing a causal
link between the potential mode of action or
mechanism and allergic lung disease.
Mechanistic effects research helps confirm
1-7
-------
Exposure-
Study Design
Effects-
Study Design
Risk Assessment
Study Design
the associations observed in
the exposure assessment and
could lead to the identification
of specific fungi species
involved in producing allergic
lung disease. Epidemiological
studies in children provide
important information for the
design of risk assessment
approaches to protect children
exposed to fungi and help
shape the design of future
studies. Risk assessment
approaches are being
developed based on results
from the exposure assessment
and effects research, all of
which provide the scientific
basis for development of risk
management options and
remediation strategies, if
necessary. Once a remediation strategy has
been implemented, future studies will be
designed to evaluate the effectiveness of the
strategy. Depending on the outcome of
these studies, additional research on
exposure, effects, and risk assessment
models may be initiated to devise a more
effective risk assessment-risk management
approach.
What is the Health Risk to
Children following
Exposure to Molds?
Assess spores in homes before
and after remediation
Develop animal model
Test spores
Epidemiological studies
Develop remediation
strategy for children
exposed to molds
Develop risk assessment
approaches to protect
children exposed to molds
Figure 1-6 Integrated ORD Research Program
on Molds and Asthma in Children
Strategic Research Directions
Research to Improve the Scientific Foundation of
Human Health Risk Assessment:
• Harmonizing Cancer and Noncancer Risk
Assessments
•Assessing Aggregate and Cumulative Risk
•Evaluating the Risk to Susceptible Human
Subpopulations
1.3.2 Research Themes
Based on input from Regional and Program
Office risk assessors and ORD scientists,
future ORD research will focus on two
strategic directions (see text box), including
research to improve the scientific foundation
of human health risk assessment and
research to enable evaluation of public
health outcomes from environmental risk
management decisions. Research to
improve human health risk
assessment will emphasize three
themes, i.e., harmonizing cancer and
noncancer risk assessments,
assessing aggregate and cumulative
risk, and evaluating risks for
susceptible and highly exposed
subpopulations. ORD has deter-
mined that there is a need to develop
the scientific basis for harmonizing
the use of mechanistic information in
cancer and noncancer risk
assessments. Research on assessing
aggregate and cumulative risk
addresses the need to develop risk assess-
ment/risk management approaches to
-------
evaluate multichemical/multipathway
exposures to environmental agents, while
research on risks to susceptible and highly
exposed subpopulations focuses on under-
standing variability in human responses to
environmental agents. Susceptible subpopu-
lations also include populations of people
that are differentially exposed to environ-
mental agents. These themes are discussed
in greater detail in Chapter 2.
ORD will also initiate research to
enable the evaluation of public health
outcomes from risk management actions.
This program will include new EPA efforts
to measure and monitor improvements in
environmental public health following risk
management actions as underscored by
requirements that EPA evaluate the success
of its environmental programs and deci-
sions. Success will be measured by changes
in health outcomes and indicators resulting
from risk management decisions. EPA has
traditionally relied on "process" measures
(e.g., decreased emissions, number of sites
cleaned up) to measure public health benefit
indirectly. ORD's future research program
will seek to identify and validate health
events that can serve as true public
health outcome measures (Figure 1-7).
The regulatory and scientific bases for
this part of the research program are
described in greater detail in Chapter 3
of this document.
1.4 STRATEGIC PRINCIPLES
The following strategic principles
will be used in developing and imple-
menting ORD's research program on
human health:
Collaboration across ORD - As
described previously, the intramural
ORD program is organized around the Risk
Assessment-Risk Management Paradigm,
i.e., NERL, NHEERL, NCEA and NRMRL
(Figure 1-2). ORD's future research
program will focus on more complex
environmental problems requiring
collaboration and synergy between the
various Laboratories and Centers in ORD.
Scientists in Program and Regional Offices
are viewed as collaborators as well as
clients, and collaborative relationships will
be established to design and conduct studies
related to human health risk assessment and
risk management.
Focus and Broad Application - A
research strategy to improve human health
risk assessment and management must
emphasize selected high-priority issues with
outcomes expected to have a wide effect on
risk assessment. ORD will focus the core
human health research program on environ-
mental pollutants, which is consistent with
the expertise and infrastructure ORD has
developed over the last several years.
However, as knowledge gaps are identified
for other classes of environmental agents,
such as microbes and bioaerosols, research
will be initiated to address specific questions
related to those agents.
Research
T
r
Assessing
Risk:
Exposure .
Dose/Effects
Risk Assessment
i
Informing Risk
Management
Options:
Regulatory
Risk Management
k
w
Assessing
Effectiveness of
Decisions:
Health Indicators
Exposure Indicators
Figure 1-7 Role of Research in Assessment of Health Outcomes
in the Risk Assessment and Risk Management Process
1-9
-------
Support EPA's Mission - The research
must address knowledge gaps in risk assess-
ment identified by Program and Regional
Offices or raised by specific regulatory or
legislative requirements. Results should
have tangible benefits to all groups interes-
ted in improved risk assessments (i.e., states,
local governments, industry, nongovern-
mental environmental organizations,
communities, international governments).
ORD's research will result in products and
information that have direct and practical
applications in risk assessment. ORD
scientists will also identify issues that may
be important to the future of risk assessment
that are not major concerns to programs and
regions at the present time.
Outreach, Coordination, and Partner-
ship with External Scientific Community -
ORD will develop an integrated research
program utilizing its intramural scientific
capacity in conjunction with extramural
grants and cooperative and interagency
agreements. In addition, efforts have been
and will continue to be made to identify and
foster collaboration with other federal and
state agencies, as well as academic and
private organizations, that complement
ORD's research efforts (see Appendix D).
1-10
-------
2. RESEARCH TO IMPROVE THE SCIENTIFIC FOUNDATION
OF HUMAN HEALTH RISK ASSESSMENT
ORD's human health risk assessment
program is based on the assumption that
major uncertainties in risk assessment can
be reduced by understanding the funda-
mental principles of how, at what level, and
how often humans are exposed to pollutants;
how much of the toxic moiety arrives at the
target site; and the basic biological changes
that lead to a toxic or adverse health effect.
Research questions related to harmonizing
risk assessment, assessing aggregate and
cumulative risk, and evaluating risk to
susceptible and highly exposed subpopu-
lations will be framed to address knowledge
gaps and interrelationships of events along a
continuum from source through exposure
and dose to effect (Figure 1-3). The overall
objective of ORD's human health research
program is to link exposure, dose, and effect
approaches along this continuum in order to
provide an integrated information base for
scientifically defensible risk assessment and
risk management decisions.
2.1 Research on Harmonizing Risk
Assessment Approaches
2.1.1 Scientific Uncertainties
Assessment of health risk from expos-
ures to environmental agents has tradition-
ally been performed differently depending
upon whether the response is a cancer or a
noncancer health effect. This practice has
been based on a limited understanding of the
mode of action of toxic substances. Histori-
cally, cancer was thought to be largely the
consequence of the direct interaction of a
carcinogen with DNA to produce a heritable
change in a single cell that eventually pro-
duced a tumor. It was thought, therefore,
that the dose-response for such a mechanism
would not show a biological threshold but
would be linear at low doses. This led EPA
to employ a science policy that cancer risk
should be estimated by a linear, nonthres-
hold dose-response method.
On the other hand, a threshold has been
generally assumed for noncancer effects
based on considerations of compensatory
homeostasis and adaptive mechanisms. The
threshold concept assumes that a range of
exposures can be tolerated up to some finite
level without adverse effects. This
threshold will vary from one individual to
another, so that there will be a distribution
of thresholds in the population. Except for
some pollutants, such as the criteria air
pollutants, evaluating human risks for
noncancer effects has generally involved the
determination of a level of daily exposure
that is likely to pose no appreciable risk of
deleterious effect during a lifetime.
The disparate approach for assessment
of cancer and noncancer endpoints has been
questioned (e.g., NRC, 1994). It now
appears that carcinogens can affect many
cellular targets and biochemical and biolog-
ical processes that eventually lead to the
formation of tumors. Such targets may in-
clude DNA, which contains the genes that
control cell growth, or biochemical
processes involved in cell growth regulation,
cell signaling, and cell-to-cell communica-
tion. Other mechanisms may involve cell
toxicity and death, perturbation of hormonal
systems, and suppression of the immune
system. Many of these mechanisms may
have thresholds of response, as discussed in
the proposed new cancer risk assessment
guidelines (U.S. EPA, 1996, 1999a). It has
also been hypothesized that threshold
considerations may not be applicable to all
noncancer effects, e.g., lead-induced
cognitive deficits in children. Furthermore,
our emerging understanding of the
mechanisms of carcinogenesis and other
health effects suggests that the underlying
2-1
-------
basis for certain noncancer and cancer
endpoints may have common precursors.
For example, pollutant-induced toxicity can
cause altered biological function, cell death,
and tissue regeneration while surviving cells
compensate for that injury by increasing cell
proliferation which may result in tumor
formation if continued unchecked. Thus,
the primary precursor effect may be related
to both the cancer outcome and other types
of noncancer biological effects.
Understanding an agent's mechanism of
action will be crucial to a more accurate
prediction and characterization of hazard
and risk and will be the basis for developing
harmonized approaches for all health
endpoints. Harmonization in this context
refers to the development of a consistent set
of principles and guidelines for drawing
inferences from scientific information. It
does not mean that a single
methodology should be used for
the assessment of all toxicities and
pollutants. Instead, it emphasizes
the need for consistent application
of all pertinent information on
toxicity, dosimetry, mode of
action, and exposure in all risk
assessments regardless of the
nature of toxicities or pollutants.
ORD will focus its research to
improve the foundation of these
risk assessment methods by
seeking to better understand the
mechanisms or modes of action
that are common to cancer and
noncancer health effects. One of the goals
of ORD's research on harmonization is to
improve general understanding of mechan-
isms and modes of toxic action.
2.1.2 Research Objectives
The following research objectives
provide the framework needed to develop an
integrated research program that will
harmonize risk assessment approaches:
Develop, validate, and standardize
emerging technologies or methods to
study mode or mechanism of action;
Develop the biological basis for under-
standing mode or mechanism of action;
Develop a basis for comparing risk
across all health endpoints using
mechanistic information;
Develop principles for the use of
mechanistic data to select the most
appropriate risk assessment models; and
Develop principles for the use of
mechanistic data to reduce or replace
uncertainty factors in risk assessments,
especially for inter- and intra-species
extrapolation, including approaches for
linking dosimetry models, such as phar-
macokinetic models, with empirical or
pharmacodynamic models for effects of
pollutants with similar or different
modes of action.
Exposure Scenario
Interspecies
Differences
Absorbed Dose
S
Target Dose
*
Internal Dose
Mode or Mechanism
of Action ,
Precursor Effects
Early Biological
Effects
2.1.3 Research Approach
Exposure Research. Specific exposure
issues have not been identified within the
context of the harmonization of risk assess-
ment approaches. Research to characterize
the various exposure pathways to relevant
pollutants is described in Section 2.2 under
the theme of Aggregate and Cumulative
Risk and includes a description of the
magnitude and nature of the pollutants to
2-2
-------
which people are exposed, as well as the
timing and sequence of those exposures.
Research on differential exposure of
susceptible and highly exposed
subpopulations is described in Section 2.3.
Dose Research. It is hypothesized that
there may be common biological effects that
serve as precursors to various health effects.
For example, some pollutants may cause
multiple effects, both cancer and noncancer,
through initially similar mechanisms, such
as adduction of DNA or binding to a
receptor. Subsequent events must differ in
order to produce different effects. Other
pollutants can cause multiple effects through
multiple mechanisms, often through the
formation of metabolites with different
biological activities. In either case,
knowing the biologically effective dose
of the active pollutant at the target site is
crucial for elucidating mechanisms and
modes of cancer and noncancer health
effects for risk assessment. Research on
dose will identify the biologically effec-
tive dose of parent compound or
metabolites in target tissue and attempt to
relate those levels to the presence of early
biological and precursor effects at the
molecular, biochemical, cellular, organ,
and organismal levels (see schematic on
previous page). This information will,
therefore, be crucial for studies attempting
to elucidate mode or mechanism of action.
