EPA/600/R-16/025
May 2016
ESTIMATING GREENSPACE EXPOSURE AND BENEFITS
FOR CUMULATIVE RISK ASSESSMENT APPLICATIONS
U.S. Environmental Protection Agency
Office of Research and Development
National Center for Environmental Assessment
Greenspace and Cumulative Risk Assessment Technical Meeting
4-5 May 2015
Cincinnati, OH
-------
oEPA EPA/600/R-16/025
May 2016
Final
www.epa.gov/research
Estimating Greenspace Exposure and Benefits for
Cumulative Risk Assessment Applications
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
li
-------
DISCLAIMER
This report highlights the presentations, discussions, and practical suggestions offered by
the meeting participants only; this report does not necessarily reflect the views or policies of the
U.S. Environmental Protection Agency or any other entity. This report does not present
consensus opinions of the meeting participants.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
111
-------
ABSTRACT
This document provides a summary of the technical meeting on greenspace and
cumulative risk assessment (GS-CRA) convened May 4-5, 2015 in Cincinnati, OH, by the
U.S. Environmental Protection Agency (EPA) Office of Research and Development (ORD)
National Center for Environmental Assessment (NCEA). This report highlights the
presentations, discussions, and practical suggestions offered by the meeting participants;
however, the report does not present consensus opinions of the meeting participants.
Meeting Objective: Identify and evaluate approaches and appropriate data sources for
measuring greenspace and exposure, and examine the distribution of health impacts of
greenspace (e.g., across socio-economic status, sensitive populations), including risk reductions,
from a cumulative risk assessment perspective, with attention to uncertainty in reporting and
measurement.
Approach: The meeting was structured to focus on (1) approaches and tools for
estimating greenspace (GS) exposure, and (2) potential risks and benefits of GS exposure for
human health and insights for cumulative risk assessment (CRA) applications. Meeting
participants shared duties in presenting relevant research on agenda sub-topics and leading group
discussions.
Findings: Both GS assessments and CRAs are relatively new approaches for
characterizing both the health benefits and risks associated with complex environmental
exposures. While existing evidence supports that GS effects are primarily beneficial for human
health, GS assessments strongly depend on the factors specific to the places and populations of
interest, which can differently influence the duration, frequency, and type of human exposure to
various types and quantities of GS. Quantification and qualification of dose-response
relationships related to GS exposure is limited for GS assessments, largely due to uncertainty
around GS exposure measures and the mechanisms of action between GS engagement and
human health outcomes.
The report was prepared by Argonne National Laboratory, supported by the EPA under
an interagency agreement through U.S. Department of Energy contract DEAC02-06CH11357, in
collaboration with the EPA ORD NCEA Organizing Committee and the GS-CRA Technical
Work Group. The meeting participants reviewed and refined this report before final review and
clearance by the EPA.
11
-------
TABLE OF CONTENTS
LIST OF TABLES v
LIST OF FIGURES v
LIST 01 ABBREVIATIONS AM) ACRONYMS vi
AUTHORS, CONTRIBUTORS, AND REVIEWERS vii
ACKNOWLEDGEMENTS ix
EXECUTIVE SUMMARY x
1. INTRODUCTION 1-1
1.1. MEETING PURPOSE AND SCOPE 1-1
1.2. OBJECTIVES AM) APPROACH 1-1
1.3. PARTICIPANTS 1-2
1.4. REPORT ORGANIZATION 1-2
2. FRAMING CONTEXT FOR CUMULATIVE RISK AND GREENSPACE (GS)
ASSESSMENTS 2-1
2.1. CUMULATIVE RISK ASSESSMENTS 2-1
2.2. GREENSPACE (GS) ASSESSMENTS 2-2
2.3. CONCEPTUAL MODELS 2-3
3. GREENSPACE (GS) EXPOSURE ASSESSMENTS AND METRICS 3-1
3.1. TECHNICAL PRESENTATIONS 3-1
3.1.1. Exposure Assessment Approaches 3-1
3.1.2. Tree Cover Measurements 3-5
3.1.3. Access to Greenness 3-6
3.1.4. Built Environment 3-8
3.1.5. Design and Environment Psychology 3-9
3.1.6. Specific Populations: Exposure 3-11
3.1.7. Exposure Metrics, Links to Health 3-12
3.2. KEY EXPOSURE CONSIDERATIONS 3-13
3.3. ASSESSMENT APPROACHES AND METRICS 3-14
4. GREENSPACE (GS) AM) HEALTH 4-1
4.1. TECHNICAL PRESENTATIONS 4-1
4.1.1. Respiratory Effects 4-1
4.1.2. Reproductive Effects 4-4
4.1.3. Obesity and Physical Activity 4-5
4.1.4. Cardiovascular Disease and Mortality 4-5
4.1.5. Neurologic/Neurodevelopmental Effects 4-6
4.1.6. Psychosocial Effects 4-7
4.1.7. Attention Restoration/Cognition 4-8
4.1.8. Economic and Community Benefits 4-8
4.1.9. Specific Populations: Health 4-9
4.2. KEY EFFECT CONSIDERATIONS 4-9
4.2.1. Specific Qualities of Respiratory Studies 4-10
iii
-------
TABLE OF CONTENTS (continued)
4.2.2. Specific Qualities of Neurological, Psychosocial, and Attention
Studies 4-10
4.2.3. Qualities of Studies of Other Health Outcomes 4-11
4.3. ASSESSMENT APPROACHES AND MEASURES 4-11
5. DISCUSSION SYNTHESIS AND CONTEXT FOR CUMULATIVE RISK
ASSESSMENT (CRA) 5-1
5.1. ISSUES COMMON TO GREENSPACE (GS) AND CUMULATIVE
RISK ASSESSMENTS 5-1
5.2. EVOLVING FIELD OF GREENSPACE (GS) ANALYSES 5-10
5.3. APPROACHES TO ADDRESS RANGE OF GREENSPACE (GS)
TYPES AND SPATIAL AND TEMPORAL SCALES 5-11
5.4. AVAILABLE EXPOSURE METRICS: APPLICATIONS AND
LIMITATIONS 5-11
5.5. IMPORTANCE OF ENGAGEMENT 5-12
5.6. MAINLY BENEFICIAL EFFECTS 5-13
5.7. LIMITED QUANTIFICATION OF EXPOSURE-RESPONSE
RELATIONSHIPS 5-14
5.8. UNCERTAINTIES ASSOCIATED WITH EXPO SURE AND HEALTH
MEASURES USED IN DIFFERENT GREENSPACE (GS) STUDIES 5-14
5.9. UNCERTAINTIES ASSOCIATED WITH INNATE DIFFERENCES IN
GREENSPACE ASSESSED IN DIFFERENT STUDIES 5-16
5.10. FUTURE RESEARCH DIRECTIONS 5-17
5.11. STRENGTH OF EVIDENCE FOR CAUSALITY, IDENTIFYING MAIN
ENDPOINTS 5-21
5.12. SUMMARY FINDINGS 5-24
6. CITED REFERENCES AND ADDITIONAL RELEVANT PUBLICATIONS 6-1
6.1. CITED REFERENCES 6-1
6.2. ADDITIONAL RELEVANT PUBLICATIONS 6-9
APPENDIX A: PARTICIPANT BIOSKETCHES A-l
APPENDIX B: MEETING AGENDA AND TECHNICAL PRESENTATIONS B-1
APPENDIX C: WORKING DRAFT GLOSSARY C-l
iv
-------
LIST OF TABLES
1 -1. Meeting participants 1-4
3-1. Presentation highlights: greenspace (GS) metrics and exposure 3-2
4-1. Presentation highlights: health effects of greenspace (GS) 4-2
5-1. Greenspace (GS) exposure measures and metrics and health applications 5-2
LIST OF FIGURES
2-1. Overarching conceptual model illustrating features of greenspace (GS) exposures
and health effects 2-6
2-2. Greenspace (GS) characteristics and functions that influence exposures and
interactions 2-7
5-1. Strength of evidence for selected health effects 5-23
v
-------
LIST OF ABBREVIATIONS AND ACRONYMS
AR
attention restoration
ART
Attention Restoration Theory
ASPPH
Association of Schools and Programs of Public Health
AT SDR
Agency for Toxic Substances and Disease Registry
BE
built environment
CAU
census area unit
CCAAPS
Cincinnati Childhood Allergy and Air Pollution Study
CORINE
Coordination of Information on the Environment
CDC
Centers for Disease Control and Prevention
CRA
cumulative risk assessment
CREAL
Centre for Research in Environmental Epidemiology (Centre de Recerca en
Epidemiologia Ambiental)
CRESH
Center for Research on Environment, Society and Health
CVD
cardiovascular disease
DALY
disability-adjusted life year(s) (lost)
EPA
U.S. Environmental Protection Agency
EVI
enhanced vegetative index
FEV
forced expiratory volume
GIS
geographic information system
GLUD
Generalised Land Use Database
GPS
global positioning system
GS
greenspace
GS-CRA
technical meeting on greenspace and cumulative risk assessment
ICD
International Classification of Diseases
LAI
leaf area index
LiDAR
light detection and ranging
MODIS
moderate resolution imaging spectroradiometer
NCEA
National Center for Environmental Assessment (EPA ORD)
NDVI
normalized difference vegetation index
NHEERL
National Health and Environmental Effects Research Laboratory
OKI
Ohio-Kentucky-Indiana Regional Council of Governments
ORD
Office of Research and Development (EPA)
PA
physical activity
PM
particulate matter
PM2.5
fine particulate matter (with an aerodynamic diameter of a nominal <2.5 microns)
PUFA
polyunsaturated fatty acid(s)
SES
socioeconomic status
TRAP
traffic-related air pollution
UGSI
urban greenspace index
UI-UC
University of Illinois, Urbana-Champaign
USFS
U.S. Forest Service
UTC
urban tree cover
VOC
volatile organic compound(s)
WHO
World Health Organization
vi
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
U.S. EPA GREEN SPACE-CUMULATIVE RISK ASSESSMENT (CRA) MEETING
ORGANIZING COMMITTEE
Rebecca Gernes
Glenn Rice
J. Michael Wright
Glennon Beresin
Bette Zwayer
Environmental Health Research Participant for the
Association of Schools and Programs of Public Health
(ASPPH)
U.S. EPA ORD NCEA, Cincinnati, OH
U.S. EPA ORD NCEA, Cincinnati, OH
Environmental Health Research Participant for the ASPPH
U.S. EPA ORD NCEA, Cincinnati, OH
MEETING FACILITATOR
Travis Miller
REPORT AUTHORS
Margaret MacDonell
Richard Hertzberg
Rebecca Gernes
Glenn Rice
J. Michael Wright
Glennon Beresin
Travis Miller
Julia Africa
Geoffrey Donovan
J. Aaron Hipp
Perry Hystad
Laura Jackson
Michelle Kondo
Yvonne Michael
Ohio-Kentucky-Indiana Regional Council of Governments
(OKI), Cincinnati, OH
Argonne National Laboratory, Argonne, IL
Argonne National Laboratory, Argonne, IL
Environmental Health Research Participant for the ASPPH
U.S. EPA ORD NCEA, Cincinnati, OH
U.S. EPA ORD NCEA, Cincinnati, OH
Environmental Health Research Participant for the ASPPH
OKI, Cincinnati, OH
Harvard T.H. Chan School of Public Health, Boston, MA
U.S. Forest Service (USFS), Portland, OR
North Carolina State University, Raleigh, NC
Oregon State University, Corvallis, OR
U.S. EPA, ORD National Health and Environmental
Effects Research Laboratory (NHEERL), Research
Triangle Park, NC
USFS, Philadelphia, PA
Drexel University, Philadelphia, PA
Vll
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (continued)
REPORT AUTHORS (continued)
Richard Mitchell
Mark Nieuwenhuij sen
Patrick Ryan
William Sullivan
Matilda Annerstedt van den Bosch
University of Glasgow, Glasgow, Scotland
Centre for Research in Environmental Epidemiology
(CREAL), Barcelona, Spain
University of Cincinnati, Cincinnati, OH
University of Illinois, Urbana-Champaign (UI-UC), IL
Swedish University of Agricultural Sciences, Alnarp,
Sweden
PLANNING CONTRIBUTORS
Margaret MacDonell
Kathy Eggers
Laura Bryant
Argonne National Laboratory, Argonne, IL
Argonne National Laboratory, Argonne, IL
National Council on Aging (NCA) Grantee at U.S. EPA
ORD NCEA, Cincinnati, OH
viii
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge the extensive contributions of meeting facilitator Travis
Miller, whose technical expertise and organizational skills were invaluable to the success of the
meeting. The authors also thank each invited expert for his or her excellent technical
presentations and contributions to the group discussions. In addition, we recognize Amanda
Evans for her insightful contributions to the discussions. We also thank the organizing
committee for its substantial efforts in developing this technical meeting and the further planning
and implementation contributions from Kathy Eggers of Argonne, Bette Zwayer of EPA NCEA-
Cincinnati and Laura Bryant of NCA. Finally, special thanks to Chelsey Mitchell for the lovely
artwork on our cover and to her father, Rich, for sharing it with us.
This report was prepared by Argonne National Laboratory as part of Collaborative
Interagency Agreement No. DW8992268301-0 between U.S. EPA ORD and the
U.S. Department of Energy.
Cover drawing courtesy of Chelsey Mitchell.
IX
-------
EXECUTIVE SUMMARY
ES.l. MEETING PURPOSE AND SCOPE
The Cincinnati Office of the EPA National Center for Environmental Assessment
(Cincinnati) convened a technical meeting to explore how methods and measures used to assess
greenspace (GS) could contribute to cumulative risk assessments (CRAs) and vice versa. GS
was broadly defined as open land that is at least partly vegetated, located in or adjacent to urban
or suburban areas. A key consideration is the use and effectiveness of GS as an ecosystem
service and potential risk management practice to benefit human health. The purpose of the
meeting was to inform methods and measures for assessing environmental health benefits and
risks of GS and to consider how to incorporate these factors into cumulative risk analyses.
ES.2. OBJECTIVES AND APPROACH
The main objectives of the meeting were to (1) identify how GS is being described and its
impacts assessed and (2) gain insights from GS assessments for CRA applications. To realize
these objectives, the EPA NCEA organizers brought together a group of GS experts and
practitioners from multiple disciplines to participate in joint presentations and facilitated
discussions. The presentations and discussions were topically organized by exposure and health,
and framed by conceptual models and driving questions developed by the organizers.
ES.3. KEY FINDINGS AND SUGGESTIONS
Evaluating GS effects on public health is a new and growing field of research. Published
studies show that assessments of impacts can be influenced by the measures used to describe
various attributes of the GS. Few confirming examples exist to determine whether it is valid to
apply the measures used in one study to assess GS effects for another, or to extend them more
broadly. Measures of GS exposures and related effects continue to evolve, but fully quantitative
GS health assessments are not yet available. Instead, a mixture of qualitative and quantitative
methods is used to define dose-response relationships. A number of study results show promise
for better understanding GS impacts. Researchers categorized the strength of causal
relationships (as high, medium, low) linking GS to health effects, and many paths toward
improved research have been identified. The technical work group found several areas of
consensus regarding GS impacts. In most analytical frameworks, the benefits of GS outweigh
the risks. There are benefits of GS by itself (e.g., a direct impact that reduces physiological and
psychological stress) and perhaps more often there are indirect benefits whereby GS reduces the
magnitude or effect of other exposures to stressors, lessening adverse outcomes. Common
x
-------
examples of indirect impacts that may be related to stress reduction include lower risk of
respiratory and cardiovascular disease and improved birth outcomes. Negative impacts of GS
include well-characterized exposures to environmental irritants such as pollen or mold for which
causal mechanisms are fairly well understood. For other impacts, both negative and positive,
more research is needed to understand the mechanisms and magnitudes of GS exposures and
related health outcomes. Finally, the addition of expertise in environmental psychology and
microbiomics could enrich suggested approaches and metrics for assessing GS exposures and
effects in the context of CRAs.
Five joint findings for GS assessments and CRAs can be distilled from the discussions:
1. GS effects are mainly beneficial.
Current evidence suggests that GS supports public health directly by providing a
dynamic space for exercise, social interactions, and other behaviors that are thought
to lower psychological stress and improve mood. Additional benefits of exposure to
GS appear to include improved cognition, attention restoration, and improved
immune function. Although data are limited, GS might mitigate or attenuate health
outcomes brought on by psychological stress (e.g., cardiovascular disease). A few
adverse effects from GS exposure also occur—notably respiratory and dermal
irritation related to allergens.
2. Both GS assessments and CRAs are spatially dependent.
Both assessments can be conducted at different levels of spatial extent, with
resolutions ranging from rough to highly refined. However, unlike conventional
CRAs, the meaningful attributes of a GS—beyond those associated with objectively
spatial measurements—are not well characterized.
3. Both GS assessments and CRAs strongly depend on location and population
characteristics.
Part of the planning phase of any risk assessment is to identify the scope of the
effect(s) and characterize affected population(s); CRA and GS analyses incorporate
these two factors in different ways. The scope of a CRA is often defined to increase
the tractability of the multiple stressors being addressed. Simplification can involve
placing limits on the number of chemicals, exposure pathways, or health effects to
include. With GS evaluations, the scope of the analysis generally relates to the
physical boundaries (e.g., the definition of the type and boundaries of the GS, or the
amount of GS within a defined buffer), although the set of potential health endpoints
in the nearby population is often considered in the assessment scope. The relative
absence of GS characteristics (e.g., ecological features like biodiversity, landscape
structure, and behavioral prompts like paths and overlooks) from GS assessment is a
significant shortcoming. GS assessments also exhibit a strong dependence on the
population under consideration that mirrors the way in which activity profiles of a
population (or individuals) is used when assessing exposures to chemicals in a CRA.
XI
-------
4. Quantification and qualification of dose-response relationships related to GS
exposure is limited for GS assessments. The same is true for complex chemical
mixtures typically assessed in CRAs.
A mathematical dose-response relationship linking GS exposure with any specific
health outcome(s) does not yet exist. Uncertainties in the characterization of
exposure and causality for GS are similar to the methodological limitations of
environmental chemical exposure assessment, lacking even the cursory causality
information that is available for a subset of chemicals studied in controlled animal
experiments.
5. Both GS assessments and CRAs are relatively new approaches for characterizing
complex environmental exposures. Considerable uncertainty underlies GS exposure
measures used to assess various health outcomes.
Uncertainties remain in the best available methods for quantifying and qualifying GS
exposure as well as characterizing the etiology of various health endpoints,
potentially limiting the usefulness of CRA analysis that incorporate GS. A lack of
understanding regarding the mechanism or mechanisms through which GS might
affect these health outcomes underlies many of the uncertainties in the exposure
measures. Further research and exposure classification is needed, but full
incorporation of all dimensions of GS exposure into a CRA model is unlikely.
Xll
-------
1. INTRODUCTION
1.1. MEETING PURPOSE AND SCOPE
The Cincinnati Office of the U.S. Environmental Protection Agency (EPA) National
Center for Environmental Assessment (NCEA)—Cincinnati hosted a technical meeting to
evaluate the effect of greenspace (GS) on human health from a cumulative risk assessment
(CRA) perspective. The group broadly defined GS as open land that is at least partly vegetated
and located in or adjacent to urban or suburban areas. Access and exposure to GS have been
reported to influence human health. The meeting explored how these influences can be
explicitly considered in GS assessments and CRAs.
The multiple pathways or roles through which GS potentially affects human health and
the different measures of GS used by researchers across different studies complicates existing
analyses. The mechanisms or causal pathways that best explain associations between GS and
health outcomes are uncertain. While some of the uncertainty is related to the newness of the
field and the relatively limited number of studies, some of the uncertainty also due to the
diversity of GS measures used (e.g., areal extent or plant density) and the range of causal
pathways explored (e.g., reduced air pollution via filtering, improved psychological well-being
from exposure to nature). Because GS can potentially act as either a nonchemical stressor or an
exposure modifier, it appears to be a good candidate for examination in a cumulative risk context
to help evaluate its application and effectiveness as an ecosystem service and potential risk
management practice.
The experts assembled to evaluate GS for insights into CRA (and vice versa) were asked
to review existing GS exposure measures, methods, and health effects being considered across
different fields of study, focusing on which measures are useful for assessing different health
outcomes and which are candidates for extending GS insights to CRA applications. The meeting
discussions highlighted in this report are intended to inform methods for evaluating
environmental health risks and benefits associated with GS.
1.2. OBJECTIVES AND APPROACH
The two main objectives of the meeting were to identify (1) how to characterize GS and
assess its impacts and (2) to present insights from GS assessments for CRA applications. The
approach for realizing these objectives involved identifying experts from multiple disciplines, as
outlined in the meeting agenda (see Appendix B). Sets of participants then jointly developed and
delivered presentations. A facilitated group discussion of GS measures and roles followed the
presentations. Further topics of interest were captured for discussion as time allowed. The EPA
1-1
-------
organizers also developed a set of driving questions to guide the presentation materials and group
discussions of GS toward identifying insights for CRA:
• How can existing cumulative risk assessment frameworks consider GS as it relates to
exposure assessment for human health?
• How is GS conceptualized across disciplines?
• What health outcomes are relevant to GS prevalence and access?
• Which evidence-based measures of GS provide the most applicable, reliable, and
replicable estimates for GS exposure in urban settings?
• What are the specific mechanisms for certain health benefits, and can this information be
used to inform biologic plausibility of reported associations with GS?
Twice during the month prior to the meeting, the organizers and invited participants
convened by teleconference to outline the working agenda. From the outset, participants
identified the importance of defining how key terms would be used. Reflecting collective inputs
at the meeting, GS is defined in this report as open land that is at least partly vegetated and
located in or adjacent to urban or suburban areas. Note that water also can be an important part
of what are called natural areas. Commonly referred to as blue space, water areas are not usually
included when measuring the size and shape of GS, and they are not included in GS as it has
been defined for this report.
1.3. PARTICIPANTS
The meeting participants, their affiliations, and key areas of expertise are identified in
Table 1-1.
1.4. REPORT ORGANIZATION
This meeting report on the technical meeting on GS and CRA is organized as follows:
• Chapter 2 provides overview information about GS and CRA.
• Chapter 3 summarizes the exposure presentations and synthesizes key discussion points.
• Chapter 4 summarizes the health effect presentations and synthesizes key discussion
points.
1-2
-------
Chapter 5 highlights core elements of the combined GS exposure and effect discussions
and provides context for considering GS in CRA.
Chapter 6 offers insights for cumulative risk applications.
Chapter 7 lists the references cited in this report and additional relevant publications.
Appendix A presents the biosketches of meeting participants.
Appendix B provides the meeting agenda and the technical presentations.
Appendix C presents the draft glossary distributed at the technical meeting.