The development of pharmacokinetic
models to inform studies on mode or
mechanism of action must also take into
account variables such as the duration of
exposure and possible interspecies
differences in sensitivity.
Effects Research. Central to the ques-
tion of harmonizing risk assessment
approaches is whether various modes or
mechanisms of action have a similar
necessary step (e.g., cell proliferation,
receptor interaction, response to injury or
stress, alterations in DNA repair mechan-
isms, apoptosis) leading to the adverse
effect. Virtually every toxic event in a
tissue or organism exposed to a pollutant is
modulated by a finite number of damage-
response pathways by which cells sense the
status of their internal environment.
Through these sensors, critical processes
that activate specific genes or proteins to
cause the cell to migrate, proliferate,
differentiate, or die are made by a cell's
biochemical machinery. Progress in this
area depends upon a clear understanding of
the changes in the biology of the cell
following delivery of the active chemical
moiety to target cells and the relationship of
responses with dose. Determining the pres-
ence of the active toxic moiety at specific
target sites will be crucial for these studies.
Develop
and S]
Met
Mechani
„ . . Identify Ce
Sensitive _
.,,. Processes
, Cancer a
hods
Noncancer J
^^
Compare Dose-
Response Curves
for Prototypic
Compounds
jf |r
llular Identify Common
for Precursors for
nd Cancer and
iffects Noncancer Effects
^
* Develop BBDR
stic Information PBPK Models
A significant first step in effects research
on harmonization will be the development
of sensitive and specific methods (see sche-
matic above) to study mechanism or mode
of action based on the application of
emerging technologies, especially prote-
omics and genomics. Bioinformatic
approaches will also have to be developed to
help interpret the meaning of changes
coming from multigene, microarray assays
used in hazard identification and quantita-
tive dose-response assessment. Effects
2-3
-------
research will initially focus on identifying
cellular processes (e.g., regeneration,
proliferation) that may be similar for cancer
and noncancer health effects. This research
will lead to studies that will identify
common biochemical or molecular
pathways associated with those cellular
processes. Research will then focus on
studies concerning the effects of environ-
mentally relevant doses or concentrations of
prototypic pollutants with similar putative
modes or mechanisms of action or on
pollutants sharing similar structure-activity
relationships. If a common cellular target
can be identified for specific adverse
outcomes, physiologically based
pharmakokinetic (PBPK) models will
determine target tissue levels and the
influence of duration of exposure and
interspecies variation on adverse effects.
One of the long-term goals of ORD's
research is the development of models that
take into account the sequence of early
biological events leading to adversity (i.e.,
mechanisms or modes of action) for
multiple endpoints, the shape of the dose-
response curves at low doses, and the
influence of interspecies differences. ORD
research will study both high and low doses
in order to elucidate the likely shape of the
dose-response curve and to determine
whether different modes or mechanisms
may be operating at different doses. ORD
will also focus on developing animal models
that can be extrapolated directly to humans.
Mechanistic effects research based on
emerging technologies such as proteomics,
genomics, and bioinformatics will also feed
directly into ORD's efforts to set mechanis-
tically based priorities for pollutant risk
assessments and optimize in vivo and in
vitro testing requirements through the use of
computational methods and molecular
profiling, i.e., computational toxicology (see
text box). For example, computational
methods, such as quantitative structure-
activity relationships (QSAR), could be used
Computational Toxicology
The application of models from computa-
tional and systems biology and computa-
tional chemistry for prediction and
understanding mechanisms of action
to determine which set of chemicals out of a
larger population [e.g., Toxic Substances
Control Act (TSCA) inventory] might have
the potential to produce an adverse effect
(e.g., cancer or reproductive toxicity). This
information could be used to prioritize sub-
sequent testing of this subset of chemicals
for potential human health or environmental
effects. Emerging technologies such as
genomics and proteomics could be used to
generate molecular profiles that would serve
as diagnostic tools to discriminate toxico-
logical pathways leading to different
adverse effects. Diagnostic tools could be
used to design in vitro and in vivo tests to
confirm the toxicological pathway involved
in producing the adverse effects. This
information would then be used to guide the
selection of specific testing protocols for
risk assessment. ORD will initially
demonstrate the feasibility of this approach
by focusing on prioritization and screening
assays and models for endocrine disrupting
chemicals. This class of pollutants was
chosen because ORD has considerable
experience in determining environmental
exposure levels to these chemicals, as well
as developing in vivo and in vitro tests in
response to provisions of the Food Quality
Protection Act (FQPA). The primary focus
of these studies will be on endocrine-
mediated reproductive and developmental
effects following exposure to
environmentally relevant concentrations.
Risk Assessment Methods. In develop-
ing harmonized approaches for the
assessment of risk to different health
endpoints, a key issue is to determine how
much information is needed to show that a
2-4
-------
particular toxic effect is mediated by a
specific mode of action and that the
pollutant or its metabolite is present in
sufficient quantities in the target tissue (see
schematic).
or proteomic methods. This "translational"
research will be a major challenge for EPA
as the onslaught of data generated by these
new approaches will far outpace the
research and guidance on interpretation and
application in risk assessment.
PBPK
Data Base
BBDR Data Base
Improved Risk Assessment
Methods for All
/ Risk Assessment ^^
\
Information
Required to Use
Mechanism or
Mode of Action
Choice of Risk
Assessment
Model
Use of Uncertainty
Factors and Default
D/R Assessments and
Extrapolations
Evaluation of
Data from
Emerging
Technologies
For example, the proposed cancer risk
assessment guidelines (U.S. EPA, 1999a,
1996) provide for judging the plausibility
and adequacy of available evidence for a
postulated mode of action, identifying
susceptible subpopulations, and determining
the most appropriate approaches and
methods for low-dose extrapolation. ORD
research on risk assessment methods will
focus on how to incorporate mode-of-action
information for other health endpoints.
Guidance will also be developed to
determine how different endpoints of
toxicity could develop through common
biological processes or modes of action.
One high priority for ORD research on risk
assessment methods will be prototype
assessments for both data-rich and data-poor
pollutants to illustrate how mode of action,
PBPK, and biologically based dose response
(BBDR) models may be used in lieu of
default approaches. Risk assessment
research is also needed to develop principles
to evaluate the results of studies in which
the data have been generated using genomic
Recent EPA guidance to
improve risk assessments has
emphasized the importance of
providing risk managers with a
fuller characterization of risk.
Current default approaches to
express risk for health effects
presumed to be mediated by
threshold or nonlinear modes of
action include the use of reference
toxicity values [e.g., chronic oral
Reference Dose (RfD), inhalation
Reference Concentration (RfC), or
the concept of the margin of
exposure (MOE)], i.e., the ratio of
the critical NOAEL to the expec-
ted human exposure level. Although these
risk assessment models consider all the
available data, they do not provide an
explicit estimate of variability and
uncertainty or provide information on the
consequences of exposures that exceed the
reference values or have a small MOE.
An important focus of ORD's risk
assessment research on harmonization will
be the development of approaches to charac-
terize variability and uncertainty in refer-
ence toxicity values and to provide a proba-
bilistic framework for estimating risks asso-
ciated with exposures above reference tox-
icity values. This research will examine
data underlying the various uncertainty fac-
tors commonly applied in setting reference
values, including factors for inter-species
and intra-species extrapolation (including
pharmacokinetic and pharmacodynamic
variability) and variability in responses due
to changes in exposure duration. As an
example, this research would explore the use
of probability distributions that can be com-
2-5
-------
bined to characterize the variability and
uncertainty around the reference values for
health effects. Various statistical
approaches will be explored as a means for
estimating risks above the reference toxicity
values for informing risk management
decisions and supporting economic benefits
analyses. Risk assessment methods on risk
predictive models for cancer and noncancer
effects will also be investigated.
2.2 Research on Aggregate and
Cumulative Risk
2.2.1 Scientific Uncertainties
The development of risk assessment
methodology during the 1970s and early
1980s closely followed EPA's strategy for
pollution control. Historically, EPA evalu-
ated the risks of a single pollutant in a single
exposure medium, such as lead in outdoor
air or drinking water. In reality, people are
constantly exposed to mixtures of pollutants.
Furthermore, exposure to the same pollutant
may occur from a variety of routes, includ-
ing the air, water, and food. In addition, the
composition and concentration of pollutants
in the environment is constantly changing,
Working Definitions Developed by EPA
Science Policy Council
Aggregate Exposure: The combined
exposure of an individual or defined
population to a specific agent or stressor via
relevant routes, pathways, and sources.
Aggregate Risk: The risk resulting from
aggregate exposure to a single agent or
stressor.
Cumulative Risk: The combined risks
from aggregate exposures to multiple agents
or stressors.
depending on people's activities and geo-
graphical location. It is now fully under-
stood that environmental exposure to pollu-
tants occurs via multiple exposure routes
and pathways, including inhalation, inges-
tion, and uptake through the skin. Research
on aggregate and cumulative risk will focus
on defining the multitude of ways in which
people are exposed to pollutants and charac-
terizing the subsequent effects and risks.
The FQPA directed EPA to include in its
assessment of pesticide safety the risk asso-
ciated with the cumulative effects of chemi-
cals that have a common mechanism of
toxicity and to consider aggregate dietary
and non-occupational sources of pesticide
exposure. However, EPA's efforts to assess
aggregate and cumulative risk go far beyond
the FQPA and pesticides. For example, the
Office of Water must assess risks from mix-
tures of disinfectants and their byproducts
and must balance those risks against the
risks of toxic microbes in the drinking water
supply. The Air Program Office needs
methods to assess risks from mixtures of
criteria air pollutants and sources containing
a mixture of hazardous air pollutants. The
Waste Program Office deals with mixtures
of many different chemical classes found
together in the soil, water, and air of waste
sites and their surroundings. In addition,
EPA's Program and Regional Offices deal
with communities that may be more highly
exposed than average and subject to a
variety of other stressors such as poverty,
lack of access to medical care, inadequate
nutrition, and stresses associated with living
near landfills, incinerators, and/or heavy
industry. To encompass all these concerns,
this document defines aggregate exposure
and cumulative risk broadly in accordance
with the working definitions developed by
EPA's Science Policy Council (see text
box).
2-6
-------
The traditional approach to assessing
aggregate and cumulative risk is to focus
primarily on individual pollutants and their
sources. The pollutants are initially traced
through the environment, and the concentra-
tions and doses of each chemical are estima-
ted separately. The toxicity and risks from
the multiple stressors are added or combined
using the basic methods in EPA's Chemical
Mixtures Guidelines (U.S. EPA, 1986,
2000c) to determine risk. This pollutant-
based approach has most often been applied
for estimating exposures and risks for
specific locations or scenarios (e.g., risks
associated with a hazardous waste site).
The objective of ORD's research pro-
gram on aggregate and cumulative risk is to
provide methods, models, data, and guid-
ance for assessing human health risk so that
EPA can protect the health of the public and
environment more effectively. ORD's
research program on aggregate and cumu-
lative risk will take two approaches:
chemical-focused and population-based.
The chemical-focused approach may be
better suited to address the likely effects of a
specific source or a pollution control
strategy when the key variables associated
with that source can be well characterized
for specified human exposure scenarios. A
population-based approach may be better at
revealing total exposures and identifying
when important sources or important path-
ways of exposure may not have been identi-
fied. A population-based approach may also
be useful in assessing public health out-
comes because the objective of any control
policy is to decrease public exposure and
risk. ORD research will build on these two
approaches to develop scientifically robust
aggregate and cumulative risk assessment
methods, including how to identify impor-
tant stressors to a population, combine risk
over several stressors, define risks that
accumulate over time, and assess the inter-
action between stressors. The research
program for aggregate and cumulative
research consists of several interrelated
research efforts, all of which add critical
components to the overall aggregate/cumu-
lative risk assessment effort.
2.2.2 Research Objectives
The following research objectives pro-
vide the framework to develop an integrated
research program on aggregate and cumula-
tive risk:
Q Determine the best and most cost-
effective ways to measure human
exposures in all relevant media,
including pathway-specific measures of
multimedia human exposures to
environmental contaminants across a
variety of relevant microenvironments
and exposure durations and conditions;
Q Develop exposure models and methods
suitable for EPA and the public to assess
aggregate and cumulative risk, including
mathematical and statistical
relationships among sources of
environmental contaminants, their
environmental fate, and pathway-
specific concentrations; models linking
dose and exposure from biomarker data;
and approaches to assess population-
based cumulative risk, including those
involving exposure to stressors other
than pollutants; and
Q Provide the scientific basis to predict the
interactive effects of pollutants in mix-
tures and the most appropriate
approaches for combining effects and
risks from pollutant mixtures.
2.2.3 Research Approach
Exposure Research. One goal of ORD's
research program is to develop methods and
approaches for measuring exposures and
identify exposure factors accounting for
aggregate and cumulative exposure. In
assessing aggregate and cumulative risk, the
focus will be on measuring exposure and
2-7
-------
estimating biologically relevant doses in
exposed individuals. Considerable progress
has been made over the past two decades
toward developing personal measurement-
based methodologies for assessing human
exposures in either a population of concern,
or in the population at large. The Total
Exposure Assessment Methodology
program and the National Human Exposure
Assessment Survey (U.S. EPA, 1999b) have
demonstrated the techniques and values of
measuring personal exposures. In addition,
the CDC continues to improve their methods
for measuring pollutants and their metabo-
lites in blood and urine and have recently
begun reporting exposure data for a repre-
sentative sample of the U.S. population.
These measurement-based methods add to
our arsenal of approaches to address
aggregate and cumulative risk.
Exposure research on cumulative and
aggregate risk will build upon the problem-
driven research being conducted under other
research strategies (see Appendix A) and
focus principally on describing how people
come into contact with pollutants. As a
result of this emphasis, important compon-
ents of this research will be to: (1) identify
and characterize major factors, including
time-activity patterns, that contribute to
human variability in aggregate and
cumulative exposure, and (2) to conduct
studies to determine distributions of
aggregate and cumulative exposure for the
general population and for specific
susceptible or targeted subpopulations
(see schematic). Exposure research will
integrate an understanding of exposure
pathways and human contact with
pollutants into probabilistic human
exposure models that account for both
aggregate and cumulative exposures.
These exposure models will then be tested
against the exposure and exposure factor
data generated through targeted laboratory
and field measurement studies, including
population and epidemiological studies.
The resulting data will be used to improve
our understanding of human exposure and
refine the exposure models. The ultimate
objective of this research will be to
assemble and integrate a knowledge of
human exposures into models that describe
those exposures and to combine the source
models, the transport and fate models, and
the probabilistic exposure models into an
integrated modeling framework that can be
linked and effectively employed by the risk
assessor. The framework is designed to link
a variety of source, exposure, exposure-
dose, and dose-effect modules into a
comprehensive source-to-effects modeling
framework characterizing and assessing
user-specified aggregate and cumulative
exposures and risks. The resulting tools,
models, and framework will then be
disseminated to scientists and risk assessors
as they work to solve specific programmatic
problems as outlined in ORD's research
strategies (see Appendix A).
Exposure to Dose Research. When
exposures to an agent occur via multiple
routes, they must be converted to a common
basis, usually some measure of dose, to
evaluate the risk of aggregate and
cumulative exposure. Ideally, the common
Identify Major Factors
that Contribute to
Variability in
Human Exposures
Determine Distributions
of Aggregate and
Cumulative
Exposure in Population
Develop Probabilistic
Exposure Models
Test Models in Lab and
Field Measurement Studies
Source Models
Integrated Modeling
Framework
Dose-Effects
Models
Fate and
Transport
Models
2-8
-------
metric would be the biologically effective
dose, that is, the dose to the target organ,
tissue, cell, or molecule that causes the toxic
or adverse health effect (see schematic).
Mechanistic Data
Route
Activity
^k
Biolo
Effeoti
Patterns
Multiple Pollutants
[ S
gically
VQ Dose
\
Bion
-E
i
/
narkers ^ — "
xposure
ffect
/
Analytical
^Methods
Exposure Data
Humans
"Statistical Models
Dosimetry
Models
Risk Assessment for
Aggregate and Cumulative
Exposures
The biologically effective dose may be the
pollutant itself or one or more metabolites
and may be affected by many factors. For
example, contemporaneous exposure of a
single pollutant by more than one route can
result in different proportions of parent
compound or metabolites than would be
predicted from one route alone. The route
of exposure may also modulate the internal
dose of systemic toxicants at the target
tissue due to alterations in physiological
parameters (e.g., breathing rate due to an
irritant) or pharmacokinetic parameters
(e.g., induction of enzymes). Human
activity patterns may also affect the
biologically effective dose. A pesticide, for
example, may contact the body through
inhalation of dust from contaminated
surfaces, the diet, and as a result of hand-to-
mouth activity. People may be exposed
occupationally as well as incidently away
from their place of work. People may also
be exposed to low background levels and
also, by virtue of special intermittent
activities, to bursts of higher exposure.
Finally, the biologically effective dose may
be affected by exposure to more than one
pollutant. Multiple pollutant exposures
might change the metabolic transformation
of the pollutants in the mixture, resulting in
different biologically effective
doses than seen after exposure
to the pollutants in isolation.
Ingestion of otherwise
innocuous substances, because
of enzyme induction, might
also increase the rate of
formation of a lexicologically
relevant metabolite of a pollu-
tant of environmental concern.
Knowledge of the biologically
effective dose provides the
basis for developing dosimetry
models that can be used in
assessing risk of aggregate and
cumulative exposures.
Measuring the biologically
effective dose in humans, however, is not
easily accomplished and is therefore not
usually attempted. More often, a surrogate
for the biologically effective dose, such as
the absorbed dose (the amount of substance
crossing an absorption barrier such as the
skin, the lining of the lung, or the lining of
the gastrointestinal tract) or the level of
pollutant in human blood, urine, or other
biological tissue is measured or estimated
and used in the aggregate assessment as the
common metric. In some notable cases
(e.g., concentration of lead in the blood,
carboxyhemoglobin), the human biomarker
can also be used as a quantitative predictor
of effects. An exposure biomarker is an
exogenous substance or its metabolite(s)
measured in a compartment within an
organism, whereas an effects biomarker is a
measurable change within an organism that
can be recognized as an established or
potential health impairment. Exposure
biomarkers are actual evidence of internal
dose. However, only a few biomarkers,
such as urinary metabolites, are relatively
easy to measure in exposure field studies.
2-9
-------
With proper research, such biomarkers can
be used with pharmacokinetic models to
estimate, via a "back calculation," the
biologically effective dose and even the
exposure that occurred. Thus here, the
exposure-to-dose continuum is actually used
in reverse for "dose-to-exposure"
estimations.
Identification and characterization of
biomarkers and development of methods to
use them will be a high priority for ORD's
research on aggregate and cumulative risk.
Development of analytic methods to
measure biomarkers and methods for their
analysis and interpretation will be necessary
for exposure and dose assessment. This will
require contributions from ORD's research
on effects to provide the scientific basis for
the development of sensitive and specific
biomarkers based on mechanistic studies.
Combined with proper modeling techniques
and some knowledge of possible exposure
patterns and measurements, biomarker data
can be used to estimate dose and exposure.
Research in this area will also focus on the
development and/or implementation of
advanced statistical methods to help formu-
late and use dosimetry models for estimating
exposure from biomarkers. This is espe-
cially important as more and more
biomarker measurements are taken and their
results are made available. For example,
CDC is publishing on the internet the results
of such measurements taken in the
population. Those data, often representing
"snap-shots in time" will have to be
interpreted using a variety of modeling and
statistical tools to determine the meaning of
these data with respect to exposure and
dose.
Initially, ORD's dose research will focus
on the development of a suite of route-
specific models for use in dose-response
assessment of cumulative and aggregate
exposures. This will build upon the
dosimetry-based approach in the current risk
assessment guidelines, extend it to oral and
dermal exposures, and use it to evaluate
aggregate exposures. As the program
progresses, dose models will be expanded to
describe and predict chemical disposition
within the body resulting from aggregate
and cumulative exposures. ORD's
dosimetry models will enable users to
estimate biologically relevant doses
resulting from exposure to multiple
pollutants and multiple pathways of
exposure. The most immediate phases of
this research will concentrate on aggregate
exposures. In addressing cumulative risk,
models will be first developed for those
cases involving exposure to multiple
compounds with common modes of action.
The next phase will begin to address those
cases in which compounds may act with
different modes of action.
ORD realizes that there must be signifi-
cant integration between research on expos-
ures, dose, and effects to study the problem
of aggregate and cumulative risk adequately.
ORD has already implemented plans to
facilitate a multidisciplinary approach to this
problem. For example, scientists from
NERL, NHEERL and NCEA, as well as
scientists from the Office of Pesticide
Programs (OPP), are working on a
collaborative research project to develop
methods and models for assessing the
exposure, dose, and aggregate and cumula-
tive risk of pyrethroid mixtures. In addition,
NERL and NUEERL sponsored an
Exposure to Dose Modeling Workshop in
July 2001 to begin linking quantitative
modeling in a Human Health Risk
Assessment context. This meeting
examined a number of issues related to
source, exposure pathways, doses in
toxicology and epidemiological studies,
pharmacokinetic modeling of mode of
action, effects, and dose-response modeling.
Presentations at the meeting were followed
by a discussion of research directions and
options for linking models. Significant
2-10
-------
opportunities for collaboration and model-
sharing between principal investigators in
both laboratories were identified. Projects
have been initiated to integrate to a greater
extent the modeling efforts of the two
laboratories. Interactions between the
exposure, dose, and effects research
programs and the risk assessment methods
are also being developed.
Effects Research. The FQPA indicates
that EPA must consider the cumulative
effects of pesticides and other chemicals
having a common mode or mechanism of
toxicity. Understanding cumulative risk
requires knowledge about mechanisms or
modes of action and an understanding of
how chemicals will interact in mixtures.
The principal effects issue for cumulative
risk is the possibility that chemicals in
mixtures may interact in a nonadditive
manner. There is evidence that the
assumption of dose additivity may not hold
for all mixtures of pollutants. For example,
research has indicated that antagonism can
occur at high concentrations of some
mixtures of pollutants; whereas synergistic
interactions have been noted at the low end
of the dose-response curve for other
mixtures. Understanding the conditions
under which nonadditive interactions will
occur between pollutants is needed to
support risk assessment approaches for
cumulative exposures.
ORD's effects research program on
mixtures will test various assumptions
concerning the behavior of pollutants in
defined mixtures containing major or key
known constituents at concentration ratios
resembling real-world mixtures. To develop
quantitative models, it is crucial in these
studies to understand dose-response
behavior and the pharmacokinetic character-
istics of each pollutant. Much of this
information can be derived from projected
work on the development of methods and
mechanistically based dose-response
models. It is likely that a systematic
approach to the study of mixtures will
require the development of new
investigative tools such as genomics and
proteomics so that effects of multiple pollu-
tant interactions can be studied in rapid
fashion.
The overall approach of ORD's effects
research on chemical mixtures will be to
identify key biological processes (see
schematic) that could be used in testing for
various health endpoints and in determining
Identify Precursor or
Adverse Effects for
Assessment ^^\
Information on Mechanism
or Mode of Action
Dose-Response Assessment
of Prototypic Agents to Test
Principle of Additivity
Single vs Combined
Different Ratios
Different Order
Studies on Complex Mixtures
Principles of Chemical
Interaction
effects of pollutants based on their mechan-
ism or mode of action and environmental
relevance. Initial studies will focus on dose-
response curves for pollutants in isolation,
and then pollutants will be tested for evi-
dence of antagonism, potentiation, or syner-
gism with other pollutants in the mixture.