1-3
-------
Table 1-1. Meeting participants
Name
Organization
Key expertise
Invited GS experts
Julia Africa
Harvard T.H. Chan School of
Public Health
Ecological infrastructure, biophilic design,
and restorative landscapes
Geoffrey Donovan
U.S. Forest Service (USFS),
Pacific Northwest
Environmental economics, urban tree
benefits, safety, and public health
J. Aaron Hipp
North Carolina State University
Built environment (BE) and health
behaviors, physical activity
Perry Hystad
Oregon State University
Environmental epidemiology, greenness,
and chronic health effects
Laura Jackson
U.S. EPA National Health and
Environmental Effects Research
Laboratory (NHEERL)
Ecosystem services, urban ecosystems
Michelle Kondo
USFS, Philadelphia
Environment, public health, and safety;
urban stabilization/sustainability
Yvonne Michael
Drexel University School of Public
Health
Epidemiology, psychosocial factors in
health, healthy aging, women's health
Richard Mitchell
University of Glasgow
Influence of physical and social
environments on population health
Mark Nieuwenhuijsen
Centre for Research in
Environmental Epidemiology
(CREAL), Barcelona
Environmental exposure and health impact
assessment, epidemiology
Patrick Ryan
University of Cincinnati
(Cincinnati Children's Hospital and
Medical Center)
Environmental epidemiology, air pollution
William Sullivan
University of Illinois at
Urbana-Champaign
Greenspace and healthy, sustainable
communities; attention restoration
Matilda Annerstedt
van den Bosch
Swedish Agricultural University
Behavioral medicine, epidemiology, natural
environments, and public health
EPA Cincinnati GS-CRA team
Glennon Beresin
Association of Schools and
Programs of Public Health
Greenspace and public health
Amanda Evans
Oak Ridge Institute for Science and
Education
Cumulative exposure and risk, mixtures
Rebecca Gernes
Association of Schools and
Programs of Public Health
Greenspace and public health
Glenn Rice
EPA ORD NCEA
Cumulative exposure and risk, mixtures
1-4
-------
Table 1-1. Meeting participants (continued)
Name
Organization
Key expertise
EPA Cincinnati GS-CRA team (continued)
J. Michael Wright
EPA ORD NCEA
Environmental epidemiology,
cumulative risk assessment
Argonne CRA collaborators
Richard Hertzberg
Argonne National Laboratory
Cumulative exposure and risk, mixtures
Margaret MacDonell
Argonne National Laboratory
Cumulative exposure and risk, mixtures
Technical expert-facilitator
Travis Miller
Ohio-Kentucky-Indiana (OKI)
Regional Council of
Governments
Land use planning
1-5
-------
(This page intentionally left blank.)
1-6
-------
2. FRAMING CONTEXT FOR CUMULATIVE RISK AND GREENSPACE (GS)
ASSESSMENTS
To examine how GS analyses could inform CRA practices and potentially be
incorporated into future CRAs, the technical experts prepared presentations describing how GS
exposures are assessed and how GS can affect human health. Basic concepts underlying a CRA
and GS assessment are presented in Sections 2.1 and 2.2, respectively. Section 2.3 describes the
development of conceptual models as a way to organize key information and communicate
elements of an assessment to interested parties; an overarching conceptual model illustrating
features of greenspace (GS) exposures and health effects is also presented.
2.1. CUMULATIVE RISK ASSESSMENTS
CRA is a relatively recent and evolving field of risk analysis. CRAs are designed to
characterize and quantify, to the extent possible, the combined risks to human health or the
environment from exposures to multiple stressors, including chemical, physical, biological, and
psychosocial stressors (U.S. EPA, 2003; U.S. EPA, 2007; NAS, 2009). CRAs that focus on
human health risks are typically evaluated from a population perspective. Such CRAs consider
the given population's vulnerabilities to potentially harmful stressors (e.g., a genetic
predisposition to harm from exposures to a certain mix of stressors). The U.S. EPA (2003)
Framework identifies key elements of CRAs and observes that approaches for conducting a CRA
can range from qualitative to quantitative, depending on available data and resources. Some
CRAs have focused on communities that are more burdened than others, as part of
environmental justice evaluations. Other CRAs focus on diseases, which are multifactorial, so
applying a CRA approach helps assure that multiple factors are considered and reduces the
potential for missing a key factor.
CRAs in particular, and risk assessments in general, are conducted to help risk managers
make decisions. Ideally, CRAs should be conducted in a decision-relevant context and used to
convey to a risk manager what is known about the risks and benefits associated with the
exposure conditions for the population group of interest by considering the multiple stressors and
buffers to which the population might be exposed and associated health effects. A buffer
(sometimes referred to as a mediator) mitigates an adverse exposure or effect; in epidemiologic
studies, buffers may modify exposures, modify effect measures, or be identified as confounders.
As an example, some people consider polyunsaturated fatty acids (PUFAs) in fish to be a buffer
because some studies show decreased risk of cardiovascular disease (CVD) among people who
eat fish, and these decreases have been attributed to the PUFAs (Cohen et al., 2005a, 2005b).
2-1
-------
We note that decision makers generally have concerns beyond health risks; these can include
cost, feasibility, and social acceptance among other factors.
A CRA for environmental health differs from a classical chemical- or source-based risk
assessment in several important ways. First, a CRA typically focuses on a specific population
instead of on a pollutant source (e.g., emissions from a facility stack or effluent discharge pipe).
Consequently, the focus is on a population's health risks, reflecting all relevant sources
contributing to their exposures and other factors that influence exposure-disease relationships.
Second, in CRAs, those "exposures" are extended to include influential environmental
and population-specific conditions, and they can be quite complex—involving not only multiple
chemicals but also nonchemical stressors and other factors that directly impact public health, and
they could render some populations more vulnerable to environmental exposures.
Third, CRAs potentially can evaluate multiple health effects. This feature reflects the
complex nature of the exposures as well as joint toxicity, with the potential for toxicological
interactions.
Fourth, the complexity often requires using simplifying methods in CRAs. For example,
a risk assessor could group chemicals by similarities in health (toxicity) endpoints, in timing of
different exposures, or in their occurrence via specific pathways or environmental media.
Fifth, because of the higher dimensionality and potentially large number of interactions,
information needed to quantify risk across all key elements is typically incomplete. Thus,
conducting an uncertainty analysis is an essential aspect of a CRA. Much of the information
characterizing exposures and effects in a CRA may be qualitative; consequently, the uncertainty
analysis could lack statistical descriptors such as confidence intervals. Instead, uncertainty
analyses for CRAs tend to contain descriptions and rankings of factors judged to be most
influential.
2.2. GREENSPACE (GS) ASSESSMENTS
The evaluation of GS in terms of environmental health risks and benefits has many of the
same features as a CRA. First, the population is key to understanding potential GS impacts.
Second, one of the primary complexities associated with GS assessments are the multiple ways
to describe and measure GS as well as the various ways humans interact with GS and are
exposed to features within GSs (e.g., released pollen). Third, GS has been reported to affect
multiple aspects of human health through various suggested mechanisms. Fourth, GS exposure
measures commonly used in health assessments are considered simplifications or proxy
measures (e.g., there are few studies of health effects that examine how people actually engage
with GS).
2-2
-------
Most GS effects appear to be strongly determined by (1) the manner and extent of
interaction with the GS by nearby individuals or the community and (2) the innate aspects of the
GS. Similarly, cumulative chemical risks can be influenced by the exposure characteristics of
route, concentration, and extent of exposure (represented by exposure time, frequency, and
duration), taking into consideration overlaps of timing of exposures and effects across multiple
stressors. Finally, GS analyses have substantial uncertainty given the many ways in which GS
could affect health.
There are some important differences between GS analyses and CRAs. Whereas a CRA
is an evaluation process that may or may not focus on a specific physical location, a GS
assessment focuses on a physical entity with geographic descriptors. In trying to understand the
causal mechanisms through which GS influences public health, it might be easier to
conceptualize a GS as representing a collection of nonchemical stressors and buffers, as well as
exposure or effect modifiers. Then the GS evaluation would be similar to a CRA that addresses
multiple stressors and impacts on a specific population (or individual).
Multiple-stressor exposures for a GS assessment need to be more fully defined than they
have been for CRAs. For example, instead of defining exposure as contact with chemical,
physical, or biological stressors, the GS exposure measures might include characteristics that
suggest or indicate specific types of population interactions with the GS.
The general measures most commonly used are quantity (how large is the GS area),
quality (detailed characteristics including on-site attractions such as playgrounds or flower
gardens), and function (likely effects on environmental quality parameters or population use).
Metrics for function can include measures of environmental effects (e.g., pollen concentrations
affecting air quality) as well as the specific functions and opportunities the GS provides for the
population. Actual population use, such as frequency of physical activity (PA) within the GS
boundaries, would also be measured when possible.
2.3. CONCEPTUAL MODELS
Conceptual models are representations, usually graphical, of the assumed relationships
between sources and effects (Suter, 1999). For chemical risk assessments, many conceptual
models are easy to interpret because they show actual material flows from emission sources to an
exposed population. For more complex cases, including CRAs (perhaps including CRAs that
would evaluate GSs), the connections shown in the model could depict direct and indirect effects
of stressors and buffers on multiple endpoints. Models can also reflect complex processes and
activities that include physical, psychosocial, and biological effects.
Conceptual models could serve three important purposes in CRAs: (1) they help analysts
thoughtfully examine and clarify their assumptions concerning the potential relationships among
2-3
-------
the stressors, buffers, and health outcomes assessed; (2) they can facilitate communication
among risk analysts, risk managers, and stakeholders; and (3) they can help identify important
information gaps and research needs. Conceptual models can also help organize the data
collection, analysis, and reporting of a CRA (Suter, 1999). One goal of this technical meeting
was to develop a clear, overarching conceptual model as a way to broadly illustrate the complex
relationships between GS and human health outcomes, potentially increasing the application of
conceptual models in GS analyses.
Conceptual models for GS could range from simple schematics to highly complex flow
charts (Hartig et al., 2014; Tzoulas et al., 2007; Lachowycz and Jones, 2013). Because GS can
alter stressors (e.g., shade can reduce overall heat), stressors can influence GS (e.g., heat alters
GS health), and GS can be a source of stressors (e.g., tree and grass pollen). It is not surprising
that conceptual models depicting GS effects can potentially include several double-ended arrows
and multiple connections. Because some of these relationships are not well understood, the
conceptual models may be incomplete or the depicted relationships could be speculative.
One recommended approach that has been successfully applied in ecological risk analysis
is to create modular component models (e.g., for activities, sites, and populations that can be
recombined as needed for different settings) (U.S.EPA, 1998). Another useful approach is to
employ hierarchical models, beginning with the simplest portrayal of the most important
elements and connections. The example conceptual model in Figure 2-1 illustrates a broad
overview of relationships among GSs, different types of stressors, and human health.
This conceptual model illustrates the occurrence of and interactions among four different
types of stressors commonly considered as part of a CRA: chemical, biological, physical, and
psychosocial. These stressors are linked to the physical location and context of a specific area.
Context includes not only physical factors such as climate and seasonal trends and physical
proximity to GS, but also the characteristics of an area (e.g., socioeconomic status, community
identity, and local infrastructure and policy), as well as individual characteristics related to
person-environment interaction.
Individuals exhibit their own intrinsic characteristics (e.g., age, gender), as well as
modifiable characteristics such as perceptions of safety. The model shows that the occurrence of
and exposures to these four different types of stressors can be influenced by the quality, quantity,
and specific environmental and sociobehavioral functions associated with GS. These
characteristics and functions, which serve as the basis for GS metric development and
assessment, are further illustrated in Figure 2-2. The interactions subsequently influence
environmental quality and physiological, psychological, and social pathways; collectively, these
affect human health outcomes. Note that this model is meant to convey an overarching view of
2-4
-------
how GS can interact with both stressors in the environment and populations; it is not intended to
be exhaustive.
2-5
-------
Climate and
Seasonal Trends
Proximity to
Greenspace
Physical Location
and Context
Area
Characteristics
Individual/Subgroup
Characteristics
> Hazardous chemicals in air, water,
and soil
~ Allergens and pathogens in air,
water, and soil
1 Heat
I Noise
»Visible & built environment
> Proximity to sources
; Low socioeconomic status
> Crime, social instability,
institutional disenfranchisement
^ Quantity Characteristics
^
Quality Characteristics
r "i
Characteristics and functions influencing the nature, frequency, and
duration of exposure to/ interaction with greenspace
^ -J
1
-
w
^ Environmental Functions
i wmrnm
Environmental Quality
Psychosocial Health
Social Connectedness
Physical Health
_
¦ Human Health I
State of Physical,
Mental, and Social
Well-Being
Figure 2-1. Overarching conceptual model illustrating features of greenspace (GS) exposures and health effects.
2-6
-------
Quantity Characteristics
Quality Characteristics
Size, shape
Tree coverage
Vegetation coverage
Type (e.g. park, garden)
Species composition
Design/aesthetics
Maintenance
Access/ownership
Amenities
Functional uses
r
Characteristics and functions influencing the
nature, frequency, and duration of exposure to/
interaction with greenspace
v J
Environmental Functions
Socio-Behavioral Functions
Pollen concentrations
Noise reduction
Heat reduction
Natural hazard buffers
Pollution reduction
B§1B
Figure 2-2. Greenspace (GS) characteristics and functions that influence
exposures and interactions.
Although GS research is growing rapidly, the uncertainty regarding mechanisms of action
for the variety of outcomes is a fundamental barrier to developing accurate estimates of risks or
benefits to health from GS. Greater use of conceptual models to depict specific relationships
between GS and human health could help further strengthen ongoing advances in GS analyses.
For example, conceptual models could help illustrate known and hypothesized mechanisms
between GS exposure and specific health endpoints (e.g., CVD or ragweed allergy) and potential
combinations of GS exposures with various chemical and nonchemical stressors. In many cases,
multiple features of the same GS might offer a different profile of benefits or risks for different
groups of people. Further, detailed conceptual models could also help delineate differential
exposures and effects, as well as key assumptions and uncertainties, for specific subgroups such
as age, gender, socioeconomic status, or urban and rural populations (Suter, 1998).
2-7
-------
(Thispage intentionally left blank.)
2-8
-------
3. GREENSPACE (GS) EXPOSURE ASSESSMENTS AND METRICS
Highlights of the work group presentations on GS exposure metrics and assessments are
presented in Section 3.1 and Table 3-1. The full sets of slides that accompanied these
presentations are provided in Appendix B, in the same order as the list of participants in
Appendix A. Key exposure considerations from these presentations and the accompanying
discussions are highlighted in Section 3.2. Approaches and metrics for GS exposure assessments
are described in Section 3.3.
3.1. TECHNICAL PRESENTATIONS
Multiple experts within the topical areas identified in the agenda led the presentations on
assessing GS exposures. Key points from each presentation are highlighted below.
3.1.1. Exposure Assessment Approaches
Laura Jackson (EPA National Health and Environmental Effects Research Laboratory
[NHEERL])
Mark Nieuwenhuijsen (Centre for Research in Environmental Epidemiology [CREAL])
Matilda Annerstedt van den Bosch (,Swedish Agricultural University)
Three distinct approaches to GS exposure assessment were described from ongoing
programs in the United States and Europe. In the United States, the EPA (2015) EnviroAtlas
flittp://enviroatlas.epa.gov/enviroatlas/) is a publicly available data resource for mapping and
evaluating ecosystem services supply, demand, and drivers of change. EnviroAtlas
conceptualizes GS as providing three categories of services that benefit human health: (1) buffers
against natural and anthropogenic hazards, (2) opportunities for healthful behaviors such as
active transport (e.g., bicycling or jogging) and social interaction, and (3) supporting
environmental functions such as carbon sequestration and wildlife habitats. This resource does
not include potential negative consequences for human health, such as from pollen exposures.
Environmental data on GS are available at the watershed level (hydrologic unit code [HUC] 12)
for the contiguous United States and at fine-scale resolution for selected communities.
Estimates of GS derive mainly from 30-m (roughly 100-ft) satellite imagery and aerial
photography. EnviroAtlas also provides health-related metrics such as tree cover along roads
and streams, temperature and pollution reduction by tree cover, and walking distance to parks,
schools, and day-care centers with surrounding GS. Sociodemographic variables are included to
assess distribution and estimate the potential for improvements in population health through GS
interventions.
3-1
-------
Table 3-1. Presentation highlights: greenspace (GS) metrics and exposure
No.
Topic
Scope notes
Presenters
Key points
1
Exposure
assessment
approaches
Overview of
how GS is
determined,
illustrated by
EnviroAtlas,
PHENOTYPE,
and World
Health
Organization
(WHO) GS
indicator
Laura Jackson {EPA)
Mark Nieuwenhuijsen (CREAL)
Matilda Annerstedt van den Bosch
{Swedish Agricultural University)
• Definitions of GS and potential health and
ecosystem services
• Different levels of measurement for indicators of
interest (e.g., level 1, 2, 3 for PHENOTYPE)
• Policy indicators developed broadly for universal
application and comparability
• Overview of specific metrics at various scales
2
Tree cover
measurements
Normalized
difference
vegetation
index (NDVI),
regional to
local urban tree
cover (UTC)
Geoffrey Donovan {U.S. Forest Service
[USFS])
Perry Hystad {Oregon State University)
• NDVI commonly used in epidemiological studies;
it is unclear what it actually captured
• NDVI is potentially useful for validating finer
scale measures
• Comparison of several tree cover estimates and
epidemiological applications
3
Access to
greenness
Richard Mitchell {University of Glasgow)
Michelle Kondo {USFS)
Matilda Annerstedt van den Bosch
{Swedish Agricultural University)
• Comparison of coarse- and fine-scale
measurements in the United Kingdom
• Measures of residential greenness vs. access and
use of GS
• Overview of vacant lots as potential stressors or
ecosystem services
• No standard scientific measure of access
3-2
-------
Table 3-1. Presentation highlights: GS metrics and exposure (continued)
No.
Topic
Scope notes
Presenters
Key points
4
Built
environment
(BE)
Perry Hystad (Oregon State University)
Yvonne Michael (Drexel University)
• Objective vs. qualitative measures for BE
• Natural experiments are currently the main
approach for assessing GS and BE together
• Variables and relationships differ at different
scales
• Discussion of potential data sources for BE-GS
assessment
5
Design,
environmental
psychology
Includes
canopy shape,
way-finding
strategy,
attention/
cognition
restoration
Julia Africa (Harvard)
William Sullivan (University of Illinois,
Urbana-Champaign [UI-UC])
Richard Mitchell (University of Glasgow)
• Underlying: psychoevolutionary theory and
Attention Restoration Theory (ART)
• Biophilic design, viewable GS and insights for
CRA
• Exposure duration (defined as time per event),
frequency, concentration of GS, mode of delivery
• Cultural differences in definition and value of GS
6
Specific
populations,
exposure
Includes aging,
lower
socioeconomic
status (SES)
Yvonne Michael (Drexel University)
Richard Mitchell (University of Glasgow)
J. Aaron Hipp (North Carolina State
University)
• Exposure differences by subgroup
• Gender, race, socioeconomic position, age
• Sequence of exposure over the life course
• Intersections with accessibility
7
Exposure
metrics, links
to health
Illustrated by
attention
restoration
effect
William Sullivan (UI-UC)
Yvonne Michael (Drexel University)
• GS exposure elements include frequency,
duration, and nature of interaction with GS
• Experimental approaches for estimating GS
exposure
• Beyond physical contact; GS views linked to
reduced stress and improved performance
• Estimating exposure (dose)-response
3-3
-------
Specific metrics to estimate GS exposure and related health impacts from land cover
composition were described and categorized as indicators of healthful exposures and/or
indicators of effect measure modifiers. For example, the amount of visible tree cover from a
residence is estimated with a 50-m buffer around the residence, and this 50-m buffer estimate is
applied as an indicator for both potential engagement with natural features (healthful exposure)
and for potential protection against heat, air pollution, and night light (effect modification)
(U.S. EPA, 2015).
In Europe, the European Union has established a project to investigate the
interconnections between exposure to natural outdoor environments (in both rural and urban
settings) and better human health and well-being. Referred to as the Positive Health Effects of
the Natural and Outdoor environment in Typical Populations in different regions in Europe
(PHENOTYPE, http://www.phenotype.eu/), thist project is coordinated by the CREAL in
Barcelona, Spain. The PHENOTYPE program assesses the natural environment via a multilevel
approach using quantitative, qualitative, and in-field assessments to estimate GS exposures and
associated health effects. The approach has three levels of analysis. The first consists of
quantitative assessments that involve objective measures of natural environments, such as
large-scale vegetative coverage, and that evaluate effects with secondary health data from
epidemiological studies. The second involves assessments that use detailed secondary data to
estimate the quality of the natural environments for smaller areas, such as data on the specific
attributes of municipal parks (e.g., whether a play area is available for children, or whether there
are hiking trails). The third approach is at the most localized scale, involving primary data
collected through environmental audits and used to measure environmental quality.
These three levels provide different opportunities for GS exposure assessments. At the
broadest scale, measures such as the normalized difference vegetation index (NDVI) are useful
for estimating generalized availability of GS within an area and provide the opportunity to
compare different regions and cities. However, these measures do not provide the level of detail
necessary to estimate access to greenness, which can only be obtained under the second and third
levels of analysis. GS quality can be assessed across several categories, including ownership,
size and shape, functional uses, location, management and perception of management,
community identity, and climate factors.
The second featured European program focuses on the World Health Organization
(WHO, 2001) indicator, which was developed in response to the 2010 Parma Commitments that
aim to provide each child with access to urban GSs for play and physical activity by 2020. As
part of specific planning toward and monitoring of this goal, a measure was developed to define
GSs and appropriate spatial metrics for evaluation. The methodology for this measure was
designed for easy and widespread practical use with publicly available data, with an emphasis on
3-4
-------
screening and on comparability among regions over performance in statistical or spatial
modeling. The measure uses population distribution and geographic information system (GIS)
data to calculate and provide the ratio of people within a designated buffer area of a GS to the
city's total population, as an implicit environmental health indicator.
GSs were defined spatially as green-covered areas that are a minimum of 1 ha in size
(roughly 2.5 acres), providing a recreational use (such as a park, public garden, or zoo),
including suburban areas managed as parks and green areas adjacent to urban areas, per the
definition of urban GSs from Urban Atlas (http://www.eea.europa.eu/data-and~maps/data/urban-
atlas). In a case study of Mai mo, Sweden, a Euclidian buffer distance of 300 m (roughly
1,000 ft) around GSs was chosen to represent accessibility for the city. While intended for use in
assessing the public health endpoints of physical activity and stress, the measure does not include
estimates of quality or actual walking distances to GSs. The measure is to be validated in
relation to health data and tested in further case studies in Europe.