Two key questions are where on the dose-
response curve interactions occur and
whether interactions vary with the ratio of
the pollutants in the mixture. Additionally,
fit will be important to determine the
influence of the order of presentation of
pollutants in the mixture. Studies on
interactions between pollutants in mixtures
2-11
-------
will be used to develop principles for the
assessment of real-world mixtures.
Risk Assessment Methods. Human
populations are most frequently exposed to
multiple environmental pollutants (e.g.,
particulate matter, pesticides, microbes,
climatic stressors). Exposure to multiple
stressors could change health risks through
combining effects arising from similar
modes of action or through interactions
between nonchemical stressors that increase
or decrease the potency of environmental
agents. Research will be designed and
conducted to evaluate population-based
approaches to assess effects of total
exposures in the environment and the
interaction of chemicals with nonchemical
stressors. Because this is an emerging area,
case studies will be conducted and a concep-
tual framework will be developed incorpor-
ating results from ORD aggregate/cumula-
tive research and addressing issues of
aggregate and cumulative exposure,
mechanisms of action, and PBPK and dose-
response modeling. The objective of this
research is to develop guidance and EPA
guidelines for population-based cumulative
risk that will incorporate cumulative and
aggregate exposure to multiple stressors.
2.3 Research on Susceptible and Highly
Exposed Subpopulations
The goal of ORD's program on
susceptible and highly exposed
Subpopulations is to understand why some
people and groups are more susceptible or
highly exposed than others. Observed
variability in human responses to
environmental agents reflects differences in
biological susceptibility and exposure.
Variation in biological susceptibility
depends on intrinsic factors (e.g., life stage,
gender, genetic factors) and acquired factors
(e.g., preexisting disease, nutrition, stress,
licit and illicit drug use, cigarette smoking,
alcohol use). Variation in exposure and
dose can be influenced by many of the same
factors. In addition, factors such as occupa-
tion, location of residence, and activity
patterns that place individuals in contact
with environmental agents can cause varia-
tion in exposure. Information is needed on
how various susceptibility and exposure
factors alter responses to chemical expos-
ures. ORD research on susceptible and
highly exposed Subpopulations will focus on
three factors (i.e., life stage, genetic factors,
and preexisting disease) that have been
identified by a program office and the
scientific community as having a high
priority for risk assessment.
In addition to examining the role of each
of the three factors separately, ORD will
examine the interactions among these
factors that may result in susceptibility
greater than that caused by the presence of
only one factor. For example, a child with
autism or developmental delay might mouth
hands and toys to a greater degree than the
average child. The condition of autism
might make the child a highly exposed
individual within a group that is already
prone to higher exposures. An early
exposure might also affect the development
of the immune system, which in turn, could
make the individual more susceptible to
development of cancer later in life. Genetic
predisposition, such as lack of an enzyme
that detoxifies a particular chemical, can
make individual members of an already
susceptible life stage even more susceptible.
While life stage, genetic predisposition,
and preexisting disease provide a focus for
the strategy, there are many complex inter-
actions among risk factors that must be
considered in any research program on
susceptible and highly exposed Subpopula-
tions. Gender is not highlighted as a factor
in this research strategy, but it is recognized
that any study of the effect of life stage,
genetic factors, or disease on response to
environmental agents must understand how
2-12
-------
differences between the sexes influence
response to exposures. ORD recognizes that
stressors associated with the physical and
social environment of an individual can also
increase susceptibility to environmental
agents. In addition, factors such as socio-
economic status of individuals and
communities, access to health care,
nutrition, drug and alcohol use, and
exposure to environmental tobacco smoke
can alter the sensitivity to environmental
agents. The influence of these factors on
susceptibility and exposures associated with
life stage, genetic predisposition, and
disease will be considered in future
implementation of the strategy.
ORD will identify and study those that
are susceptible to disease that can be
induced or exacerbated by environmental
exposures. Disease endpoints that are
included within the strategy include cancer;
neurotoxicity; immune system effects;
asthma and other respiratory effects; repro-
ductive effects; and birth defects including
death, malformation, and growth alteration.
While the strategic focus is on life stage,
genetic factors, and the effects of disease on
responses to exposure, the interaction of
these factors with intrinsic factors including
gender and stressors in the physical and
social environment will be an important
consideration.
Other ORD research strategies which
address susceptible and highly exposed
subpopulations are the Strategy for Research
on Environmental Risks to Children (U. S.
EPA, 2000a) and the Asthma Research
Strategy (U.S. EPA, 200la). The influence
of life stage on responsiveness to endocrine
disrupters is also described in the Research
Plan for Endocrine Disruptors (U.S. EPA,
1998).
2.3.1 Scientific Uncertainties
Life Stage. There are specific periods or
windows of vulnerability during develop-
ment, particularly during early gestation but
also throughout pregnancy and early child-
hood through adolescence and old age, when
toxicants might permanently alter the mor-
phology and/or function of a system
(Rodier, 1980; Bellinger, et al., 1987).
Many specific adverse effects ranging from
functional impairment to anatomical birth
defects depend on an exposure at a specific
stage of development. Children may also be
more vulnerable to specific environmental
pollutants because of differences in absorp-
tion, metabolism, and excretion (NRC,
1993). In addition, children's exposures to
environmental pollutants are often different
from those of adults because of different
diets and different activities (e.g., playing on
floors and in soil and mouthing of their
hands, toys, and other objects) that can bring
them into greater contact with environ-
mental pollutants (Bearer, 1995). Because
children consume proportionately more food
and fluids, have a greater skin surface area
relative to their body weight, and breathe
more air per unit body weight than adults,
they may receive greater exposure to
environmental substances (NRC, 1993).
These health threats to children are often
difficult to recognize and assess because of
a limited understanding of when and why
children's exposures and responses are
different from those of adults.
The effect of aging on response to
environmental exposures is another area of
uncertainty based on life stage. The elderly
may respond to environmental exposures
differently than younger adults. There may
be an increased risk of cancer and degenera-
tive diseases as a function of age. The
prominence of these concerns is rapidly
elevating with the largest birth cohort in the
US, namely the "baby boomers," now
becoming senior citizens. Many of these
2-13
-------
individuals are living longer and the effect
of previous exposures may be markedly
magnified with aging. In 2002, EPA
launched an Aging Initiative designed to
study and prioritize the environmental
health risks aging citizens face. As part of
that initiative, EPA is in the process of
creating a National Agenda on the
Environment and the Aging. The intent of
this agenda is to help identify research gaps
and develop strategies that will help
scientists better understand the environ-
mental hazards affecting the health of older
Americans. EPA has held listening sessions
to solicit input from the public concerning
the development of a research program on
aging.
Genetic factors. There are a number of
genetic factors that could predispose human
subpopulations to adverse effects from
exposure to pollutants, including genetic
polymorphisms for metabolizing enzymes,
differing rates of DNA repair, and different
rates of compensation following toxic insult.
The main scientific question for this
research is whether such genetic differences
significantly influence risk at realistic, low-
dose exposures. Information on gene-
pollutant interactions as a result of long-
term exposure to environmentally relevant
concentrations of pollutants is needed.
Health status. Preexisting diseases may
influence an individual's response to
environmental toxicants by altering
xenobiotic metabolism or otherwise altering
the host's response in a synergistic, additive,
or antagonistic manner. ORD research has
shown, for example, that mice challenged
with influenza have increased mortality
from exposure to several environmental
agents including dioxin, ozone, and
ultraviolet radiation. Research is needed to
develop animal models of diseases having a
high incidence in the human population and
to determine the effects of the disease on the
dose-response curves for high priority
environmental agents (e.g., air pollutants,
pesticides).
2.3.2 Research Objectives
The Human Health Research Strategy
provides a broad framework for ORD
research in human variability. Issues
specifically related to children's risk are also
covered in more detail in the Strategy for
Research on Environmental Risks to
Children (U.S. EPA, 2000a), the Strategic
Plan for Endocrine Disrupters (U.S. EPA,
1998) and the Asthma Research Strategy
(U.S. EPA, 200la). The following research
objectives provide the framework for an
integrated research program on variability in
the human population:
Q Identify the key factors that contribute to
variability in human exposure, including
the distribution of human exposures
and behavior associated with exposure to
pollutants;
Q Improve the accuracy of dose estimation
in the general population;
Q Identify the biological basis underlying
differential responsiveness of sensitive
subpopulations of humans to pollutant
exposure; and
Q Determine how exposure, dose and
effect information can be incorporated
into risk assessment methods to account
for interindividual variability.
2.3.3 Research Approach
Exposure Research. Although an
average person may not be exposed to an
environmental agent at a level that would
cause a health concern, a small percentage
of the population may have significantly
higher exposures because proximity to
sources or activities increase likelihood of
exposure. Therefore, exposure assessments
should include distributions of exposures to
allow identification and assessment of
groups of people at risk from high-end
2-14
-------
exposures. Exposure assessments should
also account for the exposures of people
who may be especially susceptible.
Initially, ORD's exposure research will
focus primarily on children. The overall
objective is to develop a broadly applicable
probabilistic total-exposure model capable
of linking to a PBPK model to estimate
children's exposure (see schematic).
Collect Data on Exposures in
Children
^
Identify Factors Increasing
Exposure of Children
Probabilistic T otal-Exposure >^
Model
PBPK Models on
^ '
Model Estimating Children's
Exposure to Pollutants
ORD will collect data on children's
exposures and factors that influence
exposure. These data will provide input to
the development of a probabilistic model.
Status and trends in children's exposure to
environmental agents will also be character-
ized. Highly exposed subpopulations of
children will be identified, and important
sources and pathways of children's
exposures will be delineated. Residential
exposure factors for children will be
characterized by age and gender for the
national population, regional populations,
highly exposed groups, and susceptible
groups. Factors that will be characterized
include activity patterns (time spent in a
given activity and frequency of occurrence),
soil and dust ingestion rates, factors reflect-
ing transfer of environmental agents from
objects and surfaces children commonly
touch, and factors related to ingestion of
pollutant residues on surfaces. Future
research will include the aging population.
Dose Research. Dose research in the
area of susceptible subpopulations will
focus on developing probabilistic exposure
and pharmacokinetic models which estimate
doses in susceptible subpopulations,
including children and those with genetic
polymorphisms or preexisting disease (see
schematic below). This research will
provide crucial information on the
likelihood that a pollutant or its
metabolites will be present at the
target site, the concentrations in target
tissues, and whether and how the dose
varies between members of the
general population and susceptible
individuals. Measuring and modeling
the effect of susceptibility factors on
dose will help ORD design and
conduct studies of the biological
mechanisms on the cellular and
molecular levels that lead to adverse
effects. This will lead to a better
understanding of the biological bases
for differential sensitivity of
susceptible subpopulations.
D eterm inants of V ariability
C h ild r e n
G enetic Polym orphism s
Preexisting Disease
Probabilistic
Exposure
M o de 1 s
Pharm acokinetic
M o d e 1 s
M odels to Estim ate Dose or
Concentration in Target Tissue for
Susceptible Subpopulations
2-15
-------
For the near term, ORD will continue its
focus on children. Broadly applicable
PBPK models and methods will be produced
that allow better quantitative
characterizations of dose to target tissue in
developing organisms to replace default
assumptions in children's risk assessments.
Over the next two to three years, research on
the influence of aging on pharmacokinetic
parameters will increase.
The development and linkage of
probabilistic exposure and pharmacokinetic
models will provide valuable tools for
analyzing and utilizing data describing
variations in subpopulations for risk
assessment. A key step will be to establish
methods and approaches that can be applied
to both animals and humans to aid in
extrapolating from dose-response data
collected in animals to humans.
Effects Research. The main hypothesis
of the effects research on susceptible
subpopulations is that differences among
individuals (inter-individual) as well as the
variability in an individual's responses over
time (intra-individual) are due to biological
variability. ORD's effects research on
susceptible subpopulations will focus on
developing biological models that describe
the differential sensitivity of various
subpopulations for risk assessment,
especially the influence of life stage, genetic
factors and preexisting disease on
toxicological outcome or adverse health
effect (see schematic).
Life Stage. There is now evidence that
differential sensitivity of very young (early
postnatal, children) and elderly individuals
to certain pollutants may be related to phar-
macokinetic factors. In conjunction with
ORD's dose research program, the effects
research program will develop longitudinal
pharmacokinetic information for prototypic
environmental agents from the prenatal and
early postnatal period to senescence in
laboratory animals to determine how
specific xenobiotic metabolizing enzymes
change as a function of lifestage. Research
will also determine how biological changes
specific to some life stages (e.g., prolifera-
tive phase during development) can increase
the risk of certain pollutants. Identification
of such pharmacodynamic factors is crucial
for the protection of susceptible subpopu-
lations at different stages of development.
As in the case of research on exposure, the
initial emphasis will be on children.
Research in this area will begin to focus on
the elderly following the development of a
national research agenda on aging. An
objective of ORD's effects research will be
to link life stage-related effects at the tissue,
organ, and system levels with the underlying
effects at the cellular and molecular levels
Mechanism or
Mode of Action
Pharmacodynamic Studies on
Differential Sensitivity in Animals
-Life Span
-Genetics
-Disease
Biological Basis
Sensitivity to Spec
Interspecies I
for Differential
ific Chemicals and
ixtrapolation
Data from Clinical
-1
and Epidemiological
Studies
Biological Models Describing Differential
Responsiveness of Susceptible Subpopulations
and to develop the first-generation of
biologically based predictive models.
Information from dose-response,
pharmacokinetic, and mode-of-action
studies in animals will be incorporated into
models that more accurately predict
children's risks.
2-16
-------
Effects research currently focuses on the
effects of pollutants on early stages of
development. These endpoints include birth
defects, reproductive, neurodevelopmental,
immunological and endocrine disorders. As
more is known about the effects of pollu-
tants on infants and children, research
efforts will begin to examine the influence
of early exposure to pollutants on health
status later in life. Multidisciplinary
approaches will be developed in animal
models to examine the effect of environ-
mental pollutants on the aging process and
to develop predictive models that can be
incorporated into the risk assessment
process.
While animal studies are useful in
understanding mechanisms of toxicity and
are frequently used to predict adverse effects
in humans in the absence of human data,
animals may not always be appropriate
models for humans or for particular life
stages. Extrapolation between animals and
humans can be uncertain. Coordinating the
development of animal models in ORD will
be closely coordinated with human research.
Epidemiological studies will be crucial
to understanding whether certain groups are
more susceptible to environmental contami-
nants than others, and such studies will be
conducted by all Laboratories and Centers in
ORD. Hypothesis-based human epidemio-
logical and clinical studies will be necessary
to identify and confirm that adverse effects
occur in humans, to identify risk factors, to
develop dose-response relationships in
humans, and to improve extrapolations from
animal data to humans. Human studies will
be conducted as needed for high-priority
environmental agents and to assist in model
development and validation.
In the Children's Health Act of 2000,
Congress directed the National Institute of
Child Health and Human Development to
establish a consortium of federal agencies,
including EPA and the CDC, to design and
implement a National Children's Study.
This study will follow a cohort of children
from as early in pregnancy as possible to
adulthood to evaluate the effects of chronic
and intermittent exposure on child health
and human development. The goal is to
enroll at least 100,000 children in the study.
Exposure information will be collected for
preconception exposures, at several times
during pregnancy, and at several ages after
birth; and outcome data will be collected
during pregnancy, infancy, childhood, and
beyond, perhaps focusing on developmental
milestones of potential susceptibility in each
of several age ranges. Biological specimens
from the parents and children will be
collected. Children will be followed at least
through their primary school years and,
preferably, into adulthood. ORD is
participating in the planning and design of
the study and developing and testing
methods for data collection.
Genetic Differences. ORD's effects
research on genetic influences will address
the hypothesis that individuals harboring
genetic polymorphisms in metabolic genes
may have increased vulnerability to health
effects following exposure to some pollu-
tants. ORD research has shown, for
example, that people who are phenotypic for
rapid acetylation have higher levels of
urinary mutagens following exposure to
heterocyclic amines in food. The main
scientific question for this research is
whether such genetic differences signifi-
cantly influence risk. This research will
focus on the influence of genetic factors on
long-term exposure to low levels of pollu-
tants. The role of other genetic factors in
susceptibility, such as differing rates of
DNA repair and compensatory responses to
toxic insult, will also be investigated.
Genotype can alter exposure and phar-
macokinetic and pharmacodynamic factors.
If individuals are genetically predisposed to
2-17
-------
autism and if autism results in higher
mouthing of objects and hands, then genetic
factors could lead to higher exposures.
Genotype can also alter transport proteins,
clearance, metabolic profiles, and target
organ sensitivity. Therefore, research will
be conducted with the goals to develop
exposure and pharmacokinetic and pharma-
codynamic data and models with which to
estimate dose and response in individuals
with genetic polymorphisms. The use of
these data in dose-response assessment will
be similar to that described in the discussion
of life stage.
Disease. Preexisting diseases may
influence the response to environmental
toxicants by altering xenobiotic metabolism
or otherwise altering the host's response in a
synergistic, additive, or antagonistic
manner. Preexisting dysfunction of a target
organ may alter the type or severity of a
chemically induced adverse effect on end
organ function, such as the effect of air
pollutants on lung function in individuals
with asthma or the effect of exposure to a
neurotoxicant on children with
neurobehavioral dysfunction. In the near
term, research will focus on the
development of animal models of diseases
having a high occurrence in the human
population (e.g., asthma, bronchitis, hyper-
tension) and on determining the effects of
the disease on the dose-response curves of
high priority environmental agents (e.g., air
pollutants, pesticides). Mechanistic research
will establish animal models that employ
specific host traits that are characteristic of
the disease and represent "risk factors" for
increased sensitivity to chemicals. Once
effects have been established using these
animal models, studies will be conducted to
extrapolate from animal data to human
effects and across levels of biologic organi-
zation. Epidemiological studies will also be
used to identify possible associations
between exposure to a specific pollutant and
manifestation of a disease. Such associa-
tions will then be tested in in vitro or in vivo
animal models. Data derived from these
studies can be used to assess the possible
increased risk to pollutant exposure in
individuals with preexisting disease.
Research on health status will continue to
focus on asthma and other respiratory
diseases and air pollution; studies on other
diseases and pollutant classes will be
conducted as time and resources allow.
Effects research will be conducted with
a goal to develop pharmacokinetic and phar-
macodynamic data and models with which
to estimate dose and response in individuals
with a preexisting disease. ORD will
examine the physiological and biological
changes associated with common diseases
and how these changes influence the phar-
macokinetics and pharmacodynamics of
environmental agents. This information will
lead to improved models and methods for
accounting for the effects of disease on
individuals' responses to environmental
agents.
Risk Assessment Methods Research.
The results of ORD's research in exposure,
dose, and effects, along with research
supported by other government agencies and
nongovernmental sponsors, will be used to
develop improved methods, models, tools
and databases for risk assessments of
susceptible and highly exposed subpopula-
tions (see schematic).
Biological Models on
Phamaoodynamic
Differences
PBPK Models on
Susceptible
Subpopulations
\^
Exposure Variables
/
/
Improved Methods, Models, Tools
and Databases for Risk Assessment
of Susceptible and Highly
Exposed Subpopulations
2-18
-------
ORD will use pharmacodynamic data and
PBPK models from research on effect and
dose to develop better dose-response
methodologies to account for susceptibilities
of various life stages and to evaluate the
adequacy of the current default uncertainty
factor of 10 in accounting for human
variability for noncancer health effects.
ORD risk assessment methods research will
also analyze data on exposure factors,
human activity patterns, and environmental
concentrations, including those generated by
the exposure research program on pesticides
and air pollutants, to quantify the important
factors used in exposure assessment and to
evaluate representativeness of the data based
on factors such as life stage, genetics, and
pre-existing disease. Databases on
physiological and pharmacokinetic factors
for various life stages will be developed to
aid in development and implementation of
PBPK models. Dose-response
methodologies for specific life stages,
accounting for differences between children
and adults, will be developed. Distributions
of exposure factors measured in ORD
studies will be incorporated into the
Exposure Factors Handbook (U.S. EPA,
1997b, 2000b). Finally, ORD will develop
guidance for performing risk assessments
for children, the elderly, and those with
preexisting diseases and for taking into
account genetic variation in risk assessment.
2-19
-------
3. RESEARCH TO ENABLE EVALUATION OF PUBLIC
HEALTH OUTCOMES FROM RISK MANAGEMENT ACTIONS
ORD's current Human Health Core
Research Program began a new effort to
evaluate the public health outcomes of EPA
policies in 2001. The issue of "accounta-
bility," i.e., making sure that our research
and pollution control programs produce
measurable benefits in public health, is an
important one. In order to better protect the
country's air, water, and land resources,
EPA must go beyond its current reliance on
process indicators such as decreased emis-
sions or discharges and measure actual
changes in the status of ecological condition
and human health. This issue is funda-
mental to the first goal of ORD's Strategic
Plan, i.e., to support EPA's mission "to
protect human health and to safeguard the
natural environment-air, water, and
land-upon which life depends." With the
advent of the GPRA and calls for EPA to
stress and demonstrate outcome-oriented
goals and measures of success, research is
needed to enable evaluation of actual public
health outcomes from risk management
(emission or exposure control) actions.
Furthermore, in 2001, EPA announced an
"Environmental Indicators Initiative" to
improve the EPA's ability to report on the
status of and trends in environmental condi-
tions and their effect on human health and
the nation's resources. ORD will work with
the Office of Environmental Information to
develop and publish a "State of the Environ-
ment Report" using available national level
data and indicators to describe human health
environmental conditions and human health
concerns. Part of that report will identify
data gaps and research needs and discuss
opportunities for partnering with other
research organizations in filling those gaps.
Estimating public health benefits of EPA
regulatory decisions and rule-making or, in
a more general sense, evaluating public
health outcomes from risk management
actions will be a challenge. It will involve a
number of disciplines grounded in both the
physical and social sciences and
increasingly must take into account the
economic and behavioral aspects of human
decision-making. The remainder of this
chapter outlines the basic conceptual
strategy of an emerging research program to
evaluate the environmental public health
consequences of EPA regulations,
regulatory or non-regulatory programs, or
other risk management activities. Both the
precise methods to be used and the
outcomes that will be studied have not yet
been chosen.
Evaluating public health outcomes from
risk management actions is clearly linked to
assessing human health risks. EPA risk
assessors and risk managers must consider
the uncertainties associated with the risk
assessment process, including the upper as
well as the lower bounds of such uncer-
tainty. Coupled with these uncertainties is
the fact that EPA very often estimates the
future benefits of public health outcomes in
a politically charged environment. Depend-
ing on the desired human health protection
endpoint, final decisions often rest with
national and state policy makers and deci-
sion officials. These officials take scientific
findings into account along with a number
of other considerations that assist them in
making more informed public policy
decisions.
Generally, EPA has not prepared retro-
spective evaluations examining whether the
intended benefits in protecting public health
were realized once an EPA decision had
been in place for a period of time. One
exception to this was the decision to ban
lead in gasoline and other products, the
subsequent tracking of reduced blood-lead
levels in children as a result of the ban, and
3-1
-------
the epidemiological studies to confirm the
linkage between elevated blood level levels
and reduced cognitive development in
children. The confounding influences of
various factors (e.g., age of exposure, dura-
tion of exposure, exposure to other pollu-
tants alone or in complex mixtures) offer
challenges at every turn in evaluating public
health outcomes. As EPA develops and
implements a research program advancing
the evaluation of public health outcomes,
either prospective or retrospective, partici-
pants and observers must recognize that the
program will take years or perhaps decades
to develop and fully implement. It will
involve a number of organizations both
within and outside of EPA working in
partnership to collect and analyze data and
then to use that data in methodologies and
tools to objectively determine the effective-
ness of risk management decisions on public
health outcomes.