3.1.2. Tree Cover Measurements
Geoffrey Donovan (U.S. Forest Service [USFSJ)
Perry Hystad (iOregon State University)
Several data sources and methods were described for determining tree cover for large
areas. The NDVI is derived from satellite imagery of chlorophyll and is the measure most
widely applied in epidemiological studies of surrounding greenness and health. Several studies
use GIS to derive the mean NDVI value for the determined buffer distance(s) around the point of
interest (e.g., a residence or a school) as a measure of surrounding greenness (Agay-Shay et al.,
2014; Amoly et al., 2014; Dadvand et al., 2014; Hystad et al., 2014). Recently, the standard
deviation of NDVI rather than the mean value has been used as a measure of variation in
greenness, which could differentiate areas with both green and built features (Periera et al.,
2012).
Other satellite-derived imagery sources include the moderate resolution imaging
spectroradiometer (MODIS) NDVI, enhanced vegetative index (EVI), U.S. Forest Change
Assessment Viewer (ForWarn), and the MODIS leaf area index (LAI). These sources can
provide more detail on the type of greenness and phenology, but they are not frequently used in
GS exposure assessment because of lower resolution imagery. Land cover classification is
another measurement of large-scale greenness. Land cover classifies imagery along several
categories, which can range from basic classifications (e.g., tree canopy, water, buildings) to
more explicit categories for land use and vegetation cover (e.g., parkland, roadways, industrial
areas). Available at different scales and resolutions, land cover data sets are particularly useful
3-5
-------
in measuring change in greenness over time. Light detection and ranging (LiDAR) is another
sensing method measuring elevations on Earth using a pulsed laser (by illuminating a target and
analyzing the reflected light to estimate distance). This method can detect vegetation cover and
land use in three dimensions and at high resolution. However, data are more difficult to interpret
from this method so it is not used as frequently as NDVI and other measures of land cover.
At a smaller scale of analysis, data on urban tree cover (UTC) and land use are
increasingly available at the municipal level. These can provide information on more specific
attributes such as tree species composition, tree removal and loss (e.g., due to emerald ash borer
infestation or to widen streets), and individual park features and access points. With the variety
of measures available, the emphasis is on determining exposure metrics relevant to the health
outcome of interest, the population, and area characteristics. Much is yet to be learned about
specific patterns of GS use that are difficult to ascertain from measurements based on imagery.
There is a growing emphasis—especially for health data—on integrating large-scale objective
measures such as those that can be collected with NDVI, with individual data like those gathered
from global positioning satellites (GPS) and more subjective data obtained from surveys and
environmental audits. Additionally, measurements of GS from a street-level perspective rather
than bird's eye imagery may be useful in assessing visible greenness. With a consensus neither
sought nor reached, the following topics were also discussed at the meeting:
• NDVI as predictor of health effects: What exactly does NDVI measure? Is analysis of
mechanisms possible using only NDVI?
• Residential greenness, passive proximity versus active engagement for access and use:
What are the health outcomes most associated with each, and what is their potential joint
influence on various outcomes?
• What are the best methods to consider temporality, such as seasonal variation in both GS
and health effects, and acute versus chronic exposures?
• Is there a threshold for exposure? Consider floor versus ceiling for visible effects of GS.
3.1.3. Access to Greenness
Richard Mitchell (University of Glasgow)
Michelle Kondo (USFS)
Matilda Annerstedt van den Bosch (,Swedish Agricultural University)
3-6
-------
Different definitions of GS were discussed, as were other natural features that could be
included (e.g., water bodies/features, or "blue space"), acknowledging that no single consensus
definition exists for what constitutes GS. Research highlights were shared from the Center for
Research on Environment, Society and Health (CRESH, http://cresh.org.uk/), which examines
GS using land cover data at multiple resolutions. Scotland's map compiles GS data from all
32 Scottish council areas, a political subdivision unit, and classifies 23 unique spaces using
primary and secondary codes to capture spaces with multiple functions, such as woodland areas
within parks. Using GIS, a buffer analysis can produce a measure of the proportion of green
land cover within an area, as well as the population within a certain distance to GS. Different
data sets can produce different results using this method. Differences can be illustrated by
comparing the percentage of GS based on three different GS data sets. These data sets include
(1) the Coordination of Information on the Environment (CORINE,
http://www.eea.europa.eu/publications/CQR04andcover) data set, derived from satellite
imagery; (2) the Ordnance Survey MasterMap (the United Kingdom's most detailed
vector-based data: https://www.ordnancesurvev.co.uk/business-and-
government/products/mastermap-products.htmO; and (3) CRESH's own model-based estimates
of the percentage of GS land cover.
Two key considerations for assessing access and exposure to greenness are: (1) the scale
of analysis (e.g., neighborhood, municipality, or regional) and (2) the assumption of accessibility
through measures of proximity and coverage. A short distance to objectively measured GS does
not necessarily mean that GS is accessible (e.g., it could be a private park). Similarly, a measure
of GS coverage in an area does not provide information on the frequency, duration, and nature of
use for populations of interest. Other challenges in measuring access were discussed such as
visible greenness, exposure in indoor spaces, and individual time spent in a GS. Monitoring
devices using GPS sensors are one method used to inform GS exposure estimates with time and
location data, although these tend to be used only in relatively small studies.
Vacant lot greening efforts in Philadelphia provide a valuable opportunity for considering
a natural experiment design and for conducting GS access research. Findings are inconsistent for
the potential health benefits and risks of greened vacant lots. Using GS audits and resident focus
groups and interviews, a recent study (Heckert and Kondo, under review) examined the potential
impact of vacant lot greening on residential perception and access. Results indicate that many
residents did not notice when a lot had been greened, and they were unsure how to interact with
the space in this new condition, but continued negative perceptions were reported around
remaining vacant lots.
This overview led to a renewed group discussion of GS definitions with specific attention
to perceptions of quality or design that might influence associated health effects. For example,
3-7
-------
while an overgrown vacant lot could contribute to an NDVI-based measure of GS, it might
actually have adverse effects for nearby residents. Additionally, factors such as design and
community involvement around projects such as vacant lot greening could substantially
influence perceptions and ultimate community impacts of newly greened spaces. The effect of
participatory design and accessibility (e.g., perceived access, perceived safety, perceived
function) regarding green interventions is an area needing further research.
3.1.4. Built Environment
Perry Hystad (iOregon State University)
Yvonne Michael (Drexel University)
Key aspects of GS accessible or implemented through changes in the built environment
were identified. Defined as the human-made spaces in which people work, live, and play, the
built environment is often determined locally, through land use planning, zoning ordinances, and
design guidelines. Parks, trails, green roofs, community gardens, and green stormwater
infrastructure are all examples of planned GSs in the built environment. Increasing access to GS
is an emerging policy priority in many urban areas for its potential ecological and social benefits,
and the evaluation and validation of such interventions are developing areas of research.
Natural experiments provide an opportunity to evaluate the influence of GS while
minimizing some of the confounding inherent in observational studies and improving the causal
inference available from cross-sectional studies. Addressing the definition of GS as part of
urban open space, the impacts of transforming blighted vacant land into GS from a study in
Philadelphia were discussed, including reductions in gun assaults, vandalism, and self-reported
improvements in stress and exercise (Kondo et al., 2015; South et al., 2015).
In studies of the built environment, the measurement of GS includes objective
measurements such as land use and NDVI, residential proximity to parks, and street-level audits
using Google Earth. Subjective measures such as qualitative GS audits and residential surveys
are also used. It is important to control for other built environment factors when evaluating the
effect of GS because that can be highly correlated with other variables. A study of GS and
reproductive outcomes indicated a lower risk for several birth outcomes in higher NDVI areas
after controlling for air pollution, park proximity, walkability, and noise (Hystad et al., 2014).
Further, qualitative assessments of microenvironments have been found to impact social
functioning and could modify the effect of GS exposure (Brown et al., 2008). Finally,
crowdsourcing studies are another avenue of analyzing visible GS, with some finding subjective
classifications of "happy," "beautiful," and "quiet" in images with greenery compared to images
with no greenery (see Mappiness, http://www.mappiness.org.uk/).
3-8
-------
3.1.5. Design and Environment Psychology
Julia Africa (Harvard T.H. Chan School of Public Health)
William Sullivan (University of Illinois at Urbana-Champaign)
Richard Mitchell (University of Glasgow)
Design features can affect exposures to GS. Considering connections between humans
and nature from a psychoevolutionary perspective relative to human history, the built
environment can be viewed as a novel living space for humans compared to natural
surroundings. Current research tends to create a false dichotomy between psychological and
physiological responses to GS; we know less about how strong the relationships are between
individuals and landscape characteristics that support observed health effects. Some research
supports a link between nature and human health regardless of the type of exposure, finding
reduced Cortisol levels for both those who viewed a forested area and those who walked through
it (Park et al., 2010). Meanwhile, the potential for adverse impacts of GS also exists, as
illustrated by increased crime, air pollution (e.g., pollen), or an individual's fear of certain
natural features (e.g., dark areas or poison ivy). GS in rural environments is also a growing area
of research that presents many questions, considering the different perspectives and experiences
of rural populations with nature and associated impacts.
Theoretical background and experimental evidence of Attention Restoration Theory
(ART), which hypothesizes that mental concentration improves after exposure to nature, builds
on the early work by Kaplan and Kaplan (1982, 1989) and Kaplan (1995). The ART researchers
assert that attention takes one of two forms in the brain: involuntary or directed. Involuntary
attention requires little effort and produces little mental fatigue. Items, ideas, or places that are
fascinating draw on involuntary attention. Conversely, directed attention requires a high degree
of effort and focus and is used for tasks such as learning, problem solving, and planning. The
focused effort required for directed attention fatigues mental processing more easily than
involuntary attention. Mental fatigue can lead to inattentiveness, irritability, and impulsive
decision making. According to ART, brief exposure to natural settings and scenes activates
involuntary attention, which allows for rest and recovery of the ability to focus attention.
Interrupting directed attention with restorative stimuli could lead to improvements in memory
and performance. In the two experimental studies of ART discussed, the performance of
participants was measured in response to different exposures to GS. Measures of attentional
performance before and after the GS exposure indicate that exposure to greenness is associated
with improvements in cognition and memory.
It is worth noting that ART is one of several theories used by environmental
psychologists to describe observed effects of nature on human consciousness, well-being, and
3-9
-------
health. Other notable contributions include the biophilia hypothesis (Wilson, 1984; Kellert and
Wilson, 1993), the savanna hypothesis (Orians and Heerwagen, 1992), the habitat theory, and
prospect-refuge theory (Appleton, 1975). A more recent contribution by Heerwagen (2006)
described a framework for "features and attributes of buildings linked to well-being needs and
experiences" (e.g., views of outdoor nature, natural lighting, interior plantings), reflecting the
relationship between nature and health in anthropocentric terms. Cramer and Browning (2008)
consolidated these observations in three broad experiential categories describing human-nature
relationships: nature in the space, natural analogs, or nature of the space (see Ryan et al., 2014).
Further discussion focused on the contribution of specific natural design features to
observed responses. In a classical exposure-response model, design features influence positive
or negative appraisal of an individual's adaptive capacities, with a negative appraisal increasing
the likelihood of psychological and physiological stress responses such as elevated Cortisol,
blood pressure, and heart rate (Cohen et al., 1995). All current psychoevolutionary theories
described above suggest that natural environments are more likely to soothe or positively
stimulate our neurobiology as compared with most features of the built environment. Notably,
Edward O. Wilson created the term "biophilia" to describe the "innate tendency [in human
beings] to focus on life and lifelike processes," suggesting that these responses are cross-cultural
and ahistorical.
Design features that echo the movement, variability, and periodicity found in nature
through stimulation of the five senses can influence physiological and psychological responses.
Examples of natural design elements include the use of fractal and Fibonacci sequences found in
babbling brooks, dappled sunlight, or flower petal structures; the use of natural fibers and
materials like stone in built environment settings; and the elevation of natural "soundscapes" like
birds singing as a component of GS. Exposure was discussed in terms of frequency, duration,
and intensity of immersion over various intervals; the interplay between indoor exposure to
natural design elements, and a potential "priming" effect for outdoor exposures was briefly
discussed as an area warranting future research.
The government of Singapore has incorporated periodic exposure to GS as part of its
Nature Pyramid, a food and healthy living guide patterned after the U.S. Food and Drug
Administration's food pyramid. Singapore's example illustrates the potential for integrating GS
exposure into the urban fabric (Beatley, 2012). The Nature Pyramids model has inspired the
Biophilic Cities Network (http://www.biophiliccities.com), a group of urban planners, landscape
architects, and public health clinicians that seeks to transform our urban model to support our
innate affinity for and exposure to nature.
3-10
-------
3.1.6. Specific Populations: Exposure
Yvonne Michael (Drexel University)
Richard Mitchell (University of Glasgow)
J. Aaron Hipp {North Carolina State University)
The potential for differential exposure to GS for specific population subgroups involves
considering differences across factors such as life stages, gender, and socioeconomic variables.
The need to address specific populations stems from the multiple functions of GS and multiple
pathways of GS exposure. GS can be considered as having both salutogenic (beneficial to
health) and pathogenic (detrimental to health) effects, and as such, has the potential to increase
or decrease inequalities in health through public health interventions (e.g., path creation and
health walk promotion). Population subgroups are described in relation to both differential
exposure and susceptibility. To illustrate the exposure aspect, proximity or access to GS could
be a function of environmental justice or cultural and behavioral factors determining frequency,
perceptions, and type of activities occurring in the GS. Gender differences reported for GS and
health relationships emphasize the importance of capturing quality as well as quantity of
greenspace and not assuming uniform health benefits of GS for all population subgroups
(Richardson and Mitchell, 2010). In a recent analysis, women were found to be likely to use GS
at lower levels than men, although women reported using GS for similar purposes (Miller et al.,
2014). Considering susceptible populations, certain subgroups might have particularly positive
or adverse responses to a certain GS. For example, an asthmatic child could be more at risk of
an asthma attack in areas with high tree pollen, and that potential could be magnified in the
presence of other common urban irritants such as diesel exhaust from heavily trafficked
corridors.
In addition to particular health conditions, attention to subgroups with different social,
cultural, and environmental norms and expectations was discussed as influential when estimating
exposure. Specific grouping categories include gender, socioeconomic status, race, ethnicity,
and life course. While some data suggest that the quality of a GS is more important to women,
the evidence is mixed regarding differences in use among the men and women who do access
GS. Whether and how this affects the estimates of GS exposure is an area needing further
research to assess whether the pathways affecting access or use could differ between men and
women.
Studies of GS often examine socioeconomics variables such as income and level of
education. In general, results indicate that access to quality GS decreases with socioeconomic
position and among minority race communities (Wolch et al., 2014), and there is evidence that
GS use is lower among lower-income and minority groups even when GS is available (Jones
3-11
-------
et al., 2009; Hipp et al., 2013a). After controlling for socioeconomic position, Suminski et al.
(2012) found that nonwhite Midwestern communities had the least access to GS and the number
of amenities within GS/parks. Others have shown that for GSs that are used, minority races and
those of Hispanic ethnicity are more active in the GSs compared to whites (Floyd et al., 2008);
this result was associated with amenities within the spaces, such as sport facilities. Perceptions
of GSs are also important in determining access, and evidence suggests that different groups
perceive spaces differently. Considering exposure, it was suggested that assuming access via
proximity alone is insufficient for understanding the complex relationship between proximity,
access, and use among different groups.
Finally, considerations for exposure were discussed from a life-course perspective, which
examines the changes in certain influential factors for different age groups. Risk and exposure
were again discussed in terms of access and susceptibility at certain stages of development and
life. For example, children might have limited access to GS depending on their proximity,
mobility, and the rules of their household. Older adults may also have limited access, but for
other reasons than children do. Sensitivity to allergens and chronic conditions can also vary by
age. Evidence was presented that adults are more likely to access parks than either children or
older adults, whereas teens are more likely to access parks within a mile of their homes (Cohen
et al., 2006). The discussion centered around the need to consider differences in exposure and
susceptibility in further research studies, which have rarely considered data based on actual visits
and activities in GSs.
3.1.7. Exposure Metrics, Links to Health
William Sullivan (University of Illinois at Urbana-Champaign)
Yvonne Michael (Drexel University)
A number of metrics can be used to assess exposure and response related to GS.
Frequency, duration, and the type of the interaction with the given GS were discussed as central
components of the exposure evaluation. Exposure is not limited to physical contact with nature;
research studies indicate that having a view of GS is associated with improvements in mental
performance and stress reduction (Kuo and Sullivan, 2001; Ulrich et al., 1991). Although
specific metrics can differ across studies, most research includes a measure of canopy cover or
canopy density as an initial environmental exposure, with qualitative and individual information
on use added as available or feasible. Controlled experimental studies have implemented canopy
density as a "dose" of GS, with participants exposed to varying amounts of green imagery (Jiang
et al., 2015).
3-12
-------
Biomarkers such as Cortisol levels and skin conductance have been used to measure stress
responses across different GS exposure settings. With the ability to control short-term exposures
to greenness, studies such as these are building evidence of acute exposure and health response
(Jiang et al., 2014a, 2014b). Additional research is needed to develop measures of long-term
exposures and effects. Note that canopies are not the only GS metric warranting further
examination; beyond blue space, a number of vegetation types can be identified that are not
related to trees. In addition, surface features that are not vegetation (and are simply colored
green) are captured in certain measures of GS. Therefore, a clear understanding of each GS
measure is important to their relevant applications and potential combinations and extrapolations
across studies.
3.2. KEY EXPOSURE CONSIDERATIONS
The technical meeting discussions identified several essential considerations for assessing
exposure to GS and natural environments. The exposure measures evaluated are commonly
guided by the health outcome of interest. Coverage and distance measures are widely used in
ecological studies of various health and other population-level outcomes, including reproductive
and respiratory outcomes, mortality, and housing values (Mitchell and Popham, 2008; Dadvand
et al., 2012a, 2012b; Hystad et al., 2014; Li et al., 2015). Measuring exposure to assess health
outcomes such as psychological stress or attention restoration (AR) requires different, often more
complex accounts of the nature, duration, and frequency of an individual's interaction with GSs.
Location is another essential component of exposure assessment, not only to estimate
coverage and proximity to vegetation but to account for the local contexts shaping the nature of
and interactions with GS in a specific area. These contexts can include physical features of the
GS like topography, land use, ownership patterns, and other cultural and sociodemographic
norms (Lachowycz and Jones, 2013; Berland et al., 2015).
After establishing the health effect or effects of interest and the context of the GS, the
exposure assessment commonly considers qualitative and quantitative features of the actual GSs.
Quantitative measures are largely objective and address elements such as size, shape,
distribution, and distance to the GS(s) in an area. Quality estimates relate to the nature of the
space and the types of activity or engagement it can provide; amenities, maintenance, public
accessibility, and functional uses are common qualitative measures (Nieuwenhuijsen et al.,
2014). Examples of common measures are discussed in Section 3.3 and summarized in
Table 5-1.
To the extent possible, exposure assessments should also consider the reasons for and
nature of the populations' engagement with the GS (e.g., for physical activity, to attend a social
event, to fish for leisure), as well as the frequency and duration of the interaction; this includes
3-13
-------
potentially different impacts of short-term versus long-term engagement with natural spaces.
Such assessments should allow for multiple types of engagement, perhaps even within a single
trip. The report authors acknowledged that this level of detail is not often feasible for large
ecological studies. Finally, the frequency, duration, and nature of the interaction with GSs often
vary across different populations (Rosso et al., 2011; Ord et al., 2013; Sullivan et al., 2014).
In addition to accounting for the proximity and nature of engagement with the GSs
themselves, assessments should evaluate potential exposure to other environmental and social
stressors and the implications for susceptible populations. Such exposures can occur within,
adjacent to, or outside the GS setting or interaction area; for example, trees in a park may filter
some pollutants in the air, but overall exposure to air pollution may increase if the park is near a
major highway or intersection. Engagement with GS can also have adverse effects for certain
subgroups, namely those susceptible to aeroallergens or asthma triggers (Tzoulas et al., 2007,
Lovasi et al., 2013).
Several factors in the social environment, from local policies and cultural norms to
traffic, noise, or crime rates, can inform attitudes and engagement with GS. Community and
individual perceptions of safety, function, and accessibility may differ across subgroups and can
affect the nature and duration of activity within and around GSs (Mitchell et al., 2015). These
complex, place-based factors can influence the use of GS as well as present additional stressors
that can mediate or otherwise intersect with the benefits or risks associated with GS engagement.
3.3. ASSESSMENT APPROACHES AND METRICS
A tiered approach is generally applied to assess GS exposures, with the level of detail
depending on the population and health outcome of interest, available metrics, and time and
other resources available to those conducting the assessment. Considerations of scale, data
accessibility, and comparability across different areas have led many researchers to rely on
large-scale spatial estimates of GS, while questions of access and quality are determined at
smaller units of analysis. At the largest scale, satellite-derived imagery can produce estimates of
world-wide vegetation cover at various resolutions.
One of the most widely used metrics is the NDVI, which is based on the natural infrared
light-reflective properties of chlorophyll. The methodology and interpretation of the NDVI are
detailed elsewhere (Weier and Herring, 2011); the index ranges from -1 to 1, with lower values
indicating little to no greenness. Daily NDVI values are available from 1972 to present at a
30-m resolution, the smallest available for global satellite imagery. While different types of
earth image data such as LiDAR and MODIS are available, NDVI is a widely applied
measurement due to its availability, relative ease of interpretation, and comparability across
studies.
3-14
-------
A second resource for large-scale GS measurement is land cover data. These data
provide information on different types of land use and vegetation. Large-scale land cover
databases such as CORINE in Europe and the National Land Cover Database in the United
States are widely used for GS quantification (Richardson and Mitchell, 2010; Annerstedt et al.,
2012). Land cover data can distinguish among different types of GS, where NDVI cannot. Such
a metric is useful for tracking greenness over time, both for research purposes and for policy
development goals, including those identified in the WHO/Europe Parma Commitments
(Annerstedt van den Bosch et al., 2014a, 2014b).
At smaller spatial scales, land use and ownership information is available for many
municipalities and counties across the United States and Europe, often at an individual parcel or
lot level. These types of data provide more contexts for GS within the built environment, for
example, allowing the user to distinguish a public park from a private cemetery, determine
access points, or observe the types of features (roads, parking lots, buildings) that surround the
GSs. Assessments at this level have much more detail than available from satellite imagery and
are mostly used for studies at the municipal or neighborhood scale. Created and managed
locally, land use data sets are more difficult to compare across studies, as classification and
resolution may vary (Nieuwenhuijsen et al., 2014).
Information can also be analyzed with area demographics, such as those collected by the
U. S. census or the American Community Survey, which are available at census block levels and
higher. For more qualitative spatial assessments, open-source resources such as OpenStreetMap
can be used to identify specific GS features, such as parks, trails, or playgrounds (Agay-Shay
etal., 2014).