The Presidential Commission on Risk
Assessment and Risk Management (1997)
has supported the need for EPA to measure
the effectiveness of public health interven-
tions (see text box).
The Presidential Commission on Risk Assessment
and Risk Management points out the need for
progress in several scientific areas "...if we are to
improve our ability to implement and measure the
effectiveness of public health interventions.
Specifically, we need to:
(1) Link studies of exposure and studies of adverse
health or ecological outcomes;
(2) Determine regional differences in disease
prevalence and disease incidence trends and risk
factors;
(3) Develop good baseline and surveillance
information about incidence rates of diseases
specifically linked to environmental causes; and
(4) Identify the most important environmental
causes of diseases." (page 47, Vol. 1)
The National Research Council (1997) also
noted a lack of consensus concerning
appropriate indicators of health status that
could be used to measure the performance of
environmental health programs. This led
the Council of State and Territorial
Epidemiologists, the CDC, the Agency for
Toxic Substances and Disease Registry, and
EPA to begin developing a set of public
health indicators to track adverse health
events related to the environment. The Pew
Environmental Health Commission (Pew,
2000) has also recommended a nationwide
tracking of priority chronic diseases such as
asthma and respiratory diseases and of
exposures to environmental pollutants such
as polychlorinated biphenyls, metals, and
pesticides.
Chapter 2 of the Human Health
Research Strategy sets forth priorities for
improving the science of human health risk
assessment. These improvements will result
in more effective and longer lasting risk
management actions and will contribute to
public health outcomes that can be achieved.
Chapter 3 describes research enabling more
informative and reliable evaluations of
public health outcomes (e.g., improved
estimates of actual reductions in risks to
public health via exposure and effects
data) from risk management actions.
These two chapters taken together will
provide the foundation for successful
implementation of the Human Health
Research Strategy in the years to come.
3.1 SCOPE AND DEFINITIONS
As discussed in Chapter 1, there are
great similarities in information needs for
risk assessment and risk management.
This is because understanding the
efficacy of an EPA decision requires a
comparative analysis of risks before and
after implementation of risk management
actions (see Figure 3-1). At the same
time, various risk management actions
3-2
-------
Health/Exposure Research
• Laboratory Studies
• Human Studies
• Population/Field Studies
• Models
I
Risk Assessment
• Hazard/Dose-Response
Assessment
• Exposure Assessment
• Risk Characterization
I
Evaluation of Health Outcomes
• Health Indicators
• Exposure Indicators
Risk Management Actions
I
Risk Management Research
Figure 3-1. Role of analysis of health outcomes in the risk management decision process.
must be applied within the framework of
maximum achievable risk reduction that is
efficient, cost-effective, and long-lasting.
Important issues need to be addressed that
require research targeted at the most robust
possible evaluation of public health
outcomes from risk management actions.
This chapter stresses the identification
of existing information and the creation of
new information that can be used in
evaluating public health outcomes from
risk management decisions. Reflecting
the close relationship between risk
assessment and risk management, this
public-health-outcomes research program
is included in the Human Health Research
Strategy for two reasons: (1) the need to
link more closely risk assessment and risk
management so as to improve human
health risk assessments and (2) the need to
improve the scientific basis for evaluating
public health outcomes from risk
management actions.
It is essential for the research described
in this chapter to be based upon a common
set of definitions. Haddix and others in their
Prevention Effectiveness: A Guide to
Decision Analysis and Economic Evaluation
(1998) offer a set of useful definitions
adopted for this research strategy (see text
box).
Definitions of Key Terms (Haddix et al., 1998)
Effectiveness: The improvement in health outcome that a
prevention strategy can produce in typical community-based
settings (p. 146).
Efficacy: The improvement in health outcome that a
prevention strategy can produce in expert hands under ideal
circumstances (p. 146).
Outcome Measure: The final health consequence (e.g.,
cases prevented) on an intervention (p. 149)
3-3
-------
The remainder of this chapter discusses the
scientific uncertainties underlying the
evaluation of public health outcomes from
risk management actions and describes the
research approach used to meet the
objectives of ORD's public-health-outcomes
research program.
3.2 SCIENTIFIC UNCERTAINTIES
The basic philosophy behind the EPA's
public health policies is that regulatory or
other risk management actions are taken
with the intent of preventing or reducing
releases of pollutants of concern. This
philosophy assumes that exposure preven-
tion or reduction will lead to measurable
reductions in specific human health effects.
However, actual reductions in health effects
will depend on the proportional relation-
ships between the pollutant releases and
health risks from a given source as well as
whether the health risks are influenced by
other sources and factors not being con-
sidered. The degree of certainty and
directness of the links between source,
exposure, and effect influence the validity of
this assumption. Behavior of individuals
and communities in reducing risk are addi-
tional important variables.
Unfortunately, in most cases, this link-
age has a very poor quantitative scientific
foundation; and health-protective default
assumptions are generally used in cases of
uncertainty or lack of information. If this
linkage were better forged scientifically, the
predictability of risk management action
effectiveness would be more accurate. Even
so, actual effects will need to be measured
to evaluate whether the predictions (or the
prediction approach) are correct. The
optimal approach is to compare a health risk
assessment before and after the risk manage-
ment action has been employed. This is,
however, a very complex and challenging
undertaking because a systematic evaluation
framework does not always exist given the
unethical nature of randomized controlled
clinical trials in humans with toxic agents.
Prospective assessments of risk often use
multiple approaches with varying degrees of
sensitivity, uncertainty, and reliability.
Furthermore, even if prospective assess-
ments were reliable, they may not be
suitable for a retrospective analysis. For
example, an epidemiological study with
sufficient sensitivity for prospective risk
assessment may not have the statistical
power to detect the expected risk reductions
in the size of the community affected. In
addition, if the expected public health out-
come is the lessening of a chronic effect
(e.g., cancer), it may take many years to be
detected with current risk assessment
approaches that use cancer incidence as the
outcome. Finally, some risk management
actions create multiple and perhaps
disparate benefits and possibly unintended
consequences. This causes great difficulty
in the analysis of management actions
because the unintended effect has to be
identified and evaluated.
Long-Term Goal: Provide the scientific
understanding and tools to assist EPAand others
for use in evaluating the effectiveness of public
health interventions from risk management
actions.
Key Scientific Questions: How can the most
effective tools, systems, methods, and models be
identified, discovered, or developed and
integrated into a decision-making framework, to
assist federal, state and local decision-makers in
evaluating changes in public health as a result of
risk management actions? What is the ability of
this framework to quantify such changes
accurately?
3-4
-------
3.3 SCIENTIFIC OBJECTIVES
Two research objectives were developed
based on the three research issues described
below. The research questions were
developed in accordance with the Long-
Term Goal and Key Scientific Questions
described in the text box on the previous page.
The research questions serve as the foundation
upon which to develop a coherent
framework for an integrated research
program and include the following:
D The kinds of policies, regulations, or
actions that should be evaluated to
determine the efficacy of risk
management actions;
D The approaches available to address the
effectiveness of risk management
actions on public health; and
D The improvements needed to the
approaches and the development of a
useful framework for evaluating public
health outcomes.
Admittedly, achieving the long-term
goal and answering the key scientific and
associated research questions will take
considerable time and effort. ORD's three
national Laboratories and two national
Centers have agreed to work on this public-
health-outcomes research program together;
however, neither ORD nor EPA proposes to
undertake this research alone. The research
program described here will be a daunting
undertaking and one that must rely on the
contributions and collaboration with a
number of different organizations. It will
require feedback loops, engagement, and
partnering with other organizations both
within and outside EPA if it is to succeed.
In close collaboration with research
partners, ORD's research will provide the
scientific understanding and tools to assist
EPAand others in evaluating public-health-
outcomes resulting from risk management
actions. The new public health outcomes
research program is designed to address the
long-term goals and key scientific questions
in a stepwise fashion from reductions in
release, through reductions in exposure, to
improvements in public health. It is not
designed to be an expansion of the EPA's
epidemiological research program; but will
rely on collaborations with, and data and
information from, other federal, state, and
public health organizations. Ultimately, the
tools, systems, methods, models, and the
framework within which they operate should
measure or reliably estimate changes in
human health risks with a known level of
precision and accuracy. This precision and
accuracy should be sufficient to allow EPA
to determine how its regulatory decisions
and risk management actions contributed to
those changes. Two specific objectives of
ORD's research program emerge:
D Establish linkages between sources,
environmental concentrations, exposure,
effects, and effectiveness such that a
change in a public health outcome
consequent to a risk management action
can be determined by measuring or
modeling any one of these linked steps;
and
D Improve tools, systems, methods, and
models by which EPA and others can
measure or predict changes in public
health outcomes following risk
management actions.
It should be noted that a substantial part
of the research on the complex relationship
between sources and environmental quality
(i.e., fate, transport, and transformation) is
contained within problem-driven research
programs (e.g., particulate matter, air toxics,
hazardous waste; see Appendix A).
Research on the effectiveness of public
health outcomes will provide the linkages to
these other related research programs.
General precedents indicate the feasi-
bility and utility of meeting these two
3-5
-------
objectives. For example, effectiveness
evaluations have been conducted for diverse
risk management actions (e.g., for pharma-
cologic therapy, vaccine efficacy, and
smoking cessation). These evaluations are
becoming more commonplace, and several
groups have attempted to provide guidance
for the conduct of such studies (Gold et al.,
1995; Graham et al., 1998; Haddix et al.,
1998).
3.4 RESEARCH APPROACH
In developing research program priori-
ties and a deeper understanding of the
relationships between risk management
actions and public health outcomes, it will
be necessary to select cases to study based
on the suite of risk management actions that
might be employed by the EPA. A decision
on the appropriate number and scope of the
case studies will be made after further delib-
erations in workshops and other fora both
internal and external to EPA. Particular
emphasis will be placed on policies, regula-
tions, or actions attendant to risk manage-
ment that EPA has developed, is developing,
or may need to develop within the next 10
years. This type of approach will require
close collaboration with EPA's Program and
Regional Offices. Study sites and the
selection of appropriate research approaches
will vary depending upon the environmental
exposures and effects of interest.
To ensure full coverage of the possible
risk management alternatives, classes of risk
management actions will be identified as the
first step in the case study process. These
classes of action include, but are not limited
to, those that reduce exposure to pollutants
currently in the environment, dispose of or
redistribute substances currently in the
environment, and those that license (or
allow) new substances into the environment
or allow levels of substances already in the
environment to be increased. Coupled with
these classes of risk management actions
will be an identification of their implications
for evaluating public health outcomes.
Efforts to ascertain the effectiveness of
risk management actions will depend on the
selection of pertinent research approaches
and appropriate indices of public health
exposure and effects outcomes. An
evaluation of the public health outcome of a
risk management decision should answer
two questions:
D Did the risk management action actually
prevent, reduce, eliminate, or modify
exposure to the pollutants of concern?
D Did this prevention, reduction,
elimination, or modification result in
disease prevention and improved public
health?
Four approaches might be used to assess
public health outcomes, including health
outcome/effect studies, population exposure
studies, field sampling of environmental
media, and measuring changes in source
emissions. Coupled with this will be the
need to investigate and evaluate the
performance of models used to estimate
outcomes when measurement data may be
inaccessible or too costly to collect except
as a representative sample. These
approaches are ordered in terms of their
ability to determine human exposures and to
link them with public health outcomes;
however, this ordering does not mean that
an approach listed before another approach
is necessarily more feasible. Using these
approaches effectively in evaluating public
health outcomes from risk management
actions will require linking them in the
development of a framework or model.
Each of these areas can be improved, in
some cases as a result of the risk assessment
research program discussed in Chapter 2.
However, there are some special needs for
evaluating regulatory efficacy for public
health protection. Thus, a careful analysis
and prioritization of the approaches vis-a-vis
3-6
-------
the risk management action classes
described above are essential.
Although the above approaches are
listed discretely, perhaps the greatest
challenge of the public-health-outcomes
research program will be to elucidate the
linkages among them. Ultimately, this will
vastly increase the feasibility and accuracy
of both prospective and retrospective risk
assessments. Given the immense number of
scenarios to be evaluated, models of this
process are needed. Such models are under
development as part of the core research
program described in Chapter 2, but
additional models are likely to be required
that incorporate the special needs of an
retrospective assessment and more
thoroughly link the approaches under
consideration.
To assess the strengths and weaknesses
of evaluations of public health outcomes
from risk management actions, a logical first
step will be to use existing approaches and
to evaluate available databases that compile
pollutant release information and environ-
mental concentrations, health endpoints, or
both. Appendix E lists some databases and
other sources that contain information that
could be used to correlate health endpoints
with concentrations of pollutants. Such an
exercise will likely identify priorities for
future research. Better ways to measure
changes in effects (or in indicators of
effects, exposure, indicators of exposure,
environmental concentrations, or source
strength) are needed, together with programs
to measure the effects before and after
implementation of the EPA's decisions.
Risk management tools are needed that
express the EPA's understanding of the
cost-effectiveness and long-lasting nature of
risk management actions and that convey
that understanding to other regulatory
offices, the regulated community, and the
public. Finally, a framework to link models
all the way from source to human health
effects provides more confidence in
exposure-dose-response relationships
through a thorough understanding of the
critical processes within, and linkages
between, each component of the human
exposure-dose-response sequence.
3.5 RESEARCH IMPLEMENTATION
The ultimate goal of ORD's new public-
health-outcomes research program is to
provide a set of fully developed frameworks
and a suite of technical tools, systems,
methods, and models that assist EPA and
others in evaluating public health outcomes
from risk management actions. The overall
intention is to quantify environmental public
health outcome trends that could (or should)
change in response to risk management
decisions (e.g., regulations, implementing
emission control technologies) that affect
the environmental exposures that are the
primary risk factors in the causality of the
outcome. The research program will require
the full participation and active engagement
of stakeholders at all levels, both internal
and external to the EPA. It must leverage
the research program with other public- and
private-sector organizations involved in
similar or compatible efforts since that is the
only way it will succeed. The long-term
goal to provide the scientific understanding
and tools to assist EPA and others in
evaluating the effectiveness of public health
outcomes resulting from risk management
actions is extremely ambitious and research
in this area will proceed in a step-wise and
incremental fashion as described below.
Developmental phase. This phase will
provide a comprehensive state-of-the-
science evaluation of currently available
domestic and international tools, systems,
and methods, along with frameworks that
are being, or could be, used in evaluating
public health outcomes from a variety of
risk management actions. It will of
necessity partner with EPA Program and
3-7
-------
Regional Offices and will seek to engage
organizations outside EPA that are
positioned to engage in a public health
outcomes research program.
Investigation phase. This phase will
implement a detailed multiyear research
plan for improving various tools, systems,
and methods (existing and new) to evaluate
public health outcomes from risk manage-
ment actions. A preliminary compendium
of tools, systems, and methods, along with
selected framework(s), will be developed for
the multiyear plan. Pilot investigations and
studies on evaluations of health and expos-
ure information will also be conducted,
leading to further refinements of the
frameworks.
Delivery phase. This phase will provide
a set of fully developed frameworks and a
suite of technical tools, systems, and
methods for use by various stakeholders.
This compendium will be closely coupled
with illustrations and training on its use,
along with case studies targeting decision-
makers at multiple levels. As discussed
above, a near-term objective of this research
is to develop a framework and a multiyear
plan for undertaking research on evaluating
public health outcomes from risk
management actions. Recommended next
steps include the following:
D Conduct workshops, in consultation with
federal, regional, state, and local
decision-makers and other interested
parties, to develop a comprehensive
state-of-the-science evaluation and to
identify the elements of a possible
framework(s) for evaluating public
health outcomes from risk management
actions;
D Describe a set of specific cases/situa-
tions that are potential targets for case
studies (including rationale) for
evaluating public health outcomes from
risk management actions;
D Through ORD's STAR program, issue a
request for assistance for the develop-
ment of statistical techniques using
environmental and human health data in
evaluating public health outcomes and
conduct case studies to test these
techniques;
D Assess state-of-the-science approaches
for evaluating how human health is
impacted by risk management actions;
D Identify the policies and regulations that
would most likely benefit from the use
of a framework and set of tools that
evaluate public health outcomes from
risk management actions;
D Understand how various decision-
makers at the national, regional, state,
and local levels currently use, or might
use in the future, various frameworks
and tools for evaluating public health
outcomes from risk management
actions; and
D Identify a set of environmental health
indicators that can be used to evaluate
the effectiveness of risk management
actions on public health.
Components of the research program
must address such factors as the likelihood
for case studies to be informative and useful
and the composition of research designs to
achieve the desired long-term goals of the
research program. It will be critical that
pilot studies fill evidence gaps that provide
cause and effect linkages between toxicants
and environmental public health outcomes.
Clearly, this new research program on the
evaluation of public health outcomes of risk
management activities is an undertaking too
broad for EPA alone. While both the pre-
cise methods to be used and the outcomes
which will be studied have not been
decided, we expect much of the data,
particularly on the public health side, will
have to come from collaborations with our
sister agencies such as the CDC. A multi-
disciplinary approach to tying emissions,
fate, transports of toxins; environmental
3-8
-------
exposures; biomarkers of exposure; and
early indicators of disease, as well as
morbidity and mortality, requires the
unification of many scientific disciplines.
We expect broad teams of researchers, from
several ORD laboratories and partners from
other federal agencies will work together to
carry our this research on an case-by-case
basis.
3-9
-------
4. REFERENCES
Bearer C. 1995. How are children different from
adults? Environ Health Perspect 103 (Suppl. 6):
7-12.
Bellinger D, Leviton A, Waternaux C, Needleman
H, RabinowitzM. 1987. Longitudinal analysis of
prenatal and postnatal lead exposure and early
cognitive development. New Eng JMed 316:
1037-1043.
Centers for Disease Control and Prevention (CDC).
2001. National Report on Human Exposure to
Environmental Chemicals. National Center for
Environmental Health, CDC, Atlanta, GA
Centers for Disease Control and Prevention (CDC).
2003. Second National Report on Human Exposure
to Environmental Chemicals. National Center for
Environmental Health, CDC, Atlanta, GA
Food Quality Protection Act (FQPA). 1996. Federal
Insecticide, Fungicide and Rodenticide (FIFRA) and
Federal Food, Drug and Cosmetic Act (FFDCA) As
Amended by the Food Quality Protection Act
(FQPA) of August 31, 1996; U.S. EPA, OPP,
document #730L97001, March 1997.
Gold MR, Siegel JE, Russel LB, Weinstein MC.
1995. Cost Effectiveness in Health and Medicine.
New York: Oxford University Press.
Graham JD, Corso PS, Morris JM, Segui-Gomez M,
Weinstein MC. 1998. Evaluating the cost-
effectiveness of clinical and public health measures.
Ann Rev Public Health 19: 125-152.
Haddix AC, Teutsch SM, Shaffer PA. 1998.
Prevention Effectiveness: A Guide to Decision
Analysis and Economic Evaluation. New York:
Oxford University Press.
International Programme on Chemical Safety (IPCS).
1999. IPCS Workshop on Developing a Conceptual
Framework for Cancer Risk Assessment. 16-18
February, Lyon, France.
National Research Council (NRC). 1983. Risk
Assessment in the Federal Government: Managing
the Process. Washington, DC: National Academy
Press.
National Research Council (NRC). 1993. Pesticides
in the Diets of Infants and Children. Washington,
DC: National Academy Press.
National Research Council (NRC). 1994. Science and
Judgment in Risk Assessment. Washington, DC:
National Academy Press.
National Research Council (NRC). 1997. Building a
Foundation for Sound Environmental Decisions.
Washington, DC: National Academy Press.
Pew Environmental Health Commission. 2000.
America's Environmental Health Gap: Why the
Country Needs a Nationwide Health Tracking
Network. Baltimore, MD.
Presidential Commission on Risk Assessment and
Risk Management. 1997. Volume 1: Framework for
Environmental Health Risk Management. Volume 2:
Risk Assessment and Risk Management in Regulatory
Decision-Making.
Rodier PM. 1980. Chronology of neuron
development: animal studies and their clinical
implications. Dev Med Child Neurol 22:525-545.
Schlosser PM, Bogdanffy MS. 1999. Determining
modes of action for biologically based risk
assessments. Regul Toxicol Pharmacol 30:75-79.
U.S. Environmental Protection Agency (U.S. EPA).
1986. Guidelines for health risk assessment of
chemical mixtures. Federal Register 51:34014.
U.S. Environmental Protection Agency (U.S. EPA).
1995. Science Advisory Board, Committee on Indoor
Air Quality and Total Human Exposure. Human
Exposure Assessment: A Guide to Risk Ranking,
Risk Reduction, and Research Planning. U.S.
EPA-SAB-IAQC-95-005.
U.S. Environmental Protection Agency (U.S. EPA).
1996. Proposed guidelines for carcinogen risk
assessment. Federal Register 61:56274-56322.
U.S. Environmental Protection Agency (U.S. EPA).
1997a. Update to ORD 's Strategic Plan. Office of
Research and Development. Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA).
1997b Exposure Factors Handbook. Office of
Research
and Development. Washington, DC. EPA/600/P-
95/002Fa.
U.S. Environmental Protection Agency (U.S. EPA).
1998. Strategic Research Plan for Endocrine
4-1
-------
Disrupters. Office of Research and Development.
Washington, DC. EPA/600/R-98-087.
U.S. Environmental Protection Agency (U.S. EPA).
1999a. Guidelines for Carcinogen Risk Assessment.
Risk Assessment Forum. U.S. EPA, NCEA-F-0644.
U.S. Environmental Protection Agency (U.S. EPA).
1999b. National Human Exposure Assessment
Survey. National Exposure Research Laboratory,
National Center for Environmental Assessment,
Office of Research and Development. Research
Triangle Park, NC.
U.S. Environmental Protection Agency (U.S. EPA).
2000a. Strategy for Research on Environmental Risks
to Children. Office of Research and Development.
Washington, DC. EPA/600/R-00/068.
U.S. Environmental Protection Agency (U.S. EPA).
2000b. Child-Specific Exposure Factors Handbook.
External Review Draft. Office of Research and
Development. NCEA-W-0853
U.S. Environmental Protection Agency (U.S. EPA).
2000c Supplementary Guidance for Conducting
Human Health Risk Assessment of Chemical
Mixtures. Risk Assessment Forum. EPA/630/R-
00/002.
U.S. Environmental Protection Agency (U.S. EPA).
2001a. Asthma Research Strategy. Draft. Office of
Research and Development. Research Triangle Park,
NC.
U.S. Environmental Protection Agency (U.S. EPA).
200 Ib. Strategic Plan for the Office of Research and
Development. Washington, DC.
4-2
-------
APPENDIX A
ORD Research Plans and Strategies
Final Research Plan for Microbial Pathogens
and Disinfection By-Products in Drinking
Water (U.S. EPA, 1997)
This research plan describes ORD's research to
support EPA's drinking water regulations
concerning disinfectants, disinfection byproducts,
and microbial pathogens. The research plan
identifies key scientific and technical information
gaps and provides guidance to both intramural
and extramural research programs regarding
priorities and sequencing of research.
Research Plan for Arsenic in Drinking Water
(U.S. EPA, 1998a)
This research plan provides guidance to improve
the scientific understanding of health risks
associated with arsenic in drinking water and to
support improved control technologies for water
treatment.