Data from large-scale surveys are also used to measure access to and use of GS. These
data cover a larger extent than many field-based measures, but they are distinct from land cover
and GIS-based measures of exposure. In such surveys, respondents are asked to report their
access to and/or use of GS. Sometimes these are relatively crude measures. In the European
Quality of Life surveys, for example, respondents are asked to rate their ease of access to
recreational//green areas, and whether they use them. However, sometimes the measures are
sophisticated, with environment type, visit purpose, and activity carefully captured. The English
Monitoring Engagement with the Natural Environment survey is an example. In general, the
more sophisticated the capture of exposure to and use of natural environments, the less
sophisticated any accompanying health metrics tend to be. Self-reported general health, for
example, might be recorded for use as a predictor of contact with nature, rather than as a
consequence. Yet, larger surveys capturing exposure to and/or use of GS in less detail often also
capture useful self-reported health metrics such as mental well-being. Where fine-scale
3-15
-------
geographic identifiers are available, the measures of natural environment captured in all these
kinds of survey can also be compared to surrounding GS (Grigsby-Toussaint et al., 2015).
The most detailed methods used in GS exposure analysis occur in the field; GS audits,
questionnaires, or monitoring data from participants can provide a full picture of perceptions,
lifestyle decisions, and actual engagement with GS (Kondo et al., 2015). GS and street audits are
conducted either in person or by using street imagery, such as Google Street View, and they
assess physical features as well as real-time population counts.
Questionnaires can provide information such as self-reported physical activity, stress
levels, and perceptions of nearby GS availability, safety, and self-reported engagement
(Nieuwenhuijsen et al., 2014; Kondo et al., 2014). Although many researchers obtain biomarker
data in clinical settings, monitors worn by participants can also record several types of behavior
and exposure, including real-time location and movement, environmental exposures, heart rate,
and accelerometry (Almanza et al., 2012; Adlakha et al., 2014; Ryan et al., 2015).
Studies are beginning to investigate responses to GS exposures in experimental settings
(Jiang et al., 2014a, 2015; South et al., 2015). Online methods such as crowdsourced image
analysis are also being used to assess built environment quality (Hipp et al., 2013b; also see
Mappiness, http://www.mappiness.ore.uk/). While these methods can produce a more complete
picture of individual exposure, they require significant human and technological resources.
Often, research that uses monitoring or in-person audits is applied for specific places or projects,
such as a neighborhood block, school, or a planned event or intervention. Studies using smaller
sample sizes or specific geographical areas commonly face limitations to external validity and
the ability to compare among different populations and areas.
There is emerging consensus that GS exposure assessments should consider more
detailed exposures related to duration and frequency, population subgroup, and the specific
design and communication attributes of the GSs being studied. Temporal variation in both GS
and human engagement can affect exposure over short-term periods, but also by season and
across the life course (Rosso et al., 2011). Additional research is needed to identify the
appropriate distances and times at which to evaluate GS, and whether these may vary across
different pathways (U.S. EPA, 2015).
Hunter et al. (2015) present evidence that specific features of a space, such as design of
amenities and species composition, can affect perceptions and usage of GSs. However,
communication and local context is important to understand how GSs are integrated with a
specific area and how to optimize engagement surrounding an intervention to improve public
health (e.g., activity programming for a new multi-use path) (Nieuwenhuijsen et al., 2014;
Hunter et al., 2015). Developing and validating metrics to account for these issues is an ongoing
effort in GS exposure assessment research.
3-16
-------
4. GREENSPACE (GS) AND HEALTH
The technical presentations on health effects of GS are summarized in Section 4.1, and
key points are highlighted in Table 4-1. Key health considerations identified from these
presentations and the subsequent group discussions are highlighted in Section 4.2. Approaches
and measures for GS health assessments are described in Section 4.3.
4.1. TECHNICAL PRESENTATIONS
Multiple experts led the presentations on assessing health effects of GS within the topical
areas identified in the agenda. Key points from each are highlighted below. While the state of
research on each of these outcomes differs substantially, this section aims to present the current
established and hypothesized pathways for GS exposure and the mechanisms through which it
may impact specific populations and health outcomes.
4.1.1. Respiratory Effects
Patrick Ryan (University of Cincinnati)
Geoffrey Donovan (USFS)
An evaluation of studies of allergic reactions that focused on respiratory conditions such
as asthma indicated that some chronic effects are linked to repeat occurrences of acute effects,
such as chronic inflammatory airway disorder associated with repeated asthma attacks.
Mechanisms are well understood for respiratory effects from pollen, air pollution, and specific
indoor allergens (e.g., dust mites, cockroaches, and pets). Less understood are how these factors
might interact with conditions in the physical or social environment to influence asthma (i.e.,
environmental conditions that can exacerbate or moderate the frequency and severity of asthma
attacks). Some are related to GS (e.g., pollen and physical activity), while others are related to
community or physical attributes (e.g., crime rate, weather, and community resources).
Sociobehavioral benefits from GS are well established. Benefits include crime reduction, stress
reduction, and an increase in people's physical activity.
Environmental benefits are less pronounced and somewhat dependent on the metrics
used, but they include reductions in heat, noise, and air and water pollution. Among the
deleterious environmental impacts of GS are increased exposure to allergens, certain volatile
organic compounds (VOCs), and pesticides/fertilizers (e.g., from GS maintenance).
4-1
-------
Table 4-1. Presentation highlights: health effects of greenspace (GS)
No.
Topic
Scope notes
Presenters
Key points
1
Respiratory
effects
Includes
allergy/
asthma and
beneficial
effects (e.g.,
air filtration)
Patrick Ryan (University
Cincinnati)
Geoffrey Donovan (USFS)
• Potential benefits of GS exposure
and mechanism
• Potential risks of GS exposure
and mechanism
• Exposure characterization issues
• Previous findings and
preliminary results from
asthma/allergy cohort: NDVI
more associated than access
2
Reproductive
effects
Geoffrey Donovan (USFS)
Perry Hystad (Oregon State
University)
Yvonne Michael (Drexel
University)
• Consistent findings for improved
birthweight with higher GS
exposure
• Small signals but potentially
significant at population and
economic scale
• Discussion of birthweight as
outcome—"blunt" measure,
could be refined to other markers
• NDVI more associated than
access, more evidence needed to
establish mechanistic pathways
3
Obesity and
physical
activity
Matilda Annerstedt van den
Bosch (Swedish Agricultural
University)
J. Aaron Hipp (NC State
University)
• Evidence of urban parks as
settings for health benefits related
to physical activity
• Uncertainties around proximity
vs. activity in parks
• Use of big data: crowdsourcing,
monitors, and mobile devices
• Policy implications
4
CVD and
mortality
Includes
cause-
specific and
all-cause
mortality
Perry Hystad (Oregon State
University)
Mark Nieuwenhuij sen (CREAL)
• Mortality difficult to relate to GS;
requires large sample size
• Small reduction in all-cause
mortality from GS
• Reduced GS via removal of ash
borer-infested trees related to
increased respiratory and CVD
deaths
• Moderate/high evidence
regarding GS and non-CVD
effects
4-2
-------
Table 4-1. Presentation highlights: health effects of greenspace (GS)
(continued)
No.
Topic
Scope notes
Presenters
Key points
5
Neurologic/
neurodevelop-
mental effects
Mark Nieuwenhuij sen (CREAL)
Patrick Ryan (University
of Cincinnati)
• Mental health disorders fourth
leading cause of disability
adjusted life years (lost) (DALYs)
• Indirect benefits on child neuro
health via noise, mixed re air
pollutants
• Direct benefits of recovery from
fatigue
• Reduction in stress and crime
• Possible mechanisms of increased
physical activity and social
interactions
6
Psychosocial
effects
Michelle Kondo (USFS)
Matilda Annerstedt van den
Bosch (Swedish Agricultural
University)
Julia Africa (Harvard TH Chan
School of Public Health)
• Stress biology is multifaceted
• Expanding use of mobile
technology for "GS exposure"
• Some focused research: recently
greened vacant lots, use of virtual
reality nature scenes and sounds
• Mechanisms not always
understood (social stressors like
racism or perceived inequities
diminish resilience and increase
the need for stress mitigation
activities or resources like GS)
7
Attention
restoration/
cognition
J. Aaron Hipp (NC State
University)
Laura Jackson {EPA)
• Four mechanisms proposed
• Attention restoration related to
school performance and elderly
cognition
• Nature appreciation and social
interaction can reduce stress (Note
that stress is distinct from mental
fatigue, which can result from
stressful and delightful
experiences, and the underlying
physiological and neural pathways
differ.)
8
Economic and
community
benefits
Includes
property
values,
crime/safety
Michelle Kondo (USFS)
Geoffrey Donovan (USFS)
• Perceptions: safety vs. property
attractiveness vs. energy savings
• Strong influence of community
involvement
4-3
-------
Table 4-1. Presentation highlights: health effects of greenspace (GS)
(continued)
No.
Topic
Scope notes
Presenters
Key points
9
Specific
populations
and health
Includes
age, lower
socioeco-
nomic status
(SES)
Richard Mitchell (University
of Glasgow)
Patrick Ryan (University
of Cincinnati)
• Mixed benefits per gender
• Equigenesis not established yet
• More benefits to those in financial
difficulty
• Differential impacts on population
age groups needs info on effect
susceptibility vs. age
Highlights of the Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS) that
focused on traffic-related air pollution (TRAP) showed that closer proximity to objectively
measured GS (NDVI) was associated with a lower probability of asthma, although closer
distance to the nearest park was associated with higher probability for asthma for both low- and
high-TRAP conditions (Ryan et al., 2015).
4.1.2. Reproductive Effects
Geoffrey Donovan (USFS)
Perry Hystad (iOregon State University)
Yvonne Michael (Drexel University)
A recent evaluation of GS effects on reproductive outcomes identified three possible
causal pathways (Kihal-Talantikite et al., 2013). The most plausible biological pathway is the
psychosocial pathway where GS affects maternal stress through a psychoneuroendocrine
mechanism. Health improvements included promotion of psychological restoration,
improvement of attention, and reduction of stress and anxiety.
A physiological pathway where GS affects maternal health has been identified but is only
considered hypothetical at this stage. Postulated physiological benefits include changes
(improvements per reduced stress) to mental disorders, cardiovascular disease, and metabolic
disruptions. The third pathway is the reduction of environmental risk factors, including air
pollution, noise, and microclimates (mainly heat). Most studies investigated some relationship to
probability of low birth weight (James et al., 2015; Dzhambov et al., 2014). The GS effect on
low birth weight was small, with mixed evidence of GS impact on preterm and very preterm
births.
4-4
-------
4.1.3. Obesity and Physical Activity
Matilda Annerstedt van den Bosch (,Swedish Agricultural University)
J. Aaron Hipp {North Carolina State University)
Evaluations of the benefits of physical activity (PA) have considered both direct impacts
and improved activity associated with GS. The direct health benefits are widespread, from
reduced early mortality to reduced obesity rates to increased cellular antioxidants. For most
endpoints, there is strong causal evidence for the benefits of PA. The exception is urbanized
areas with high levels of noise and air pollution, where PA with increased breathing rate results
in higher pollutant exposure. Research shows mixed results; in one study, Dadvand et al. (2014)
showed a relationship between living close to parks and 60% higher relative prevalence of
asthma in children, but found benefits of greenspace for body weight and sedentary behavior.
Considering potential effects related to physical activity, biking along high-use roads could
potentially reduce the benefits of PA (due to adverse health effects associated with exposure to
traffic-related pollutants). While GS size and proximity correlate with PA, the type of activity is
influenced by the specific user group; the attributes and facilities (such as fields and trails) also
matter, as do communication and programming efforts around GSs (Hunter et al., 2015).
Measurement of PA is varied. Actigraph, an accelerometer device, is widely validated in
epidemiology studies of PA. New tools include mobile devices and crowdsourcing annotations
of outdoor scenes (Adlakha et al., 2014; Hipp et al., 2013b). Results show temperature and
seasonal dependence and suggest that winter monitoring should focus on malls and other large
indoor walking areas. Several policy weaknesses were identified regarding planning and design
of GS, mostly that empirical evidence is rarely used, particularly information on GS quality
(Veal, 2012).
4.1.4. Cardiovascular Disease and Mortality
Perry Hystad (iOregon State University)
Mark Nieuwenhuijsen (CREAL)
Relating mortality to GS is complicated. While mortality is an observable, discrete
endpoint, it has multiple causes and is best measured in terms of early mortality (e.g., years of
life lost). Mortality studies require large population sizes, and results are often expressed as risk
ratios. Most often, GS is represented by its area as percentage of the census area unit (CAU),
while some researchers use NDVI at the CAU or a buffer zone. All-cause mortality reduction
was small (8%) when comparing highest to lowest GS metrics (Gascon et al., 2016). In a natural
experiment, infestation of emerald ash borers was related to increased respiratory and CVD
4-5
-------
mortality, with a greater effect in counties with higher household income: 6.8 additional
respiratory disease deaths per 100,000 adults and 16.7 additional CVD deaths per 100,000
(Donovan et al., 2013). Other composite analyses show GS "greenness" having moderate to
high strength of evidence for many health endpoints, but low to moderate strength for CVD
(Gascon et al., 20162016; Jonker et al., 2014; Takano et al., 2002; Pereira et al., 2012). CVD
studies are often cross-sectional and limited to short-term influences on risk factors (e.g., change
in blood pressure) so they may not reflect long-term influences. Preliminary results from the
Nurses' Health Study for the period 2000-2010 showed no association between NDVI and CVD
(James et al., 2015). GS exposure was most often represented by NDVI.
4.1.5. Neurologic/Neurodevelopmental Effects
Mark Nieuwenhuijsen (CREAL)
Patrick Ryan (University of Cincinnati)
Mental health and behavioral disorders were reviewed in general and for child
development in particular. One condition, unipolar depressive disorders, was the fourth leading
cause of disability-adjusted life years, or disability-adjusted life years (lost) (DALYs) (WHO,
2001). One review report of 4 child studies and 24 adult studies found limited causal evidence
for surrounding greenness and mental health, with one complication of differences in how
exposure assessment was conducted (Gascon et al., 2015). GS can reduce exposures to air
pollution, heat, and noise. The impact of heat reduction on neurobehavior/mental health is
unknown. Noise is associated with neurobehavior and cognition problems in children. GS (trees
and shrubs) can reduce noise by 5-10 decibels. While many air pollutants adversely affect
central nervous system conditions, GS reductions of pollutant concentrations were small and
some GS can exacerbate pollutant-caused conditions (e.g., increased nitrogen dioxide around
street trees in urban canyons where tall buildings affect air circulation or increased pesticide
exposures from their use in GS areas).
The Cincinnati Childhood Allergy and Air Pollution Study of TRAP showed a few
statistically significant GS benefits, but only for areas with high TRAP and only for a few
markers (e.g., hyperactivity at age 7 with GS measured by NDVI). In general, GS improves
neurobehavi oral conditions and mental health by recovery from fatigue; reduction in stress and
crime; and physiological measures, including changes in stress indicators (e.g., salivary Cortisol,
blood pressure). Possible mechanisms include direct effects from viewing and being near GS
and the presumption that GS increases physical activity and social interaction, which contribute
to improved mental health, although evidence remains inconsistent, particularly for physical
activity (Dadvand et al., 2015; Maas et al., 2008; Ord et al., 2013; Lee and Maheswaran, 2010).
4-6
-------
By measuring regional cerebral blood flow, one recent study has shown that GS exposure
reduces rumination, which is indicative of depression, as well as neural activity in an area of the
brain linked to risk for mental illness (Bratman et al., 2015).
4.1.6. Psychosocial Effects
Michelle Kondo (USFS)
Matilda Annerstedt van den Bosch (,Swedish Agricultural University)
Julia Africa (Harvard TH Chan School of Public Health)
An overview of stress biology included the pathways of adaptation, resilience, and
pathophysiology. Highlights of mobile measurement technology include wearable monitoring
devices such as wristbands, chest straps, and headsets that can capture biomarkers of stress and
mood. Results from several studies showed health benefits of various GS elements and
measures. The viewing of recently "greened" vacant lots resulted in decreased heart rate (South
et al., 2015), and a separate study found improved mental health endpoints for those who moved
to greener area compared to those who moved to an area with less urban GS (Alcock et al.,
2014). Experimental exposures to virtual nature paired with natural sounds (although not to
virtual nature alone) resulted in improved physiological stress recovery (Annerstedt et al., 2013).
In some studies, GS was associated with improved social cohesion, noted by increased social
contact and sense of belonging to a community (Kuo et al., 1998; Kweon et al., 1998; Maas
et al., 2009a). GS characteristics most related to improved social cohesion included safety
perceptions, well-maintained GS areas, and GS engagement such as community gardens (Francis
et al., 2012; Hartig et al., 2014). Measurement of GS using only NDVI is clearly not adequate
for measuring how GS affects social cohesion. Exposure to nature has also been found to
increase prosocial behavior (Zhang et al., 2014). Although some results derive from reduced
stress, other causal mechanisms (e.g., those based on social cohesion) are not well understood;
however, there is strong evidence on social isolation having a significant negative health effect
(e.g., decreased cognitive function, CVD) (Boss et al., 2015; Samuel et al., 20142014).
Entrenched institutional disenfranchisement in subpopulations based on race, ethnic group,
education, or economic status could diminish resilience to the adverse impacts of stress. As a
social setting, GS can play a role in supporting adaptive behaviors at both community and
individual levels (Waverijn et al., 2014).
4-7
-------
4.1.7. Attention Restoration/Cognition
J. Aaron Hipp {North Carolina State University)
Laura Jackson (EPA NHEERL)
Four prevailing mechanistic theories were presented related to GS impacts on health that
involve psychological and mental function. Stress reduction theory (Ulrich et al., 1991) was
presented previously at this meeting and suggests that GS exposures result in decreased Cortisol
(Ward Thompson et al., 2012) and blood pressure (Hartig et al., 2003), with concomitant
lessening of other health consequences (e.g., dietary disruptions). Attention restoration theory
hypothesizes that GS is linked to improved attention and cognitive function (Hartig et al., 1996),
including improved academic performance in children (Taylor et al., 2001). One key aspect of
nature appreciation theory is that the presence of GS can lead to general mood improvement as
well as changes to mental health for those who are positively predisposed towards nature. Social
interaction theory says GS impacts are related to interpersonal interactions and mostly related to
reduced stress as measured by improved feelings of well-being, with a postulated link to
increased social cohesion.
4.1.8. Economic and Community Benefits
Michelle Kondo (USFS)
Geoffrey Donovan (USFS)
Other benefits of GS exist in the areas of safety, housing/property values, and energy
expenses. A growing number of studies have established a connection between GS and safety
perception and crime occurrence (Bogar and Beyer, 2015; Kondo et al., 2015), although with
mixed results. Some GS may trigger fear of crime. For example, an open brown area might be
perceived as better for personal safety than lush GS because of fewer and smaller hiding places
for criminals; similar perceptions were found for taller trees as safer than shrubs and short trees
(Fisher and Nasar, 1992; Nasar et al., 1993). GS in a public housing development (Kuo et al.,
1998) and introduced GS on vacant lots (Garvin et al., 2012) were linked with increased sense of
personal safety. In addition, multiple studies have shown GS is associated with fewer
occurrences of crime of multiple types, ranging from narcotics-related crimes (Kondo et al.,
2014) to property crimes (Donovan and Prestemon, 2012; Kuo and Sullivan, 2001) and violent
crimes (Wolfe and Mennis, 2012; Kuo and Sullivan, 2001; Branas et al., 2011). It may be that
maintained GS is having a reverse broken-windows effect, where environments that appear
cared-for signal fewer opportunities for would-be criminals. Much of the positive impact of GS
on inner city populations is tied to community involvement and connectedness (e.g., a
4-8
-------
community-initiated garden could be seen as more beneficial than a city-initiated grassy field)
(Kondo et al., 2015). The economic benefits of GS on neighborhoods are more difficult to
determine. While shade trees can reduce energy consumption for air conditioning units, the cost
savings are much less than the increase in the house's value, indicating that GS perception and
attractiveness might be more important than energy costs. Street trees were found to increase
house value by $7,000, and neighboring houses within 30 m (100 ft) showed value increases
(Anderson and Cordell, 1988; Donovan and Butry, 2010; Donovan and Butry, 2011).
4.1.9. Specific Populations: Health
Richard Mitchell (University of Glasgow)
Patrick Ryan (University of Cincinnati)
Two distinct conditions have motivated the study of whether GS affects specific
populations differently. One is inequality, where different socioeconomic status (SES) or other
demographic groups have different access to GS or receive different exposures to GS. Evidence
so far is mixed regarding gender differences, where some studies show improvement for men
(e.g., CVD) but not for women, while others show opposite gender effects (Richardson and
Mitchell, 2010; Roe et al., 2013; Astell-Burt et al., 2014). The other condition is equigenesis,
where conditions and processes tend to improve health equality. Mental health inequality
(related to income) seemed smallest for GS with highest/easiest access. Escaping to nature
showed more benefit (per mean life satisfaction) for people in financial difficulty than those
living comfortably. It is unknown how the equigenic effect, if real, happens. Perhaps it is
because impacts are more readily apparent for people in poorer health (e.g., by GS access)
compared to those in good health. Impact of GS is likely to vary across age groups and across
specific health endpoints. Understanding age-group susceptibility to each major endpoint is then
critically tied to understanding the mechanisms by which GS affects those endpoints.
4.2. KEY EFFECT CONSIDERATIONS
Except for indirect influences of GS, such as reduced psychological stress, the
information on causal mechanisms relating to GS health effects is generally scant. Of the
various effects studied so far, well-known measurement methods are commonly used for some
(e.g., asthma and CVD). The difficulties arise in the various ways to describe (and measure) the
GS exposure. Childhood asthma is the most well studied respiratory effect in terms of general
etiology regarding certain air pollutants such as ozone and particulate matter (PM), but evidence
of relevant mechanisms for GS is mixed. Some effects (with asthma as a notable example) have
4-9
-------
been shown to reflect toxicological interactions among heat, pollution, and neighborhood
violence (Gupta et al., 2010).
For many endpoints, community or personal interaction with the GS is key to the
beneficial impacts and sometimes-adverse impacts (e.g., pollen exposures among allergic
populations). Some beneficial effects include CVD improvement from increased exercise, which
can be enhanced by the presence and use of GS. Some SES characteristics are variable in terms
of GS impacts (e.g., using GS as a retreat is more beneficial to those in financial difficulty than
those well-off). The health effects studied seem highly influenced by the way GS is measured or
described, as well as by multiple other personal and community characteristics, including the
type and extent of personal engagement with GS and certain genetic vulnerabilities.
4.2.1. Specific Qualities of Respiratory Studies
Ambient air is not fixed to a specific location. Thus, studies across multiple locations are
highly desirable to account for variations of air quality and respiratory disease that are not
necessarily related to GS. Respiratory disease affects large population segments, especially
children with asthma (and parents coping with affected children) and those of all ages with
adverse reactions to allergens. Childhood asthma has been heavily studied because of
environmental regulations on ozone and PM. There is much more information on modes of
action for asthma than for most outcomes, but little of that data mechanistically relates to GS.
Causal pathways linking psychological stress and asthma and pollen to respiratory effects in
allergic people have been targets of multiple studies. These studies help evaluate short- and
long-term exposure to and engagement with GS.
4.2.2. Specific Qualities of Neurological, Psychosocial, and Attention Studies
Neurological, psychosocial, and attention studies are grouped because they share many
features and peculiarities related to effect measures and causal pathways. Many neurological,
psychological, and behavioral measures involve combinations of physical and biochemical
measurements with judgments about what constitutes an abnormal response. Some outcomes
such as those related to mental health status, child neurodevelopment, and child academic
performance are of significant public health importance (as described in studies from the United
States and Europe), so even small improvements could be deemed socially significant. The
subjective nature of the effects characterizations can increase the variability across studies.
Some psychosocial effects are linked to community interactions and can then reflect GS
engagement by the community, such as cooperative planting of gardens. For the mediating
factors or outcomes generally termed psychological stress, biomarkers exist that help evaluate
4-10
-------
small changes, such as blood pressure and Cortisol levels, can help establish causal connections
to even short-term GS engagements.
4.2.3. Qualities of Studies of Other Health Outcomes
The interpretation of GS studies of other health outcomes also can be affected by the
characteristics of the studies. For example, attributing reductions in CVD mortality to GS
exposures may be difficult unless the study population size is quite large. Reductions in obesity
seem mostly related to increased physical activity (frequency or quality); thus, the two aspects
ideally are measured together. Some measurements are not strictly health outcomes but do
contribute to community or personal sense of well-being, potentially affecting health. These can
include property value, perceptions of community safety and property attractiveness, and a sense
of community involvement.
4.3. ASSESSMENT APPROACHES AND MEASURES
The health effects in the GS studies are generally monitored and measured using standard
tools, biomarkers, and protocols. Typical measures and effects include the following:
• Psychological measures: attention restoration, social ties, quality of life, and life
satisfaction through questionnaires or surveys, either standardized or study-specific.
• Physiological stress measures: serum Cortisol and salivary Cortisol as biomarkers that can
be used to measure stress level from allostatic load; also see the further measures listed
below for CVD.
• CVD measures: blood pressure, heart rate, heart rate variability, T-wave amplitude (e.g.,
measured via electrocardiogram applications), myocardial infarction, stroke, and
cardiovascular-related hospitalizations and mortality.
• Neurobehavioral measures: neuroimaging techniques, such as electroencephalogram,
functional magnetic resonance imaging, functional near-infrared spectroscopy, arterial
spin labeling, electrode cap, and cognitive/neurobehavioral/attention-repetition tests.
• Reproductive measures: birth weight (notably small for gestational age and preterm
delivery); maternal stress and health (both a risk factor itself and an influence on other
risk factors).
• Respiratory measures: forced expiratory volume (FEV), forced vital capacity,
asthma-allergy morbidity, childhood asthma emergency room visits.
• Mental health/well-being: general health questionnaire, strengths and difficulties
questionnaire, neurobehavioral AR tests, and emotional and psychosocial tests.
4-11
-------
• Mortality: both all-cause and cause specific (e.g., CVD mortality).
• Other measures and combinations: life expectancy, personal characteristics (e.g., CVD in
women), and short-term influences on other risk factors (e.g., blood pressure).
Recent advances in personal monitoring technology have been used to measure physical
activity, which is an intermediate factor in several health effects (Almanza et al., 2012; Adlakha
et al., 2014). These technologies allow people to measure activity durations and locations.
Similarly, recent advances in automated sensors and information technology are producing large
quantities of different types of data, including from crowdsourced data gathering, often called
"big data." Many of these data are measures of intermediate factors or influences on health
outcomes instead of the more common morbidity data. Perhaps the best example thus far is
physical activity, which is well linked to improvements in cardiovascular function and reductions
in psychological and physiological stress. Physical activity can be one direct indicator of the
extent of personal engagement with the GS. Data can now be compiled from crowdsourced
images (stills and videos) showing where, when, and for how long individuals are engaged in
physical activity. Data from other intermediates similarly tracked by personal monitoring
combined with GPS locations include mood and types of interactions with GS. While
personal-level data are the gold standard for these types of characteristics, the compilation and
integration of such data are currently impractical for most studies. Thus, current data metrics are
typically qualitative and include self-reported exercise, time spent outdoors, and perceptions of
GS access and safety.
4-12
-------
5. DISCUSSION SYNTHESIS AND CONTEXT FOR
CUMULATIVE RISK ASSESSMENT (CRA)
Highlights of the meeting discussions are presented in Sections 5.1 through 5.10, together
with selected citations. These discussion points and suggestions are not intended to (and do not)
represent consensus among participants nor do they reflect official positions of the EPA. An
overview of measures, metrics, and data sources that are used to assess GS exposures, and to a
lesser extent, applications for health research, is presented in Table 5-1.
5.1. ISSUES COMMON TO GREENSPACE (GS) AND CUMULATIVE RISK
ASSESSMENTS
Several issues are common to GS analyses and CRAs (see Table 5-1). Because GS
analyses and the assessment of cumulative risk are both relatively new areas of scientific study,
the basic terminology used in these two areas is evolving and inconsistent across studies and
researchers. The data necessary for a risk assessment to be considered "cumulative" and the
physical features that must be present for a space to be considered a GS have not been
standardized. The inclusion of nonchemical stressors is common to both CRA and GS research.
While vulnerability can be defined differently across disciplines, both CRAs and GS assessments
can examine how best to evaluate the influence of intrinsic factors (e.g., genetics) and extrinsic
factors (e.g., social support groups and health clinics; see DeFur et al., 2007), as well as how to
integrate these sources of vulnerability into an analysis.
Part of the planning phase of any risk assessment is to identify the scope and characterize
the affected population(s); CRA and GS analyses incorporate these two factors in different ways.
The scope of a CRA is often defined to increase the tractability of the multiple factors being
addressed. Simplification can involve placing limits on the number of chemicals, exposure
pathways, or health effects to include. With GS evaluations, the scope of the analysis generally
relates to the physical boundaries (e.g., the definition of the type and boundaries of the GS),
although the list of potentially affected health endpoints in the nearby population is often
considered in the assessment scope. While few GS studies have compared results from different
definitions of assessment scope, many acknowledge the uncertainty (i.e., results may depend on
the chosen scope).
5-1
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications"
Spatial scale and resolution
Greenspace
tools and
measures
Information
resources
(organization)
Metrics
Scale
Resolution
Health
applications
Selected references
Vegetation
coverage
NDVI:
4 I.. ¦
Mean and/or variation in
vegetative cover per unit
mess
Global
30 m—250 m
250 tn
Reproductive,
respiratory, CVD,
all-cause
mortality, sleep
duration, anxiety
and depression,
stress response,
children's mental
health/behavior
Fan et al. (2011)
Dadvand et al. (2012a)
Dadvand et al. (2012b)
Wolfe and Mennis (2012)
Agay-Shay et al. (2014)
Sarkar et al. (2013)
Amoly et al. (2014)
Balseviciene et al. (2014)
Beyer et al. (2014)
Dadvand et al. (2014)
Markevych et al. (2014)
Grazuleviciene et al. (2015)
Lietal. (2015)
Triguero-Mas et al. (2015)
Tree
coverage
Street trees
Also:
MO IMS
1 (fee covet"
Street tree density per
unit; crown area
Also:
LAI:
Percent tree covet"
Municipal
Varies per
application
/, 000 tti
30 m
Respiratory,
allergic
sensitization (to
pollen),
economic
impact: crime
Donovan et al. (2012)
Lovasi et al. (2013)
Jiang et al. (2014a)
Jiang et al. (2014b)
Jiang et al. (2015)
5-2
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
Information
Spatial scale and resolution
tools and
measures
resources
(organization)
Metrics
Scale
Resolution
Health applications
Selected references
GIS: land
cover and
use, GS
coverage
National Land
Cover Database
Land cover diversity:
mean number of land
cover types per unit, land
cover, composition
United
States
30 m
All-cause mortality,
stress response,
mental health and
well-being
Fan et al. (2011)
Mitchell et al. (2011)
White et al. (2013)
CORINE:
(European
Environment
Agency)
Land cover composition
and classification, percent
GS coverage
Europe
25-ha mapping
units
Obesity, mental
health, all-cause
mortality, and more
Annerstedt et al. (2012)
Urban Atlas (as an
example of
country-wide land
use data)
Urban land use, thematic
classification into
mapping units; population
within distance to GS
European
cities
0.25 ha
Population access,
ecosystem services
Annerstedt van den Bosch
et al. (2014a, 2014b)
Larondelle et al. (2014)
Ordnance Survey
Master Map
Surface features, includes
vegetated areas >5 m2
except domestic gardens,
regardless of accessibility
(public and private)
Scotland,
United
Kingdom
5 m2
(1:1,250 map
scale)
CVD, all-cause
mortality
Richardson and Mitchell
(2010)
Generalised Land
Use Database
(GLUD)
Land cover composition,
percent GS coverage
United
Kingdom
1,000-m
low-level
super output
areas
Mental health,
children's mental
health/behavior
Alcock et al. (2014)
Flouri et al. (2014)
5-3
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
Information
Spatial scale and resolution
tools and
measures
resources
(organization)
Metrics
Scale
Resolution
Health applications
Selected references
GIS: land
cover and
use, GS
coverage
(continued,)
CORINE/GLUD
hybrid (CRESH)
Land cover composition,
percent GS coverage
Scotland,
United
Kingdom
Census area
statistics ward;
small area
level
Stress response,
well-being
Mitchell and Popham
(2008)
Richardson and Mitchell
(2010)
Thompson et al. (2012)
Roe et al. (2013)
National Land
Cover
Classification
Database
Land cover composition,
percent GS coverage
Netherlands
25 m
Psychiatric
morbidity, anxiety
and depression
de Vries et al. (2003)
Maas et al. (2009a)
Maas et al. (2009b)
Van den Berg et al. (2010)
Land use
classification
Percent parkland
Australia
Census
collection
districts
Sleep duration,
physical activity
Astell-Burt et al. (2013)
Astell-Burt et al. (2014)
Land Class
Database 11/
Conservation Area
Boundaries/Land
Information New
Zealand hybrid
Land cover, percent GS
per unit
New
Zealand
CAU
Anxiety, mental
health
Nutsford et al. (2013)
Richardson et al. (2013)
5-4
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
tools and
measures
Information
resources
(organization)
Metrics
Spatial scale and resolution
Health applications
Selected references
Scale
Resolution
GIS: land
cover and
use, GS
coverage
(continued)
UTC
Land cover composition
Municipal
1 m
Reproductive effects,
respiratory effects
Nowak et al. (2006)
Donovan et al. (2011)
Lovasi et al. (2013)
High resolution
land cover
Land cover composition
Municipal
60 cm
Economic impacts:
housing prices
Lietal. (2015)
t
cover
JO tn
/
,jj_- m
Gap analysis
I.and co vtv di versi (}' and
cot: ' ft. percent
JO tn
I
hxi t iy
co\
¦i() Iff
Urban
greenspace
Urban GS index
(WHO)
Access to GS (within
300 m) of a 1-ha
minimum size, excludes
private gardens within
housing areas, cemeteries,
buildings
European
region of
WHO
Varies per
application
Physical activity,
stress response
Annerstedt van den Bosch
et al. (2014a)
5-5
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
tools and
measures
Information
resources
(organization)
Metrics
Spatial scale and resolution
Health applications
Selected references
Scale
Resolution
GS
proximity,
distribution
Land use
Proximity to GS,
proximity to GS of a
certain area
Community,
municipal,
regional,
state
Varies per
application
Reproductive effects,
CVD, stress response,
children's mental
health, all-cause
mortality
Dadvand et al. (2012a)
Dadvand et al. (2012b)
Reklaitiene et al. (2014)
Balseviciene et al. (2014)
Dadvand et al. (2014)
Duncan et al. (2014)
Tamosiunas et al. (2014)
Grazuleviciene et al.
(2015)
Triguero-Mas et al. (2015)
OpenStreetMap
Proximity to GS,
proximity to GS of a
certain area
Global
Vector
Reproductive effects
Agay-Shay et al. (2014)
Ecological map of
Barcelona
Proximity to GS,
proximity to GS of a
certain area
Municipal
0.5 m
Anxiety and
depression
Amoly et al. (2014)
Dadvand et al. (2012a)
5-6
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
Information
Spatial scale and resolution
tools and
measures
resources
(organization)
Metrics
Scale
Resolution
Health applications
Selected references
Visible
greenness,
Google Street View
Visible GS from address
or street centerlines
Municipal,
regional
Varies per
application
Physical activity
Brownson et al. (2009)
GS quantity
EnviroAtlas
Percent GS from
walkable street
centerlines and school
and day-care parcel
centroids; percent tree
cover from busy road
edges and 30-m
population estimates;
percent GS within 250 m;
walking distance to
nearest park entrance;
percent GS, tree cover,
wetlands, and water by
census block group and
12-digit HUC
Community,
national
Varies per
application
(1 m—30 m)
Multiple hazard
buffering and health
promotional
ecosystem services
(varies per
application)
Jackson et al. (2013)
Pickard et al. (2015)
Topography,
building footprints,
ArcGIS viewshed
tool
GS viewshed
Spatially
explicit
Varies per
application
Stress response,
allostatic load
South et al. (2015)
Aerial
photography,
Google Earth
Tree counts, crown area
Site-
specific;
national
Varies per
application
Neurobehavioral
effects, stress
response, economic
impacts: housing
prices
Kuo and Sullivan (2001)
Donovan and Butry (2010)
Donovan and Butry (2011)
5-7
-------
Table 5-1. Greenspace (GS) exposure measures and metrics and health applications" (continued)
Greenspace
tools and
measures
Information
resources
(organization)
Metrics
Spatial scale and resolution
Health applications
Selected references
Scale
Resolution
GS quantity,
quality, use
GS audits
Tree counts, tree
measurements (e.g.,
diameter at breast height),
species composition,
people counts and
demographics, activities,
GS attributes
Community,
municipal,
regional,
state
Varies per
application
Physical activity,
anxiety and
depression,
reproductive effects,
economic impacts:
housing prices
Weich et al. (2002)
Araya et al. (2007)
Donovan et al. (2011)
Francis et al. (2012)
Amoly et al. (2014)
Reklaitiene et al. (2014)
Tamosiunas et al. (2014)
Perceptions
of GS, safety,
access,
quality GS
use
GS questionnaires
Perceived safety,
perceived access,
perceived use, perceived
quality, self-reported use
Community,
municipal,
regional,
state
Varies per
application
Physical activity,
perceptions of GS
Amoly et al. (2014)
Reklaitiene et al. (2014)
Sugiyama et al. (2014)
GS
proximity,
use
GPS/GIS
Proximity to GS, routes
through GS, time spent in
GS and/or near GS
features
Individual
Varies per
application
Physical activity
Michael et al. (2010)
Lachowycz et al. (2012)
Adlakha et al. (2014)
Klinker et al. (2014)
GS exposure,
activity levels
Personal monitors;
MapMyRun,
accelerometers
Physical activity levels
in/near GS
Individual
Varies per
application
Physical activity
James et al. (2014)
Klinker et al. (2014)
GS health
effects
Controlled GS
exposure settings
Cognitive performance,
stress measurements
Individual
(classroom),
streets
Not applicable
Neurobehavioral
effects, stress
response
Jiang et al. (2014a)
Jiang et al. (2014b)
Jiang et al. (2015)
aThe entries are presented in the general order of broadest application to more specific.
The measures, metrics, and resolutions not currently linked with a health application are shown in lighter font and italics, to place more attention on those with
existing health applications.
(Note that the references are provided as a group and do not necessarily line up with individual entries that are grouped within the corresponding columns.)
5-8
-------
For CRAs, the population characteristics commonly relate to personal factors affecting
susceptibility to health effects (e.g., per chemical toxicity), while also considering behaviors that
influence exposure pathways, exposure duration and timing of exposures. For GS analyses, the
population could include those residing near or using features of the GS. In GS analyses,
"population characteristics" often relate more to existing health status (e.g., psychosocial stress)
and engagement between the population and the GS. To date, the health effects associated with
GS proximity and exposures are usually characterized as protective of or otherwise beneficial to
health.
One issue common to both CRA and GS evaluations is the reliance on qualitative and
self-reported measures of engagement and behavior. Many GS parameters are indeed
quantitative, such as NDVI estimates of GS area; quantitative health correlates include serum or
salivary Cortisol levels to gauge psychological stress. Many parameters, however, are subjective
or ranked, such as responses to surveys about perception (e.g., of dangers or personal stress), and
the categorization of GS by population use and level of physical activity.
Cumulative risk evaluations that are primarily qualitative are frequently used to set
priorities (e.g., to screen for priority hazards and risks in a community), and some are explicitly
described as not providing a true risk estimate (e.g., the Cal/EPAEPA [2014] EnviroScreen
Tool). One consensus of the work group is that GS evaluations are insufficiently advanced to
allow high confidence in incorporating these analyses into a traditional risk assessment involving
chemical exposure or exposure-response relationships.
Issues common to CRA and GS evaluations include:
• Inconsistent definitions
• Multiple factors: stressors, exposures, population groups, effects
• Spatial: place-based
• Temporal: acute, short-term, chronic, intermittent, continuous
• Population characteristics: socioeconomic, cultural and other susceptibility factors
• Limited information regarding accessibility is available for inclusion into the
assessment: to GS (GS assessments); to health care, public and social services (CRA
assessments)
• Limits of extent, resolution for spatial data sources
• Health effects, exposure-response relationships (knowledge limitations regarding
potential pathways, associations versus causal mechanisms [for GS]), interactions
5-9
-------
• Indices, proxy or surrogate measures
• Generalizability potentially limited by a lack of understanding of underlying
mechanisms
o Limited generalizability of exposure across geographic locations and settings
o Limited ability to extrapolate findings across different populations and subgroups
• Often qualitative (limited quantitative estimates); can use for screening, ranking
5.2. EVOLVING FIELD OF GREENSPACE (GS) ANALYSES
GS analyses may continue to improve in several ways that potentially will make it more
feasible to generalize from the results of a study of one GS and its affected population to another
GS and its population. These improvements will make such assessments more amenable for
incorporation into risk assessments.
The metrics used to measure GS will likely continue to improve in both resolution and
specificity and could make better data available for exposure assessments in more places than
currently exist. Continued evolution is expected toward standard metrics to describe the GS and
assess potential exposures. For example, with chemical-based CRAs, "mixture" is used even for
exposures that are not coincident in time, as long as the chemicals or their effects overlap within
the bodies of the exposed individuals. Metrics are needed to address different components of
GS, characterize previous exposure to or interaction with GS, and describe multiple levels and
critical time windows for exposure in a population.
Although a few examples exist where causal pathways have been suggested to explain
how GS exposure affects a specific health outcome, further progress is expected as additional
studies of the health effects associated with GS exposures are conducted. Additionally, studies
are being conducted that apply both quantitative and qualitative approaches to evaluate the
strength of evidence associated with biological pathways that might link GS exposures with a
health outcome. Such weight-of-evidence frameworks are recommended by scientific panels
(e.g., NRC, 2009; The Presidential/Congressional Commission on Risk Assessment and Risk
Management, 1997). They are used in ecological risk assessments (e.g., Suter and Cormier,
2011), have been proposed for cancer risk assessments of chemicals (U.S. EPA, 2005), and have
been used for evaluating toxicological interactions among mixtures (U.S. EPA, 2000; ATSDR,
2004). Strength- and weight-of-evidence approaches, including meta-analysis, can be used to
synthesize information from different types of studies as well as those of varying quality.
5-10
-------
5.3. APPROACHES TO ADDRESS RANGE OF GREENSPACE (GS) TYPES AND
SPATIAL AND TEMPORAL SCALES
As an emerging field, GS exposure assessment has followed existing GS quantification
methods used in urban planning and forestry. Proximity of the GS to the nearby population has
been shown to depend on the particular metric employed (e.g., linear distance to the border of a
park versus walking distance to the entrance of the park). The bounding and areal measures of
GS have also been expressed with a few common metrics, such as NDVI and skylight detection,
as well as the inclusion of vegetation on personal property and greenery in public areas.
A number of health outcomes have been associated with GS, as described in Chapter 4.
For psychological effects, an important distinction has been made between different types of
positive changes in psychological stress, for example, between recovery from stress and recovery
from attention fatigue, and how these changes may be connected to different health outcomes.
Further studies are needed to better characterize the potential for variations in health
outcomes due to seasonal changes in GS. For example, consider that northern and central
U.S. climates can exhibit dramatic changes in levels of greenness and time spent outdoors
associated with GS depending on the season. Examining the influences of seasonally influenced
changes in foliage levels and exposures to these levels on health outcomes would be a valuable
area for future research.
5.4. AVAILABLE EXPOSURE METRICS: APPLICATIONS AND LIMITATIONS
The geospatial extent, resolution, and type of exposure metric selected for a given study
can influence the conclusions that relate GS to health, as highlighted in Chapter 4. For this
reason, replicate studies are important, as is the inclusion of different exposure metrics within a
single study. This approach can help indicate the influence of a given metric on the results, and,
ultimately, it could help build consensus regarding which metrics are more relevant than others
for a given health outcome. Examining the influence of multiple metrics might also provide
clues about the biological mechanism underlying the relationship between GS and the outcome
studied.
A central challenge for GS exposure assessment is distinguishing measures of proximity
from those of accessibility and access, as described in Chapter 3. Exposure is broadly defined
for GS, so several factors are often included that vary across applications. Exposure magnitude,
duration, and the types of interactions between people and the GS are among the more common
factors included in GS assessments. Without validation studies and sensitivity analyses on
different exposure metrics, it is difficult to determine how robust some conclusions are when
applied to different exposure conditions or scenarios evaluated for the GS assessment and CRA
application.
5-11
-------
5.5. IMPORTANCE OF ENGAGEMENT
Engagement is defined here as the manner and magnitude by which a person or
population interacts with a specific GS. The extent and type of engagement can strongly affect
the influence of GS on health. The scope of reported engagement varies considerably across
studies, from simple proximity to GS, to awareness of GS, to viewing GS through a window, to
walking near or stopping in GS, to actual physical activity in GS. The duration of engagement
has been shown to be an important factor for both the extent of exposure and the duration of the
consequent health impact. The duration of exposure can be quite complex, involving not only
the total time over a relatively long period (e.g., a year or more) but also the frequency and
duration of each separate encounter with GS. The timing of engagement might also be
important—with exposure to GS sought because of, or coinciding with, episodes of
psychological stress or poor health or other types of vulnerability that are potentially more
important at that point (for that individual) than at other times. For example, gardens or views of
GS could have a different effect on individuals in a hospital setting than the same types of GS
elsewhere (Ulrich et al., 1991). Consequently, describing the GS only by its physical
characteristics (size, shape, and location) could be inadequate.