Strategic Research Plan for Endocrine Disrup-
tors (U.S. EPA, 1998b)
This research plan addresses research needs of
biological effects for human health and wildlife
and exposure assessment of endocrine disrupters.
Integration of effects and exposure research is
emphasized to provide a complete analysis of risk.
Airborne Particulate Matter Research Strat-
egy (U.S. EPA, 1999)
This research strategy describes health, exposure,
risk assessment, and management research on
particulate matter to support EPA's review and
implementation of the National Ambient Air
Quality Standards.
Strategy for Research on Environmental Risks
to Children (U.S. EPA, 2000a)
This research strategy describes future directions
and priorities of ORD's program to reduce
uncertainties in EPA risk assessments for
children, leading to effective measures to prevent
and/or reduce risk.
Mercury Research Strategy (U.S. EPA, 2000b)
This strategy presents the scientific questions and
research goals and priorities for EPA's research
program on mercury.
Asthma Research Strategy (U.S. EPA, 2000c)
This strategy describes the research directions and
priorities to improve the scientific understanding
of environmental factors underlying increased risk
for asthma and to develop more effective risk
management control technologies to reduce and
prevent asthma cases.
Air Toxics Research Strategy (U.S. EPA,
2000d)
This strategy presents research approaches and
objectives to improve the scientific and technical
knowledge base for the assessment and manage-
ment of health risks from hazardous air pollutants.
A-l
-------
Drinking Water Contaminants Candidate List
(CCL) Research Plan (U.S. EPA, 2000e)
This plan describes the research approach and
process to provide improved scientific and
technical bases for the assessment and
management of drinking water contaminants that
are on the CCL.
U.S. Environmental Protection Agency (U.S. EPA). 1997. Final Research Plan for Microbial Pathogens and Disinfection By-
Products in Drinking Water. Office of Research and Development. Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA). 1998a. Research Plan for Arsenic in Drinking Water. Office of Research
and Development. Washington, DC. EPA/600/R-98/042.
U.S. Environmental Protection Agency (U.S. EPA). 1998b. Strategic Research Plan for Endocrine Disruptors. Office of
Research and Development. Research Triangle Park, NC.
U.S. Environmental Protection Agency (U.S. EPA). \999.AirborneParticulateMatterResearchStrategy. Office of Research
and Development. Research Triangle Park, NC.
U.S. Environmental Protection Agency (U.S. EPA). 2000a. Strategy for Research on Environmental Risks to Children. Office of
Research and Development. Washington, DC.
U.S. Environmental Protection Agency (U.S. EPA). 2000b. Mercury Research Strategy. Office of Research and Development.
Washington, DC. EPA/600/R-00/073.
U.S. Environmental Protection Agency (U.S. EPA). 2000c. Asthma Research Strategy. Draft. Office of Research and
Development. Research Triangle Park, NC.
U.S. Environmental Protection Agency (U.S. EPA). 2000d. Air Toxics Research Strategy. Draft. Office of Research and
Development. Research Triangle Park, NC.
U.S. Environmental Protection Agency (U.S. EPA). 2000e. Drinking Water Contaminants Candidate List (CCL) Research
Plan. Draft. Office of Research and Development. Research Triangle Park, NC.
A-2
-------
APPENDIX B
Organizational Chart
EPA's Office of Research and Development
Office of Science Policy
Assistant Administrator
of Research and Development
National Center
For Environmental
Research
National Exposure
Research
Laboratory
Offices of Resources and
Management Administration
National Health
& Environmental
Effects Research
Laboratory
National Center
For Environmental
Assessment
A-3
-------
APPENDIX C
Examples of Mechanistic Data Used in Risk Assessment
Pollutant
Supporting Research
Aflatoxin Mechanistic studies showed that this compound forms DNA adducts and
Bl protein adducts, causing specific mutations in the p53 tumor suppressor gene.
Because of this mechanistic information, formation of DNA adducts is now
being used to assess cancer risk in human populations.
Dioxin The understanding that essentially all the effects of dioxin are mediated via
binding to the arylhydrocarbon (Ah) receptor provides the underpinning for the
species extrapolation in the risk assessment of dioxin. The Ah receptor is
highly conserved, present, and functional in nearly all vertebrates. The current
consensus that dioxin is a known human carcinogen is based on clear animal
data, limited human data, and the presence of a common mechanism of action.
Dioxin The importance of PBPK models for risk assessment is illustrated by the
identification of an inducible hepatic binding protein by dioxin, which results
in dose-dependent sequestration of dioxin in multiple mammalian species,
including humans. This information has allowed for a better understanding of
the dose-dependent differences in the disposition of dioxin, which has led to
the conclusion that body burden is the best dose metric for risk assessment of
dioxin and related compounds. This approach allows for a direct comparison
of animal and human data, which reduces the animal-to-human uncertainty in
risk assessment.
d- A number of chemicals (e.g., d-limonene) and chemical mixtures (e.g.,
Limonene unleaded gasoline) induce kidney tumors in male rats in cancer bioassays.
Mechanistic studies have shown that kidney tumors in male rats are associated
with an increase in the level of a specific protein, cc2|j,-globulin. Because this
protein is not present in human male kidneys, risk assessors could predict that
the cancer risk in humans for chemicals acting via an cc2|j,-globulin-mediated
process will be low.
Atrazine Research from ORD showed that the effects of atrazine on mammary gland
and prostate development are associated with alterations in the hormone
prolactin. This mechanistic information is currently being used to reevaluate
the risk assessment for atrazine.
A-4
-------
APPENDIX D
Agencies Having Research Programs Complementary to ORD
The National Institute of Environmental
Health Sciences (NIEHS) achieves its
mission through multidisciplinary bio-
medical research programs; prevention and
intervention efforts; and communication
strategies that encompass training, educa-
tion, technology transfer, and community
outreach. For example, the NIEHS program
includes a trans-NIH effort to study effects
of chemicals, including pesticides and other
toxics, in children. EPA has collaborated
with NIEHS in establishing Centers for
Children's Environmental Health and
Disease Prevention to define the environ-
mental influences on asthma and other
respiratory diseases, childhood learning, and
growth and development. NIEHS and the
National Institute of Allergy and Infectious
Diseases (NIAID) are conducting the Inner-
City Asthma Study, which is a prevention
trial to develop an intervention strategy to
reduce asthma morbidity in inner-city
children and adolescents. The National
Allergen Study, being conducted by NIEHS
in collaboration with the Department of
Housing and Urban Development (HUD),
examines the relationship between allergens
and lead and how allergen exposures differ
as a function of geographic region, socio-
economic status, housing type, and ethni-
city. NIEHS and the National Toxicology
Program (NTP) are developing new technol-
ogies for high-throughput toxicity testing,
and these agencies are responsible for one-
third of all toxicity testing performed
worldwide. Long-term collaborative efforts
with NTP, particularly in the areas of
carcinogenesis, reproductive/developmental
toxicity, and neurotoxicity, are well
established. NIEHS has established the
National Center for Toxicogenomics (NCT)
to coordinate an international research effort
to develop the field of toxicogenomics. The
NCT will provide a unified strategy, a
public database, and will develop the
bioinformatics infrastructure to promote the
development of the field of toxicogenomics.
NIEHS will pay special attention to
toxicogenomics as applied to the prevention
of environmentally related diseases.
The National Cancer Institute (NCI)
conducts population-based research on
environmental and genetic causes of cancer
and on the role of biological, chemical, and
physical agents in the initiation, promotion,
or inhibition of cancer and on the biological
and health effects of exposure to radiation.
The Centers for Disease Control and
Prevention (CDC), through the National
Center for Environmental Health (NCEH),
studies health problems associated with
human exposure to lead, radiation, air
pollution, and other toxicants, as well as to
hazards resulting from technological or
natural disasters. These are mainly
surveillance and epidemiological studies.
NCEH is particularly interested in studies
that benefit children, the elderly, and
persons with disabilities. The National
Center for Health Statistics (NCHS) of CDC
is conducting the National Health and
Nutrition Examination Survey (NHANES).
NHANES is a national population-based
survey and includes data on potentially
sensitive subpopulations such as children
and the elderly. EPA is participating in this
survey with NCHS to collect information on
children's exposure to pesticides and other
environmental contaminants. CDC's
National Report on Human Exposure to
Environmental Chemicals provides an
ongoing assessment of the exposure of the
U.S. population to environmental chemicals
using biomonitoring data collected through
NHANES. The first national report
provides information about levels of 27
A-5
-------
chemicals, while the second report presents
biomonitoring exposure data for 116
environmental chemicals for the U.S.
population divided into age, gender, and
race/ethnicity groups.
The National Institute of Child Health
and Human Development (NICHD)
supports laboratory, clinical, and epidemio-
logical research on the reproductive,
neurobiological, developmental, and
behavioral processes that determine and
maintain the health of children and adults.
ORD is collaborating with NICHD, CDC,
and other Federal agencies in the design and
implementation of a National Children's
Study of 100,000 children who will be
enrolled during the mother's pregnancy and
followed throughout childhood and
adolescence. The Children's Health Act of
2000 mandated this study of environmental
influences on children's health and
development.
The National Center for Toxicological
Research (NCTR) supports fundamental
research on the effects of chemicals
regulated by the Food and Drug
Administration. Although some of the
models used by NCTR may be similar to
those used by EPA, the chemicals and
regulatory context vary significantly.
Historically, NCTR has been a leader in
developing models and principles for risk
assessment, which has led to collaborations
between EPA and NCTR scientists.
A-6
-------
APPENDIX E
Examples of Health and Environmental Databases to Evaluate
Public Health Outcomes From Risk Management Actions
Environmental Databases
Source
Database Name
Contents
EPA/ORD
EPA
EPA
EPA/OAQPS
EPA
EPA
EPA/OPPTS
EPA
EPA
NHEXAS
SDWIS/FED
STORE!
AIRS
ETS
Center for
Environmental
Information and
Statistics
TRI
CERCIS
BASINS
Exposure data for Arizona, EPA Region V, and
Baltimore
Regulated pollutant concentration in drinking
water
Surface water quality/biological monitoring
Air pollutant concentrations at 4,000 sites;
9,000 point sources
Emissions from electric utilities
Central source of environmental data/trends
Toxic compounds release inventory
Hazardous waste sites, assessment, and status
Watershed pollutants (point and area source)
and locations
A-7
-------
Health Effects Databases
VA databases
EPA/ORD IRIS
NCI SEER
CDC
Veterans
Administration
National Center for
Health Statistics
State Health Various
Departments
Insurance Various
Companies
Hazard characterization and risk numbers for
cancer and noncancer endpoints
Cancer incidence/prevalence by type and
location
Behavioral Risk Factor Incidence of contagious diseases
Survey
Major disease incidence and prevalence by
location
National Health Prevalence and incidence data in populations
Interview Survey
Disease incidence by location and time
Disease and death incidence by location, time,
and population
Health and Environmental Databases
EPA, Region 3
ATSDR
Green Communities
Initiative
Registries
Hazardous Substances
Emergency Event
Surveillance System
Environmental health, economic, and societal
indicators of impact of environmental
regulation
Follow-up individual health outcomes on
communities and populations exposed to
specific environmental contaminants
Collects information on releases and victims
A-8
-------
United States
Environmental Protection
Agency
Office of Research
and Development (8104R)
Washington, DC 20460
Official Business
Penalty for Private Us
$300
EPA6GQ/R-02/050,
September 2003
www. epa.gov
Please make all necessary changes on ihe below label,
datoch of copy, and return lo Vte address tn :ho upper
lefi-hand coiner.
II you do not wish to iccewe Ihose ien>ons CHECK HEBE [~|;
detach, or copy 1hts covet, and f etiwn to the address m the
uppef lefi-hand cofnn.
PRESORTED STANDARD
POSTAGE S FEES PAID
EPA
PERMIT No. G-35
HUMAN H
ESEARCH STRATEGY
Pfmted with vegetable-based ir*. on
eap*< than conla»is a irnnnwn or
50% posl-ecfi4ui\*f Tiber f\\efil
------- |