The concentration and composition of a GS, designed or not, strongly influence the
likelihood and frequency of engagement, which can in turn affect health and behavioral
responses. At a basic level, well-placed way-finding features such as marked trails and signs
communicate the intended uses and appropriate navigation of the space. Beyond these features,
visual and other sensory cues within the GS itself are associated with physiological and
psychological effects. A body of literature from both landscape architecture and psychology
supports benefits to psychological restoration of natural organizational structure, such as
Fibonacci sequences of leaf arrangement in plants (Douady and Couder, 1996) and fractal
branching patterns (Kuo and Sullivan, 2001).
Stimuli do not necessarily need to be visual—natural soundscapes, scents (including
phytoncides), and sensory cues found in GSGS can elicit healthful responses; additionally, there
is a line of research examining the cumulative beneficial effects on stress response. In Japan, the
concept of shinrin-yoku, or "forest bathing," refers to the unique sensory immersion of walking
through a forest and has been associated with a number of psychological and physiological
benefits; cultural meanings ascribed to plants or natural features can also function as a
psychological cue (Tsunetsugu et al., 2010). Although the perceptions and values of GS vary
across places and cultures, the design, or absence of design, can impact the species, vegetation
density, and sensory quality of GSs. Design should be considered alongside other contextual
factors, such as maintenance patterns or social stressors in the surrounding area.
5-12
-------
Because of the benefits of GS to mental well-being, another important factor is social
engagement, such as community involvement in design and maintenance of GS (e.g., public
gardens). The degree of public accessibility to GS is also a factor, as described both by physical
obstacles such as fences and gates as well as public perception regarding ownership and access
fees. One study demonstrated that while overall neighborhood vegetation did show direct stress
mitigation, this effect was counteracted by its inhibition of social support, suggesting maintained
park spaces have a stronger positive effect (Fan et al., 2011). Physical activity is another
example of an important type of engagement, whereby activity conducted within GS has shown
more benefit to health than similar activity conducted indoors. Beyond providing a setting for
social gatherings and physical activity, GS features can contribute additional benefits from
engagement with natural attributes (e.g., restorative natural sounds, smells, sights). Research is
needed to better characterize the potential combined effects of exposure to GS through multiple
and potentially overlapping pathways. GS engagement can influence perceptions directly or
indirectly. An example of indirect engagement is the benefit from trees providing shade to
reduce temperature. One such finding near Sacramento, CA is that trees on the southwest side
are perceived as much more beneficial because they provide shade to residences during the hotter
times of the day, even though measurements indicate the energy cost savings are quite small
(Simpson, 2002).
5.6. MAINLY BENEFICIAL EFFECTS
Several studies have shown GS to improve public health through the benefits of increased
exercise and the direct lowering of psychological stress. Other direct benefits of GS for
population health include natural buffers against storm surges, extreme temperatures, and noise
(Jackson et al., 2013). Indirect or secondary benefits of reduced stress and attention restoration
have also been shown, including improved immune function and test performance (Jiang et al.,
2014b). Although the data are presently limited, previously established linkages between
psychological stress and specific health effects (e.g., CVD) suggest that GS presence could also
lower the risk of those health endpoints. In contrast, a few adverse effects from GSs have been
shown too. Among the more obvious is the expected increase in allergic reactions caused by
pollen from trees and grasses (but note that not all GS measures include grass). Some tree
species (e.g., sweet gum) produce VOCs, which can increase ozone, but such contributions to air
pollution likely are minor when compared with anthropogenic sources. Some airflow channeling
by trees, for example those bounding streets, has been shown to reduce the dilution of PM, but
more studies are needed to quantify the change to see whether the increase in air concentration
and health risk is significant. Few studies have examined GS for population subgroups, but
limited findings suggest potential differences in access patterns and health outcomes by gender,
5-13
-------
race, and income (Mitchell et al., 2011, Adlakha et al., 2014), with limited evidence of increased
benefits for more susceptible populations.
5.7. LIMITED QUANTIFICATION OF EXPOSURE-RESPONSE RELATIONSHIPS
Among the strongest consensus conclusions of the work group is that an analog of a
chemical's mathematical dose-response relationship with a health outcome has not yet been
identified for the impacts of GS on health. While some GS-health relationships are supported by
multiple studies (e.g., physical activity correlated with reduced obesity), the mathematical
functions that would allow estimating a "minimum effective GS exposure" are not yet available.
Part of the difficulty with establishing such a function is the lack of consistency in quantifying
both the GS entity and public engagement with GS (i.e., quantifying the GS "exposures"). In
addition, many diseases for which GS might be preventative or mitigating (notably
noncommunicable diseases) have a complex, multifactorial etiology, which makes drawing
causal inference from epidemiological analyses alone very challenging. This lack of quantitative
relationships is a major impetus for using strength-of-evidence approaches to support decisions
about GS and health.
5.8. UNCERTAINTIES ASSOCIATED WITH EXPOSURE AND HEALTH MEASURES
USED IN DIFFERENT GREENSPACE (GS) STUDIES
The uncertainties associated with both exposure measures and various health outcomes
have impeded the progress of this GS research area. A lack of understanding regarding the
mechanism or mechanisms through which GS might affect these health outcomes underlies many
of the uncertainties in the exposure measures.
A consensus definition of GS and GS exposure did not emerge from this meeting.
However, there was consensus that GS is foremost a geographic entity, land that is at least partly
vegetated and located in or adjacent to urban or suburban areas.1 There was also consensus that
GS should be defined by its geographic attributes (location, area, shape), and that human
exposures to GS should be estimated through additional attributes such as composition,
proximity and accessibility to population(s), perceptions of the GS, and the nature of public
engagement with the GS.
A number of quantitative or qualitative measures could be used to characterize most of
these attributes. GS can be characterized by its measure of greenness using average or relative
Several studies have found beneficial effects of wilderness experiences for human health (Cole and Hall, 2010;
Hartig et al., 1991; Sachs and Miller, 1992; Vella et al., 2013); however, this workshop focused specifically on
greenspace and population effects in urban areas.
5-14
-------
NDVI, or by other attributes including proportion of coverage and private versus public
ownership. Similarly, proximity to the public can be measured by linear distance or walking
distance between a residence and a GS. Many individual measures and combinations can be
used to describe and evaluate various attributes of GS and assess "exposure." However, absent
knowledge about the underlying mechanism through which GS affects a given health outcome,
GS is often treated as a composite measure.
Spatial resolution for GS imagery can be a source of uncertainty around GS variation and
specific GS attributes within the unit of analysis. NDVI is a common example of this issue,
where results cannot pinpoint the attribute of GS (e.g., tree versus open areas, species, or design)
that is linked to an effect. Similarly, measures that rely on land use or other nonphysical data
often miss GS attributes that might influence engagement or exposure.
The differences in measures of GS itself are compounded by uncertainties in the type of
GS and actual human engagement. For example, GS size alone might not relate as well to the
potential for physical activity as other characteristics of the GS are considered (e.g., signage and
physical structures, or perceptions of safety that encourage or discourage public access and use).
For example, an attractive and well-defined entry way into GS open to the public likely
encourages visits to and activities within the GS; conversely, the presence of fences or other
indicators that GS is privately owned or off limits generally discourages activities in such areas.
Neighborhood characteristics such as crime incidence also affect attitudes, perceptions, and
ultimately engagement with GS; if an area is perceived as unsafe, public access and activity is
likely to decline. In addition to built environment components, variations in climate, seasonality,
and topography can lead to faulty comparisons among GSGS located in different regions unless
adjustments are made for such factors in the analysis. While the total area of GSGS may be the
same for two locations, the amount accessible for activity may differ substantially, for example,
if one area is steep and another is flat. As described by Wheeler et al. (2015), the type, quality,
and context of GS should be considered in assessing relationships between GS and human health
and well-being.
Many health outcomes evaluated in relation to GS also have multiple descriptors and
different levels of sensitivity. Among the more detailed and precisely defined outcomes are
respiratory function and CVD, which are usually measured using common physiological
measures (e.g., FEVi, or percentage occlusion from vascular disease) or specific medical data
(e.g., International Classification of Diseases [ICD] codes or insurance data). Among the more
vaguely described outcomes is the sense of well-being, which is not affected only by personal
history but also by whether it represents a snapshot of present conditions or the extent of change
from previously undesirable conditions. Although many studies have found a link between GS
5-15
-------
and positive well-being, measures of well-being vary across studies (e.g., self-reported or
independently assessed), which can be a challenge for validation.
Health outcome data can be reported a variety of ways, including at the individual, city,
county, or national level. In addition, some health outcomes are known to depend on age, so if
the age of the study participant is not reported, then the measured effect of the GS on the health
outcome could be biased. Similarly, several health outcomes that have been associated with GS
are potentially exacerbated by heightened psychological stress. Thus, if stress is not also
evaluated when studying those outcomes, then the reported measures of association between the
GS exposure and the outcome might not be independent.
5.9. UNCERTAINTIES ASSOCIATED WITH INNATE DIFFERENCES IN
GREENSPACE ASSESSED IN DIFFERENT STUDIES
Inconsistent definitions can contribute to uncertainty across studies. The innate
differences between places and GSGS, combined with the challenges of available data, make it
difficult to establish a consistent definition of GS exposure that is applicable and comparable
across settings.
Study design, population, and scale are challenges for validating and replicating GS
studies. Randomized control trials can be challenging for GS assessments that by definition
consider GSGS in the physical environment, the effects of which are difficult to isolate among
experimental and control groups. Experimental designs that have examined GS typically involve
a view of GS, either through a window or photograph, and compare results to a control group
with alternate or no views. This has been helpful in identifying effects of short-term exposure
(e.g., occasional use of a park) on short-term outcomes such as acute stress and attention, but it is
less suitable for longer-term exposures (such as daily walks through a GS) and related health
effects. Some of the variation across studies can be addressed by statistical approaches, as long
as certain assumptions about errors and missing data can be appropriately made (e.g., for
comparing or combining cross-sectional with longitudinal studies).
While most GS assessments have similar approaches, there are no standard measures or
approaches for different levels of detail. This limits the ability to cross-validate GS studies
among cities or data sets. GS assessment has several temporal limitations: (1) critical windows
of exposure are unknown for specific population subgroups and outcomes and (2) natural areas,
particularly trees, can grow slowly, making natural-experiment approaches time and
resource-intensive for researchers and evaluators. In a sense, 30-m resolution NDVI has become
a de facto measure of greenness due to its wide availability, and it may enable comparisons of
GS across studies and regions. However, measurements differ across studies, and comparisons
of total greenness across regions can overlook important factors such as climate when assessing
5-16
-------
GS (e.g., it is difficult to compare GS in Arizona to GS in Maine). Furthermore, a 30-m
resolution image cannot capture the different types of GS that may be essential to determining
relevant mechanisms and pathways for health. Recently, research has begun to consider
"relative greenness," which standardizes an NDVI value using the mean of the surrounding area
(James et al., 2015). Studies can also incorporate more detailed GS data, where available, and
compare results with NDVI to begin to address concerns about accuracy and validity of the
exposure data. Meta-analyses of studies using NDVI could be used to determine whether
different measures (e.g., buffer distances) are associated with certain health outcomes, which
could indicate different mechanisms.
Studies also face challenges in comparing measures of accessibility. While objective
measurements of access, such as park entrances or population within a certain distance of GS,
are attainable for most studies, these do not provide information on true patterns of use.
Measurements of perceived access and/or safety can be obtained through questionnaires, but
these are not feasible for many large-scale studies. Moreover, perceptions are not measures of
actual engagement, which must be recorded through observations, self-reporting, or monitoring.
Measuring actual GS engagement is not feasible for many studies, which thus rely on secondary
data.
5.10. FUTURE RESEARCH DIRECTIONS
Considering the evaluation of GS from a risk assessment perspective, uncertainties
around mechanisms through which GS affects human health complicate the interpretation of
many studies and make it difficult to fully inform risk management decisions. Future research
should focus on identifying mechanisms through which GS can affect various health outcomes,
which would help refine GS exposure measures and might identify other health benefits. GS
studies are also needed to integrate more effectively with CRAs and support policy analyses.
Research to date has identified five ways GS is thought to influence health; the strength
of supporting evidence for each of these varies (Hartig et al., 2014):
1. GS can provide opportunities for physical activity. Researchers have identified and
evaluated features of GS that encourage physical activity, and levels of activity have been
measured in and around GS. Importantly, improvements in both self-reported and
physiological health measures have been shown to be greater following physical activity
in GS than that conducted indoors or in highly built outdoor settings (Hartig et al., 2003;
Lee and Lee, 2013; Mitchell, 2013; Shin et al., 2013). Although existing research
appears to support the finding that GS is often used for physical activity among people
who are physically active, this finding is not generalizable. Evidence remains
5-17
-------
inconclusive on whether GS, or specific features of a GS, can themselves encourage
physical activity, or whether the mechanism is more psychological than physiological.
2. Similar to physical activity, research on GS and social cohesion focuses on the
opportunities GS affords for social interaction in gathering spaces (e.g., for picnics,
performances, or other events). Community gardens and greening projects are also part
of this research theme, for which common measures are self-reported social interactions
or sense of community. In the absence of individual data, publicly available data such as
housing tenure or crime rates can be examined to approximate community cohesion or
disorder (Ewart and Suchday, 2002; Miles, 2008). Again, it is unknown where GS fits
into this process and through what mechanisms it acts.
3. Stress reduction and cognitive restoration are perhaps the most studied sociobehavioral
categories, and researchers have used a number of study designs to examine the impact of
GS on mental health, which can in turn influence multiple outcomes (depression/anxiety,
social engagement, CVD). Studies employing experimental designs that compare mental
health outcomes associated with greened versus nongreened environments are relatively
common in the literature, and these studies consistently show linkages between greener
spaces and positive outcomes such as improved attention, reduced physiological stress,
and faster recovery times (Ulrich et al., 1991; Kuo and Sullivan, 2001; Jiang et al., 2015).
Although evidence continues to build in this area, a gap exists in the understanding of
long-term variables concerning both GS exposure and mental health outcomes. In
addition, mental health is arguably entwined with both physical activity and social
cohesion. More research is needed to understand the interplay among environment,
psychology, and behavior related to GS.
4. GSGS can also influence environmental quality as part of both natural and built
environments. Water filtration and storage by GS are well-established benefits, as are the
dissipation of ocean storm energy and phytoremediation of contaminated soils. GS
mitigates the urban heat-island effect; tree cover also provides shade and shelter from UV
exposure. Results of research on air quality and GS have been mixed. Some studies
report improved air quality around GSGS, while others have found evidence of increased
airborne allergens (aeroallergens) or concentrated air pollutants near GS, particularly
street trees, which could be detrimental to populations more susceptible to allergy or
asthma and other chronic respiratory conditions.
5. GS can contribute to biodiversity, which supports ecosystem function and the capacity to
provide hazard buffering and health promotional services to society. The biodiversity
hypothesis posits that biodiversity within GS influences the human microbiome, possibly
contributing to increased immune function and reduced allergies (Hanski et al., 2012;
Rook, 2013; Kuo, 2015). There is evidence of a number of benefits of biodiversity for
human health, from psychological benefits of viewing wildlife to the medicinal,
economic, and cultural value of native species and ecosystems (WHO/CBD, 2015).
Many combinations of GS features and attributes are possible, all of which might act in
some way to influence either the environment or social or psychological processes related to
5-18
-------
health. With an improved understanding of how GS influences each of these basic mechanisms,
the characteristics of GS most relevant to certain health outcomes could be identified. One
strategy common in GS literature is the use of GS measurements at multiple distances, which can
be helpful in identifying the mechanism. To illustrate, trees within 10 m (30 ft) of a school
window might affect the attention of students in a classroom, whereas those at a greater distance
might present opportunities for physical activity and social engagement. Consideration of
measures at multiple distances can help further refine exposure descriptors and focus efforts on
targeting the enhancement of those GS elements (or combinations) that are most likely to benefit
specific populations (and the environment).
Given the many features and functions possible for GS, characterizing its potential risks
and benefits to individual or population health calls for approaches that account for multiple
factors and potential combinations of factors, both within and beyond the GS, that could affect a
given health outcome. Also needed are approaches that capture salient aspects or characteristics
of the individual or population interacting with the GS.
No standard framework exists to evaluate all potential exposures and health effects
related to GS, either alone or in combination. An alternative that is sometimes employed by
regulatory agencies is to define a small subset of possible characteristics and establish standard
protocols for including those in research studies. This approach can help inform reasonable risk
management decisions in the near term while compiling valuable case study lessons to guide
future improvements.
The selection process is illustrated by approaches that have evolved over time for
assessing health risks of chemical mixtures at contaminated sites. Decades ago, the information
considered in estimating whether a site contaminated with multiple chemicals posed an
unacceptable health risk was relatively simple. Such considerations included the similarity of the
critical toxic effect across the chemicals and the duration-averaged environmental concentration
of each that a hypothetical person could be exposed to (e.g., via incidental soil ingestion or
drinking contaminated groundwater as tap water) (U.S. EPA, 1986, 1989). Since that time,
approaches have continued to be refined. Assessments now consider toxicities at exposure levels
higher than that associated with the critical (most sensitive) effect, adjustments to account for
increased vulnerability to certain effects from early life exposures (notably cancer), and more
sophisticated exposure and toxicity groupings that consider modes of toxic action and adverse
outcome pathways (U.S. EPA, 2000; 2005; 2007; 2014). The same kind of evolution is expected
for the methods and measures used to estimate human exposures to (including interactions with)
GS and those used to predict associated health effects.
The explosion of big data—the rapidly increasing volume, variety, and availability of
data and information from structured and unstructured sources including mobile technology,
5-19
-------
sensors, and transaction records—is of increasing interest to researchers and the public. Access
to big data has concerns, which range from issues of privacy to data validity, as well as promise.
First, regarding the data validity, many data sets rely on voluntary inputs so the participants (or
study subjects) are unlikely to be statistically representative, affecting the generalizability of
findings. For example, consider a study that analyzes data from a mobile application designed to
track exercise. These data would be provided only by people already interested in recording
their physical activity. With self-reported data from self-selected participants, it also might be
difficult to establish a true control group. Second, while some well-developed bioinformatics
approaches apply sophisticated statistical methods to yield reasonably reproducible data,
crowdsourced images of GS attributes or human activities in GS analyzed by different people are
likely to include errors caused by lack of precision and other factors, including fatigue
(considering the sheer number of images being evaluated). Much of the crowdsourced
information can be qualitative or subjective, such as self-reported exercise levels (moderate or
low) or perceptions of GS access. Large amounts of these data might not significantly reduce
inter-rater variability; that is, the population variance of self-reported judgments can be quite
high so large amounts of data would improve the estimate of only that single variance. The
advantage of crowdsourced information is that it can be used to capture multiple states and
conditions, including different time frames, seasonal and weather variations, residential versus
commercial differences in GS, and size of the metropolitan area, which can modify how land
uses are designed and managed.
GS studies that assess how GS outcomes can be effectively included in CRAs would also
be useful to promote better policy analyses and integration in this area. Studies of the same
health outcome(s) for GS in different locations can help increase the confidence in generalizing a
relationship observed in previous studies (e.g., between GS and a specific health outcome) to
extend to other GS types, locations, or populations. Alternatively, new studies might find that
results from one GS-health outcome study do not apply to other locations or populations, which
might provide insights into underlying mechanisms.
The gold standard is to develop randomized controlled trials designed to evaluate the
relationship between specific GS exposures and specific outcomes, as well as to examine specific
mechanisms of effect. Conducting experiments in the "real world" can be challenging because
of extensive spatial or temporal requirements for an intervention, and potentially detrimental
effects on study subjects. Consider even a small trial requiring a random assignment of which
locations receive a localized greening treatment and which do not. This type of research
opportunity, with the right partnerships and willing participants, is rare (Kondo et al., 2015).
Alternatively, studies can examine the impacts of projects or naturally occurring events as
natural experiments, and develop measures for evaluating the GS components of the intervention
5-20
-------
or event compared to surrounding areas. These projects can range from local initiatives such as
tree planting or trail development, to large-scale events such as the loss of trees from invasive
species or natural disasters. Such initiating events have already informed environmental and
health impact analyses. Assessment metrics and evaluation strategies for GS benefits could be
incorporated into these efforts also.
A tiered approach to GS assessment could increase comparability across studies. "First
tier" GS assessments can include widely available and comparable measurements like the size
and shape of GS (from satellite imagery or municipal sources), number of entry points to a park,
and public or private ownership information using street network and parcel data where
available. While not measures of access per se, these metrics could be useful to determine the
likelihood of access based on publicly available data. A "second-tier" approach could use
databases such as Google Street View, OpenStreetMap, and crowdsourced data for assessments
of GS quality. Crowdsourced or other big data could be strengthened with quality assurance
analyses in the form of in-person audits or personal monitoring data checks performed on a
subsample. Several validated questionnaires already measure behavioral outcomes such as
perceived safety, stress, and physical activity; large-scale cohort studies and surveys could
incorporate questions pertaining specifically to GS access, perceived access/safety, and time
spent outdoors. GS exposure estimates could then be linked to health and behavior outcomes
from national surveys such data from CDC's (Centers for Disease Control and Prevention)
National Environmental Public Health Tracking Network and the American Community Survey.
5.11. STRENGTH OF EVIDENCE FOR CAUSALITY, IDENTIFYING MAIN
ENDPOINTS
Any assessment involves questions about how to evaluate the evidence, often dealing
with different kinds of evidence, as well as questions about how to assess confidence in that
evidence. One approach for evaluating uncertainties in relationships among stressors, mediators,
and outcomes that has been outlined for CRAs is to apply a structured rating scheme that is
based on evaluating both the weight and the strength of evidence (Suter, 1993). Here, the weight
of evidence referred to the confidence in either the credibility or relevance of the type of
evidence. The term strength of evidence is often used today to reflect overall credibility,
extending from a more focused earlier definition that reflected a measure of the degree (e.g., the
likelihood based on the reported measure).2 A recent study considered the phrase "weight of
2Griffin and Tversky (1992) provide an illustrative example regarding an evaluation of a letter of recommendation
for a student written by a former teacher. The evaluator may consider "two separate aspects of the evidence:
(i) how positive is the letter? and (ii) how credible or knowledgeable is the writer? The first question refers to the
strength or extremeness of the evidence, whereas the second question refers to its weight or credence."
5-21
-------
evidence" too vague and varied in practice, and thus of little scientific use. For this reason,
strength of evidence and evidence integration were preferred (NRC, 2014). Ideally, future
approaches can consider diverse types of evidence in a rigorous, systematic, and transparent
manner that leads to a scientifically defensible conclusion regarding the nature of the relationship
(if any) between an exposure and a health outcome. As the field of GS assessment grows,
structured evidence-driven approaches will be useful for mapping causal pathways between GS
exposure and health outcomes.
Some strength-of-evidence approaches give a higher rating to observed relationships for
which plausible underlying causal pathways or mechanisms have been identified. Observed
relationships that lack evidence for an underlying causal pathway would be rated lower. Another
characteristic given a higher rating is the consistency of results across several studies. Enacting a
rating approach usually requires first evaluating multiple studies to determine where the
uncertainties lie (e.g., with the GS description or with the health outcomes reported) or with the
basic investigative methodology (e.g., where known modifying factors were not adequately taken
into account). The evidence is considered strongest for relationships that are consistently shown
and that can be explained by specific causal pathways. Examples of relatively strong
relationships that would receive a higher rating are reduced psychological stress and improved
reproductive outcomes associated with exposure to GS, for which substantial evidence exists
across a range of study designs. Figure 5-1 shows a potential ranking of the strength of evidence
for GS and different health outcomes based on workshop findings.
5-22
-------
Health effect
Pathways for GS contact
Strength of Evidence
Psychological
Attention Restoration
Cardiovascular and
Mortality
Social Cohesion
Reproductive
Physical Activity
Respiratory
Neurodevelopmental
stress/anxiety reduction; changes in air
quality, temperature
improved cognitive restoration and functions-
recovery from mental fatigue
stress/anxiety reductions-
changes in air quality, temperature
stress/anxiety reduction
stress/anxiety reduction; increased social
contacts; increased physical activity;
changes in air quality, temperature
stress/anxiety reduction
stress/anxiety reduction; changes in
biodiversity; changes in air quality; physical
activity
stress/anxiety reduction; changes in air
quality
Increased
Decreased
1. Figure is for illustrative purposes only and not intended to be comprehensive.
2. For cited resources of benefits and risks of GS for health, see EnviroAtlas' EcoHealth Relationship Browser:
http://enviroatlas.epa.gov/enviroatlas/Tools/EcoHealth RelationshipBrowser/index.html.
Figure 5-1. Strength of evidence for selected health effects.
As research on GS exposures and effects continues to evolve, evidence will be evaluated
in new ways. The insights gained are anticipated to further inform CRA methods for assessing
exposures and effects of nonchemical stressors and mediators. One suggestion for future GS
evaluations is to develop weight- or strength-of-evidence rating structures and apply those to
existing studies for some of the better-understood health effects. For these structured judgments,
key GS terms and exposure scenarios would need to be clearly defined, as would ways to express
variations when the definitions are not consistently followed. Once this type of structure has
been applied, the communications for risk managers could identify those causal pathways that
score high in two areas: (1) they are reasonably well explained by the information available, and
(2) they are consistently demonstrated in multiple studies. An alternative is to identify those
health outcomes for which multiple studies show GS benefits, even (or especially) when the GS
descriptors vary (i.e., GS benefits are robust to the selection of GS measure). The first approach
5-23
-------
is likely to be used for assessing a specific type of GS, public engagement, and health endpoint.
The second approach is more general and can be used to evaluate many types of GS because the
likelihood of some sort of net benefit is high.
5.12. SUMMARY FINDINGS
Five joint findings can be distilled from the workshop discussions:
1. GS effects are mainly beneficial. This contrasts with effects typically assessed in
CRAs, which are mainly harmful.
Current evidence suggests that GS supports public health directly by providing a
dynamic space for exercise, social interactions, and other behaviors that are thought
to lower psychological stress and improve mood. Additional benefits of exposure to
GS appear to include improved immune function, cognition, and attention restoration.
Although data are limited, GS might mitigate or attenuate health outcomes brought on
by psychological stress (e.g., cardiovascular disease). A few adverse effects from GS
exposure also occur- notably respiratory and dermal irritation related to allergens.
2. Both GS assessments and CRAs are spatially dependent.
Both assessments can be conducted at different levels of spatial extent, with
resolutions ranging from rough to highly refined. However, unlike conventional
CRAs, the meaningful attributes of a GS—beyond those associated with objectively
spatial measurements—are not well characterized.
3. Both GS assessments and CRAs strongly depend on the characteristics of the
population, including engagement with GS; scope considerations can differ.
Part of the planning phase of any risk assessment is to identify the scope of the
effect(s) and characterize affected population(s); CRA and GS analyses incorporate
these two factors in different ways. The scope of a CRA is often defined to increase
the tractability of the multiple stressors being addressed. Simplification can involve
placing limits on the number of chemicals, exposure pathways, or health effects to
include. With GS evaluations, the scope of the analysis generally relates to the
physical boundaries (e.g., the definition of the type and boundaries of the GS, or the
amount of GS within a defined buffer), although the set of potential health endpoints
in the nearby population is often considered in the assessment scope. The relative
absence of GS characteristics (e.g., ecological features like biodiversity, landscape
structure, and behavioral prompts like paths and overlooks) from GS assessment is a
5-24
-------
significant shortcoming. Concentrating on specific components of GS versus other
features in the built environment can lead to a false dichotomy between the two. In
reality, green and grey features can be closely integrated, both physically and in the
effects on behavior and population health. GS assessments also exhibit a strong
dependence on the population under consideration, that mirrors the way in which
activity profiles of a population (or individuals) is used when assessing exposures to
chemicals in a CRA.
4 Quantification and qualification of dose-response relationships related to GS
exposure is limited for GS assessments. The same is true for complex chemical
mixtures typically assessed in CRAs.
One of the strongest consensus findings of the work group is that a mathematical
dose-response relationship linking GS exposure with a health outcome(s) does not yet
exist. Uncertainties in the characterization of exposure and causality for GS are
similar to the methodological limitations of environmental chemical exposure
assessment, lacking even the cursory causality information that is available for a
subset of chemicals studied in controlled animal experiments.
5. Both GS assessments and CRAs are relatively new approaches for characterizing
complex environmental exposures. Considerable uncertainty underlies GS exposure
measures used to assess various health outcomes.
Uncertainties remain in the best available methods for quantifying and qualifying GS
exposure as well as characterizing the etiology of various health endpoints,
potentially limiting the usefulness of CRA analysis that incorporate GS. A lack of
understanding regarding the mechanism or mechanisms through which GS might
affect these health outcomes underlies many of the uncertainties in the exposure
measures. Although addressing these uncertainties represents a common research
area between the two fields, we acknowledge that the range of exposures implicated
in salutagenic GS range from botanical bioaerosols, which are easy to sample, to
neurobiological and cultural responses to the view of a specific landscape, which can
be more challenging. Further research and exposure classification is needed, but full
incorporation of all dimensions of GS exposure into a CRA model is unlikely.
5-25
-------
6. CITED REFERENCES AND ADDITIONAL RELEVANT PUBLICATIONS
6.1 CITED REFERENCES
Adlakha, D; Budd, EL; Gernes, R; Sequeira, S; Hipp, JA. (2015) Corrigendum: Use of emerging technologies to
assess differences in outdoor physical activity in St. Louis, Missouri, [first published Front Public Health
2:41], Front Public Health 3:41. doi:10.3389/fbubh.2014.00041.
Agay-Shay, K; Peled, A; Crespo, AV; Peretz, C; Amitai, Y; Linn, S; Friger, M; Nieuwenhuijsen, M. (2014) Green
spaces and adverse pregnancy outcomes. Occup Environ Med 71:562-569.
Alcock, I; White, MP; Wheeler, BW; Fleming, LE; Depledge, MH. (2014) Longitudinal effects on mental health of
moving to greener and less green urban areas. Environ Sci Technol 48(2): 1247-1255.
Almanza, E; Jerrett, M; Dunton, G; Seto, E; Pentz, MA. (2012) A study of community design, greenness, and
physical activity in children using satellite, GPS and accelerometer data. Health Place 18:46-54.
Amoly, E; Dadvand, P; Forns, J; Lopez-Vicente, M; Basagana, X; Julvez, J; Alvarez-Pedrerol, M; Nieuwenhuijsen,
MJ; Sunyer, J. (2014) Green and blue spaces and behavioral development in Barcelona schoolchildren: The
BREATHE Project. Environ Health Perspect 122(12): 1351—1358.
Anderson, L; Cordell, HK. (1988) Influence of trees on residential property values in Athens Georgia (U.S.A.): A
survey based on actual sales prices. Land Urban Plan 15:153-164.
Annerstedt, M; Ostergren, PO; Bjork, J; Grahn, P; Skarback, E; Wahrborg, P. (2012) Green qualities in the
neighbourhood and mental health—results from a longitudinal cohort study in southern Sweden. BMC
Public Health 12:337-349.
Annerstedt, M; Jonsson, P; Wallergard, M; Johansson, G; Karlson, B; Grahn, P; Hansen, AM; Wahrborg, P. (2013)
Inducing physiological stress recovery with sounds of nature in a virtual reality forest—results from a pilot
study. Physiol Behav 118:240-250.
Annerstedt van den Bosch, M; Egorov, A; Mudu, P. (2014a) A GIS-based urban green space indicator—the public
health rationale and results from a case study. Proceedings of ICUH 2014, 11th International Conference on
Urban Health, Manchester UK; March 4-7, 2014.
Annerstedt van den Bosch, M; Egorov, A; Mudu, P. (2014b) Evidence for health benefits of urban green motivates a
green public health indicator—results from a systematic review and GIS-analyses. Int Forest Rev 16(5):23.
Appleton, J. (1975) The Experience of Landscape. New York, NY: Wiley.
Araya, R; Montgomery, A; Rojas, G; Fritsch, R; Solis, J; Signorelli, A; Lewis, G. (2007) Common mental disorders
and the built environment in Santiago, Chile. Br J Psychiatry 190:394-401.
Astell-Burt, T; Feng, X; Kolt, GS. (2013) Does access to neighbourhood green space promote a healthy duration of
sleep? Novel findings from a cross-sectional study of 259,319 Australians. BMJ Open 3:e003094.
doi: 10.1136/bmjopen-2013-003094.
Astell-Burt, T; Mitchell, R; Hartig, T. (2014) The association between green space and mental health varies across
the lifecourse. A longitudinal study. J Epidemiol Community Health 68(6):578-583.
ATSDR (Agency for Toxic Substances and Disease Registry). (2004) Guidance manual for the assessment of joint
toxic action of chemical mixtures. U.S. Department of Health and Human Services, Public Health Service,
ATSDR, Division of Toxicology, Atlanta, GA (May).
http://www.atsdr.cdc.gov/interactionDrofiles/ipga.htnil.
Balseviciene, B; Sinkariova, L; Grazuleviciene, R; Andrusaityte, S; Uzdanaviciute, I; Dedele, A; Nieuwenhuijsen,
MJ. (2014) Impact of residential greenness on preschool children's emotional and behavioral problems. Int
J Environ Res Public Health 11:6757-6770.
Beatley, T. (2012) Exploring the nature pyramid. The Nature of Cities,, Aug. 7, 2012.
http://www.thenatureofcities.eom/2012/08/07/exploring-the-nature-pvramid/ (last accessed October 27,
2015).
Berland, A; Schwartz, K; Herrmann, DL; Hopton, MEME. (2015) How environmental justice patterns are shaped by
place: Terrain and tree canopy in Cincinnati, Ohio, USA 2015. CATE 8(1): Article 1.
littp://digitalcommoiis.lmii.edu/cate/vol8/issl/l.
Beyer, KM; Kaltenbach, A; Szabo, A; Bogar, S; Nieto, FJ; Malecki, KM. (2014) Exposure to neighborhood green
space and mental health: Evidence from the survey of the health of Wisconsin. Int J Env Research and
Public Health ll(3):3453-3472.
Bogar, S; Beyer, KM. (2015) Green space, violence, and crime: a systematic review. Trauma Violence Abuse [Epub
1524838015576412],
6-1
-------
Boss, L; Kang, D; Branson, S. (2015) Loneliness and cognitive function in the older adult: a systematic review.
International Psychogeriatrics 27(4):541-553.
Branas, CC; Cheney, RA; MacDonald, JM; Tam, VW; Jackson, TD; Ten Have, TR. (2011) A difference-in-
differences analysis of health, safety, and greening vacant urban space. Am J Epidemiol
174(11): 1296-1306.
Bratman, GN; Hamilton, JP; Hahn, KS; Daily, GC; Gross, JJ. (2015) Nature experience reduces rumination and
subgenual prefrontal cortex activation. PNAS (0027-8424), 112 (28), p. 8567.
Brown, HS III; Perez, A; Mirchandani, GG; Hoelscher, DM; Kelder, SH. (2008) Crime rates and sedentary behavior
among 4th grade Texas school children. Int J Behav Nutr Phys Act 5(28): 1-7.
Brownson, RC; Hoehner, CM; Day, K; Forsyth, A; Sallis, JF. (2009) Measuring the built environment for physical
activity: State of the science. Am J Preventive Med 36(4):S99-S123.
Cal/EPAEPA (California Environmental Protection Agency). (2014) California communities environmental health
screening tool, Version 2.0 (CalEnviroScreen 2.0). Guidance and screening tool (Oct.). Sacramento, CA:
Office of Environmental Health Hazard Assessment.
http://oehha.ca.gov/ei/pdf/CES20FinalReportUpdateOct2014.pdf. http://oehha.ca.gov/ei/ces2.html.
http://www.calepa.ca.gov/EnvJustice/ (last updated April 14, 2015; last accessed July 15, 2015).
Cohen, DA; Ashwood, JS; Scott, MM; Overton, A; Evenson, KR; Staten, LK; Porter, D; McKenzie, TL; Catellier,
D. (2006) Public parks and physical activity among adolescent girls. Pediatrics 118(5):el381-el389.
Cohen, S; Kessler, RC; Gordon, LU. (1995) Strategies for measuring stress in studies of psychiatric and physical
disorders. Measuring stress: A guide for health and social scientists, (pp. 3-26). New York, NY: Oxford
University Press, Inc.
Cohen, JT; Bellinger, DC; Connor, WE; Shaywitz, BAB A. (2005a) A quantitative analysis of prenatal intake of n-3
polyunsaturated fatty acids and cognitive development. Am J Prev Med 29(4):366-374.
Cohen, JT; Bellinger, DC; Connor, WE; Kris-Etherton, PM; Lawrence, RS; Savitz, DA; Shaywitz, BA; Teutsch,
SM; Gray, GM. (2005b) A quantitative risk-benefit analysis of changes in population fish consumption.
Am J Prev Med 29(4):325-334.
Cole, DN; Hall, TE. (2010) Experiencing the restorative components of wilderness environments: Does congestion
interfere and does length of exposure matter? Environ Behav 42(6):806-823.
Cramer, JS; Browning, W. (2008) Transforming building practices through biophilic design. InKellert, SR;
Heerwagen JH (Eds.) Biophilic Design. New York, NY: Wiley.
Dadvand, P; de Nazelle, A; Figueras, F; Basagana, X; Su, J; Amoly, E; Jerrett, M; Vrijheid, M; Sunyer, J;
Nieuwenhuijsen, MJ. (2012a) Green space, health inequality and pregnancy. Environ Int
40(2012):110-115.
Dadvand, P; Sunyer, J; Basagana, X; Ballester, F; Lertxundi, A; Fernandez-Somoano, A; Estarlich, M; Garcia-
Esteban, R; Mendez, MA; Nieuwenhuijsen, MJ. (2012b) Surrounding greenness and pregnancy outcomes
infour Spanish birth cohorts. Environ Health Perspect 120(10):1481-1487.
Dadvand, P; Villanueva, CM; Font-Ribera, L; Martinez, D; Basagana, X; Belmonte, J; Vrijheid, M; Grazuleviciene,
R; Kogevinas, M; Nieuwenhuijsen, MJ. (2014) Risks and benefits of green spaces for children: A cross-
sectional study of associations with sedentary behavior, obesity, asthma, and allergy. Environ Health
Perspect 122(12): 1329-1335.
Dadvand, P; Nieuwenhuijsen, M; Esnaola, M; Forns, J; Basagana, X; Alvarez-Pedrerol, M; Rivas, I; Lopez-Vicente
M; De Castro Pascual, M; Su, J; Jerrett, M; Querol, X; Sunyer, J. (2015) Green spaces and cognitive
development in primary schoolchildren. PNAS 112:26:7937-7942.
DeFur, PL; Evans, GW; Cohen Hubal, EA. (2007) Vulnerability as a function of individual and group resources in
cumulative risk assessment. Environ Health Perspect 115(5):817-824.
de Vries, S; Verheij, RA; Groenewegen, PP; Spreeuwenberg, PP. (2003) Natural environments - healthy
environments? An exploratory analysis of the relationship between greenspace and health. Environ Plan A
35:1717-1731.
Donovan, GH; Butry, DT. (2010) Trees in the city: Valuing street trees in Portland, Oregon. Land Urban Plan
94:77-83.
Donovan, GH; Butry, DT. (2011) The effect of urban trees on the rental price of single-family homes in Portland,
Oregon. Urban Forestry & Urban Greening 10:163-168.
Donovan, GH; Michael, YL; Butry, DT; Sullivan, AD; Chase, JM. (2011) Urban trees and the risk of poor birth
outcomes. Health Place 17(1):390—393.
Donovan, GH; Michael, YL; Butry, DT; Sullivan, AD; Chase, JM. (2012) Correspondence. Comment on "Green
space, health inequality, and pregnancy." Environ Int 39(1): 133.
6-2
-------
Donovan, GH; Prestemon, JP. (2012) The effect of trees on crime in Portland, Oregon. Environ Behav 44(l):3-30.
Donovan, GH; Butry, DT; Michael, YL; Prestemon, JP; Liebhold, AM; Gatziolis, D; Mao, MY. (2013) The
relationship between trees and human health: Evidence from the spread of the emerald ash borer. Am J
PrevMed 44(2): 139-145.
Douady, S; Couder, Y. (1996) Phyllotaxis as a dynamical self organizing process Part III: The simulation of the
transient regimes of ontogeny. J Theoretical Biol 178(3):295-312.
Duncan, DT; Sharifi, M; Melly, SJ; Marshall, R; Sequist, TD; Rifas-Shiman, SL; Taveras, EM. (2014)
Characteristics of walkable built environments and BMI z-scores in children: evidence from a large
electronic health record database. Environ Health Perspect 122(12): 1359-1365.
Dzhambov, A; Dimitrova, D; Dimitrakova, E. (2014) Association between residential greenness and birth weight:
Systematic review and meta-analysis. Urban For Urban Green 13:621-629.
Ewart, CK; Sue May, S. (2002) Discovering how urban poverty and violence affect health: Development and
validation of a Neighborhood Stress Index. Health Psychol 21(3):254.
Fan, Y; Das, KV; Chen, Q. (2011) Neighborhood green, social support, physical activity, and stress: Assessing the
cumulative impact. Health Place 17(6): 1202-1211.
Fisher, BS; Nasar, JL. (1992) Fear of crime in relation to three exterior site features: prospect, refuge, and escape.
Environment and Behavior 24(l):35-65.
Floyd. MF; Spengler, JO; Maddock, JE; Gobster, PH; Suau, LJ. (2008) Park-based physical activity in diverse
communities of two U.S. cities: An observational study. Am J Prev Med 34(4):299-305.
Flouri, E; Midouhas, E; Joshi, H. (2014) The role of urban neighbourhood green space in children's emotional and
behavioural resilience. J Environ Psychol 40:179-186.
Francis, J; Wood, LJ; Knuiman, M; Giles-Corti, BB. (2012) Quality or quantity? Exploring the relationship between
public open space attributes and mental health in Perth, Western Australia. Soc Sci Med 74:1570-1577.
Garvin, EC; Cannuscio, CC; Branas, CC. (2012) Greening vacant lots to reduce violent crime: A randomised
controlled trial. Inj Prev 19(3):198-203.
Gascon, M; Triguero-Mas, M; Martinez, D; Dadvand, P; Rojas-Rueda, D; Plasencia, A; Nieuwenhuijsen, MJ.
(2016) Residential green spaces and mortality: AA systematic review. Environ Int 86:60-67.
Gascon, M; Triguero-Mas, M; Martinez, D; Dadvand, P; Forns, J; Plasencia, A. (2015) Mental health benefits of
long-term exposure to residential green and blue spaces: A systematic review. Int J Environ Res Public
Health 12:4354-4379.
Grazuleviciene, R; Vencloviene, J; Kubilius, R; Grizas, V; Dedele, A; Grazulevicius, T; Ceponiene, I;
Tamuleviciute-Prasciene, E; Nieuwenhuijsen, MJ; Jones, M; Gidlow, CC. (2015) The effect of park and
urban environments on coronary artery disease patients: A randomized trial. Biomed Research Int Article
ID 403012, doi: http://dx.doi.org/10.1155/2015/403012.
Griffin, D; Tversky, A. (1992) The weighing of evidence and the determinants of confidence. Cogn Psychol
24:411-435.
Grigsby-Toussaint, DS; Turi, KN; Krupa, M; Williams, NJ; Pandi-Perumal, SR; Jean-Louis, G. (2015) Sleep
insufficiency and the natural environment: results from the US Behavioral Risk Factor Surveillance System
Survey. Prev Med 78:78-84.
Gupta, RS; Zhang, X; Springston, EE; Sharp, LK; Curtis, LM; Shalowitz, M; Shannon, JJ; Weiss, KB. (2010) The
association between community crime and childhood asthma prevalence in Chicago. Ann Allergy Asthma
Immunol 104:299-306.
Hanski, I; vonHerzen, L; Fyhrquist, N; Koskinen, K; Torppa, K; Laatikainen, T; Karisola, P; Auvinen, P; Paulin, L;
Makela, MJ; Vartiainen, E; Kosunen, TU; Alenius, H; Haahtela, T. (2012) Environmental biodiversity,
human microbiota, and allergy are interrelated. Proc Natl Acad Sci U. S. A. 109(21):8334-8339.
Hartig, T; Mang, M; Evans, GW. (1991) Restorative effects of natural environment experiences. Environ Behav
23(l):3-26.
Hartig, T; Korpela. K; Evans. G; Garling, T. (1996) Validation of a measure of perceived environmental
restorativeness. Psychol Rep 26(7): 1-64.
Hartig, T; Evans, G; Jamner, L; Davis, DS; Garling, TT. (2003) Tracking restoration in natural and urban field
settings. J Environ Psychol 23(2): 109-123.
Hartig, T; Mitchell, R; de Vries, S; Frumkin, H. (2014) Nature and health. Annual Rev Pub Health 35:207-228.
Heckert, M; Kondo, MC. (Under review) Surveyed uses and perceptions of "cleaned-and-greened" vacant lots.
Heerwagen, JH. (2006) Investing in people: the social benefits of sustainable design. Proceedings, Rethinking
Sustainable Construction '0606, Sarasota, FL, Sept 19-22. http://www.cce.ufl.edu/wp-
content/uploads/2012/08/Heerwagen.pdf.
6-3
-------
Hipp, JA; Adlakha, D; Chockalingam, R. (2013a) Social ecological constraints to park use in communities with
proximate park access. LARNet, Cyber J Appl Leis Rec Res 16(4):23-36.
Hipp, JA; Adlakha, D; Chang, B; Pless, RR. (2013b) Emerging technologies: Webcams and crowd-sourcing to
identify active transportation. Am J Prev Med 44(1) 96-97.
Hunter, RF; Christian, H; Veitch, J; Astell-Burt, T; Hipp, JA; Schipperijn, JJ. (2015) The impact of interventions to
promote physical activity in urban green space: A systematic review and recommendations for future
research. Soc SciMed 124(2015):126-256.
Hystad, P; Davies, HW; Frank, L; Van Loon, J; Gehring, U; Tamburic, L; Brauer, MM. (2014) Residential
greenness and birth outcomes: Evaluating the influence of spatially correlated built-environment factors.
Environ Health Perspect 122(10): 1095-1102.
Jackson, LE; Daniel, J; McCorkle, B; Sears, A; Bush, KFKF. (2013) Linking ecosystem services and human health:
The Eco-Health Relationship Browser. Int J Public Health 58(5): 747-755. doi: 10.1007/s00038-013-0482-
1.
James, P; Berrigan, D; Hart, JE; Hipp, JA; Hoehner, CM; Kerr, J; Major, JM; Oka, M; Laden, FF. (2014) Effects of
buffer size and shape on associations between the built environment and energy balance. Health Place
27:162-170.
James, P; Banay, R; Hart, JE; Laden, F. (2015) A review of the health benefits of greenness. Curr Epidemiol Rep
2(2): 131-142.
Jarabek, A; Crawford-Brown, D. (2014) Decision analysis approach to mode-of-action models: Explicit expression
of data quality and reliability to evaluate causality. Using AOPs for regulatory decisions: Confidence and
criteria adverse outcome pathways: From research to regulation. Poster presented at the Society of
Toxicology Meeting, Washington, DC, September 3-5, 2014.
https://ntP.niehs.nih.gov/iccvam/nieetings/aoD-wksp-2014/presentations/iarabek-508.Ddf.
Jiang, B; Chang, C-Y; Sullivan, WC. (2014a) A dose of nature: Tree cover, stress reduction, and gender differences.
Land Urban Plan 132(Dec.):26-36.
Jiang, B; Li, D; Larsen, L; Sullivan, WC. (2014b) A dose-response curve describing the relationship between urban
tree cover density and self-reported stress recovery. Environ Behav 1-23.
Jiang, B; Larsen, L; Deal, B; Sullivan, WC. (2015) A dose-response curve describing the relationship between tree
cover density and landscape preference. Land Urban Plan 139:16-25.
Jones, A; Hillsdon, M; Coombes, E. (2009) Greenspace access, use, and physical activity: Understanding the effects
of area deprivation. Prev Med 49(6):500-505.
Jonker, MF; van Lenthe, FJ; Donkers, B; MacKenbach, JP; Burdorf, A. (2014) The effect of urban green on small-
area (healthy) life expectancy. J Epidemiol Commun Health 68(10):999-1002.
Kaplan, S; Kaplan, R. (1982) Environment and cognition. New York, NY: Praeger.
Kaplan, R; Kaplan, S. (1989) The experience of nature: A psychological perspective. Cambridge, UK: Cambridge
University Press.
Kaplan, S. (1995) The restorative benefits of nature: Toward an integrative framework. J Environ Psychol
15(3): 169—182.
Kellert SR; Wilson, EO. (1993) The biophilia hypothesis. Washington, DC: Island Press.
Kihal-Talantikite, W; Padilla, CM; Lalloue, B; Gelormini, M; Zmirou-Navier, D; Deguen, S. (2013) Green space,
social inequalities and neonatal mortality inFrance. BMC Preg Child 13:191-200.
Klinker, CD; Schipperijn, J; Christian, H; Kerr, J; Ersboll AK, Troelsen J. (2014) Using accelerometers and global
positioning system devices to assess gender and age differences in children's school, transport, leisure and
home based physical activity. Int J Behav Nutr Phys Act 11:8 doi: 10.1186/1479-5868-11-8.
Kondo, MC; Gross-Davis, CA; May, K; Davis, LO; Johnson, T; Mallard, M; Gabbadon, A; Sherrod, C; Branas, CC.
(2014) Place-based stressors associated with industry and air pollution. Health Place 28:31-37.
Kondo, MC; South, EC; Branas, CC. (2015) Nature-based strategies for improving urban health and safety. J Urban
Health 92(5):800-814.
Kuo, FE; Sullivan, WC; Coley, RL; Branson, L. (1998) Fertile ground for community: Inner-city neighborhood
common spaces. Am J Commun Psychol 26(6):823—851. http://dx.doi.org/10.1023/A: 1022294028903.
Kuo, FE; Sullivan, WC. (2001) Environment and crime in the inner city does vegetation reduce crime? Environ
Behav 33(3):343-367. doi: 10.1177/00139160121973025.
Kuo, M. (2015) How might contact with nature promote human health? Promising mechanisms and a possible
central pathway. Front Psychol Front Psychol 6 (August): Article 1093.
http://dx.doi.org/10.3389/fpsvg.2015.01093.
6-4
-------
Kweon, B-S; Sullivan, W C; Wiley, AR. (1998) Green common spaces and the social integration of inner-city older
adults. Environ Behav 30(6):832-858.
Lachowycz, K; Jones, AP; Page, AS; Wheeler, BW; Cooper, AR. (2012) What can global positioning systems tell us
about the contribution of different types of urban greenspace to children's physical activity? Health Place
18(3):586-594.
Lachowycz, K; Jones, AP. (2013) Towards a better understanding of the relationship between greenspace and
health: Development of a theoretical framework. Land Urban Plan 118:62-69.
Larondelle, N; Haase, D; Kabish, N. (2014) Mapping the diversity of regulating ecosystem services in European
cities. Global Environ Change 26 (2014): 119-129.
Lee, ACK; Maheswaran, R, (2010) The health benefits of urban green spaces: A review of the evidence. J Public
Health 33(2):212-222.
Lee J-Y; Lee, D-C. (2013) Cardiac and pulmonary benefits of forest walking versus city walking in elderly women:
A randomised, controlled, open-label trial. Eur J Integr Med 6(1):5—11.
Li, W; Saphores, J-DM; Gillespie, TW. (2015) A comparison of the economic benefits of urban green spaces
estimated with NDVI and with high-resolution land cover data. Land Urban Plan 133:105-117.
Lovasi, GS; O'Neil-Dunne, JP; Lu, JW; Sheehan, D; Perzanowski, MS; MacFaden, SW; King, KL; Matte, T; Miller,
RL; Hoepner, LA. (2013) Urban tree canopy and asthma, wheeze, rhinitis, and allergic sensitization to tree
pollen in a New York City birth cohort. Environ Health Perspect 121(4):494-500.
Maas, J; Verheij, RA; Spreeuwenberg P; Groenewegen, PP. (2008) Physical activity as a possible mechanism
behind the relationship between greenspace and health: AA multilevel analysis. BMC Public Health
8:206-219.
Maas, J; vanDillen, SME; Verheij, RA; Groenewegen, PP. (2009a) Social contacts as a possible mechanism behind
the relation between green space and health. Health Place 15(2):586-595.
Maas, J; Verheij, RA; de Vries, S; Spreeuwenberg, P; Schellevis, FG; Groenewegen, PP. (2009b) Morbidity is
related to a green living environment. J Epidemiol Community Health 63(12):967-973.
Markevych, I; Tiesler, CMT; Fuertes, E; Romanos, M; Dadvand, P; Nieuwenhuijsen, MJ; Berdel, D; Koletzko, S;
Heinrich, J. (2014) Access to urban green spaces and behavioural problems in children: Results from the
GINIplus and LISAplus studies. Environ Int 71:29-35.
Michael, YL; Perdue. LA; Orwoll, ES; Stefanick ML, Marshall LM, and Osteoporotic Fractures in Men Study
Group. (2010) Physical activity resources and changes in walking in a cohort of older men. Am J Pub
Health 100(4):654-660.
Miles, R. (2008) Neighborhood disorder, perceived safety, and readiness to encourage use of local playgrounds. Am
J Prev Med 34(4):275-281.
Miller, D; Morrice, J; Aspinall, P; Brewer, M; Brown, K; Cummins, R; Dilley, R; Dinnie, L; Donaldson-Selby, G;
Gilbert, A; Hester, A; Harthill, P; Mitchell, R; Morris, S; Pearce, I; Robertson, L; Roe, J; Ward Thompson,
C; Wang, C. (2014) GreenHealth: Contribution of green and open space to public health and wellbeing.
Project No. MW/ECA/UGW/847/08. Final report for: Rural and Environmental Science and Analytical
Services Division, Scottish Government Scotland: James Hutton Institute, OPENSpace Edinburgh
University, University of Glasgow, Heriot-Watt University, Biomathematics and Statistics Scotland.
http://www.hutton.ac.uk/sites/default/files/files/proiects/GreeiiHealth-Final-Report.pdf.
Mitchell, R; Popham, F. (2008) Effect of exposure to natural environment on health inequalities: An observational
population study. Lancet 372(9650): 1655-1660.
Mitchell, R; Astell-Burt, T; Richardson, EA. (2011) A comparison of green space indicators for epidemiological
research. J Epidemiol Community Health 65(10):853-858.
Mitchell, R. (2013) Is physical activity in natural environments better for mental health than physical activity in
other environments? Soc Sci Med 91:130-134.
Mitchell, R; Richardson, EA; Shortt, NK; Pearce, JRJR. (2015) Neighborhood environments and socioeconomic
inequalities in mental well-being. Am J Prev Med 49(l):80-84.
Nasar, JL; Fisher, B; Grannis, M. (1993) Proximate physical cues to fear of crime. Land Urban Plan 26:161-178.
NAS (National Academy of Science) (2009) Science and decisions: advancing risk assessment. Washington, DC:
National Academy of Sciences, http://dels.nas.edu/dels/rpt briefs/IRA brief final.pdf.
National Research Council. (2009) Science and decisions: Advancing risk assessment. Washington, DC: National
Academies Press; 2009.
NRC (National Research Council) (2014) Review of EPA's Integrated Risk Information System (IRIS) Process;
Washington DC: National Academies Press, http://www.nap.edu/catalog/18764/review-of-epas-integrated-
risk-iiiformation-svstem-iris-process.
6-5
-------
Nieuwenhuijsen, MJ; Kruize, H; Gidlow, C; Andrusaityte, S; Anto, JM; Basagana, X; Cirach, M; Dadvand, P;
Danileviciute, A; Donaire-Gonzalez, D; Garcia, J; Jerrett, M; Jones, M; Julvez, J; vanKempen, E; van
Kamp, I; Maas, J; Seto, E; Smith, G; Triguero, M; Wendel-Vos, W; Wright, J; Zufferey, J; van den Hazel,
PJ; Lawrence, R; Grazuleviciene, R. (2014) Positive health effects of the natural outdoor environment in
typical populations in different regions in Europe (PHENOTYPE): A study programme protocol. BMJ
Open4(4):l-llll.
Nowak, DJ; Crane, DE; Stevens, JC. (2006) Air pollution removal by urban trees and shrubs in the United States.
Urban ForestForest Urban Green 4:115-123.
Nutsford, D; Pearson, A; Kingham, S. (2013) An ecological study investigating the association between access to
urban green space and mental health. Public Health 127(11): 1005-1011.
Ord, K; Mitchell, R; Pearce, J. (2013) Is level of neighbourhood green space associated with physical activity in
green space? Int J Behav Nutr Phys Activ 10:127. doi: 10.1186/1479-5868-10-127.
Orians, GH; Heerwagen, JH. (1992) Evolved responses to landscapes. InBarkow, J; Cosmides, L; Tooby, J. The
Adapted Mind: Evolutionary Psychology and the Generation of Culture. New York: Oxford University
Press.
Park, BJ; Tsunetsugu, Y; Kasetani, T; Kagawa, T; Miyazaki, YY (2010) The physiological effects of Shinrin-yoku
(taking in the forest atmosphere or forest bathing): Evidence from field experiments in 24 forests across
Japan. Env Health Prev Med 15(1): 18-26.
Periera, G; Foster, S; Martin, K; Christian, H; Boruff, BJ; Knuiman, K; Giles-Corti, B. (2012) The association
between neighborhood greenness and cardiovascular disease: An observational study. BMC Public Health
12:466-475.
Pickard, BR; Daniel, J; Mehaffey, M; Jackson, LE; Neale, AA. (2015) EnviroAtlas: a new geospatial tool to foster
ecosystem services science and resource management. Eco Serv 14:45-55.
Reklaitiene, R; Grazuleviciene, R; Dedele, A; Baceviciene, M; Luksiene, D; Sapranaviciute-Zabazlajeva, L;
Radisauskas, R; Bernotiene, G; Bobak, M; Nieuwenhuijsen, MJ. (2014) The relationship of green space,
depressive symptoms and perceived general health in urban population. Scand J Public Health 42:669-676.
doi: 10.1177/1403494814544494.
Richardson, EA; Mitchell, R. (2010) Gender differences in relationships between urban green space and health in
the United Kingdom. Soc Sci Med 71(3):568-575.
Richardson, EA; Pearce, J; Mitchell, R. (2013) Role of physical activity in the relationship between urban green
space and health. Public Health 127(4):318-324.
Roe, JJ; Thompson, CW; Aspinall, PA; Brewer, MJ; Duff, EI; Miller, D; Mitchell, R; Clow, A. (2013) Green space
and stress: Evidence from Cortisol measures in deprived urban communities. Int J Environ Res Public
Health 10:4086-4103.
Rook, GA. (2013) Regulation of the immune system by biodiversity from the natural environment: An ecosystem
service essential to health. PNAS 110(4): 18360-18367.
Rosso , AL; Auchincloss, AH; Michael, YL. (2011) The urban built environment and mobility in older adults: A
comprehensive review. J Aging Res 2011:1-10. Article ID 816106 doi:
http://dx.doi.org/10.4061/2011/816106.
Ryan, CO; Browning, WD; Clancy, JO; Andrews, SL; Kallianpurkar, NB. (2014) Biophilic design patterns:
Emerging nature-based parameters for health and well-being in the built environment. ArchNet, Int J
Architectural Res 8(2):62-76.
Ryan, PH; Son, SY; Wolfe, C; Lockey, J; Brokamp, C; LeMasters, G. (2015) A field application of a personal
sensor for ultrafine particle exposure in children. Sci Total Environ 508:366-373. doi:
10.1016/j.scitotenv.2014.11.061. Epub 2014 Dec 10.
Sachs, JJ; Miller, SR. (1992) The impact of a wilderness experience on the social interactions and social
expectations of behaviorally disordered adolescents. Behav Disord 17(2):89-98.
Samuel, LJ; Himmelfarb, CRD; Szklo, M; Seeman, TE; Echeverria, SE; Diez Roux, AV (2014) Social engagement
and chronic disease risk behaviors: the multi-ethnic study of atherosclerosis. Prev Med 71, 61-66.
Sarkar, C; Gallacher, J; Webster, C. (2013) Urban built environment configuration and psychological distress in
older men: Results from the Caerphilly study. BMC Public Health 13:695-706. doi: 10.1186/1471-2458-13-
695.
Shin, Y-K; Kim, DJ; Jung-Choi, K; Son, Y-J; Koo, J-W; Min, J-A; Chae, J-H. (2013) Differences of psychological
effects between meditative and athletic walking in a forest and gymnasium. Scand J Forest Res 28:64-72.
Simpson, JR. (2002) Improved estimates of tree shade effects on residential energy use. Energy Build
34(10): 1067-1076.
6-6
-------
South, EC; Kondo, MC; Cheney, RA; Branas, CC. (2015) Neighborhood blight, stress, and health: A walking trial
of urban greening and ambulatory heart rate. Am J Public Health 105(5):909-913.
Sugiyama, T; Giles-Corti, B; Summers, J; du Toit, L; Leslie, E; Owen, N. (2014) Initiating and maintaining
recreational walking: AA longitudinal study on the influence of neighborhood greenspace. Prev Med
57(3):178-182.
Sullivan, WC; Frumkin, H; Jackson, RJ; Chang C-YY. (2014) Gaia meets Asclepius: Creating healthy places. Land
Urban Plan 127(2014):182-184.
Suminski, RR; Connolly, EK; May, LE; Wasserman, J; Olvera, N; Lee, RE. (2012) Park quality in racial/ethnic
minority neighborhoods. Env Justice 5(6):271-278.
Suter, GWII. (1993) Ecological risk assessment. Boca Raton, FL: Lewis Publishers.
Suter, GW II. (1998) Guidelines for ecological risk assessment. Federal Register 63(93):26846-26924.
Suter, GW II. (1999) Developing conceptual models for complex ecological risk assessments. HERA 5(2):375-396.
Suter, GW II; Cormier, SM. (2011) Why and how to combine evidence in environmental assessments: Weighing
evidence and building cases. Sci Tot Env 409(2011): 1406-1417.
Takano, T; Nakamura, K; Watanabe, M. (2002) Urban residential environments and senior citizens' longevity in
megacity areas: The importance of walkable green spaces. J Epidemiol Comm Health 56:913-918.
Tamosiunas, A; Grazuleviciene, R; Luksiene, D; Dedele, A; Reklaitiene, R; Baceviciene, M; Vencloviene, J;
Bernotiene, G; Radisauskas, R; Malinauskiene, V; Milinaviciene, E; Bobak, M; Peasey, A;
Nieuwenhuijsen, MJ. (2014) Accessibility and use of urban green spaces, and cardiovascular health:
Findings from a Kaunas cohort study. Environ Health 13(1):20. doi:10.1186/1476-069X-13-20.
Taylor, AF; Kuo, FE; Sullivan, WC (2001) Coping with add: the surprising connection to green play settings.
Environ Behav 33(l):54-77. doi: 10.1177/00139160121972864.
The Presidential/Congressional Commission on Risk Assessment and Risk Management. (1997) Risk assessment
and risk management in regulatory decision-making. Washington, D.C: U.S. Government Printing Office;
1997.
Thompson, CW; Roe, J; Aspinall, P; Mitchell, R; Clow, A; Miller, D. (2012) More green space is linked to less
stress in deprived communities: Evidence from salivary Cortisol patterns. Land Urban Plan 105:221-229.
Triguero-Mas, M; Dadvand, P; Cirach, M; Martinez, D; Medina, A; Mompart, A; Basagana, X; Grazuleviciene, R;
Nieuwenhuijsen, MJ. (2015) Natural outdoor environments and mental and physical health: Relationships
and mechanisms. Environ Int 77:35-41.
Tsunetsugu, Y; Park, B-J; Miyazaki, Y. (2010) Trends in research related to "Shinrin-yoku" taking in the forest
atmosphere or forest bathing in Japan. Environ Health Prev Med 15:27-37.
Tzoulas, K; Korpela, K; Venn, S; Yli-Pelkonen, V; Kazmierczak, A; Niemela, J; James, P. (2007) Promoting
ecosystem and human health in urban areas using green infrastructure: A literature review. Land Urban
Plan 81:167-178.
Ulrich, RS; Simons, RF; Losito, BD; Fiorito, E; Miles, MA; Zelson, M. (1991) Stress recovery during exposure to
natural and urban environments. J Environ Psychol, 11:201-230.
U.S. EPA (U.S. Environmental Protection Agency). (1986) Guidelines for the health risk assessment of chemical
mixtures. EPA/630/R-98/002. Washington DC: Risk Assessment Forum, Washington, DC.
http://cfoub.epa. gov/neea/cfm/recordisplav.cfm?deid=22567.
U.S. EPA (U.S. Environmental Protection Agency). (1989) Risk assessment guidance for superfund, Volume 1:
Human health evaluation manual (Part A), interim final (RAGS Part A), EPA/540/1-89/002. Washington,
DC: Office of Emergency and Remedial Response.
http://www.epa.gov/oswer/riskassessment/ragsa/index.htm (last accessed August 7, 2015).
U.S. EPA (U.S. Environmental Protection Agency). (1998) Guidelines for ecological risk assessment. (EPA/630/R-
95/02F). Washington, DC: Risk Assessment Forum, http://www.epa.gov/sites/production/files/2014-
1 l/documents/eco risk assessroentl998.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2000) Supplementary guidance for conducting health risk
assessment of chemical mixtures. EPA/630/R-00/002. Washington, DC: Risk Assessment Forum, Office of
Research and Development, http://cfoub.epa.gov/ncea/cfm/recordisplav.cfm?deid=20533.
U.S. EPA (U.S. Environmental Protection Agency). (2003) Framework for cumulative risk assessment. EPA/600/P-
02/00 IF. Washington, DC: Office of Research and Development, National Center for Environmental
Assessment, http://www.epa.gov/raf/publications/framework-cra.htm.
U.S. EPA (U.S. Environmental Protection Agency). (2005) Guidelines for carcinogen risk assessment. EPA/630/P-
03/00IF. Washington, DC: Risk Assessment Forum, http://www2.epa.gov/osa/guidelines-cafcinogen-risk-
assessment.
6-7
-------
U.S. EPA (U.S. Environmental Protection Agency). (2007) Concepts, methods, and data sources for cumulative
health risk assessment of multiple chemicals, exposures and effects: a resource document. EPA/600/R-
06/013F. Cincinnati, OH: Office of Research and Development, National Center for Environmental
Assessment, http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid= 190187.
U.S. EPA (U.S. Environmental Protection Agency). (2013). Percent of tree cover within 26 m of a road edge [Data
fact sheet]. EnviroAtlas: September, 2013. Available online at
http://enviroatlas.epa.gov/enviroatlas/DataFactSheets/pdf/ESC/Estimatedpercentoftreecoverwithin26mofar
oadedge.pdf (last accessed August 10, 2015).
U.S. EPA (U.S. Environmental Protection Agency). (2014) ORD framework on adverse outcome pathways to be
used by OECD to improve chemical safety assessments. Research Triangle Park, NC: Office of Research
and Development, National Health and Environmental Effects Research.
http://www.epa.gov/nlieerl/articles/2011/Chemical Safety Assessments.html (page last updated April 23,
2014; last accessed August 8, 2015).
U.S. EPA (U.S. Environmental Protection Agency). (2015) Enviro Atlas, http://enviroatlas.epa.gov/enviroatlas/ (last
updated July 23, 2015; last accessed August 8, 2015).
Van den Berg, AE; Maas, J; Verheij, RA; Groenewegen, PP. (2010) Green space as a buffer between stressful life
events and health. Soc Sci Med 70:1203-1210.
Veal, AJ. (2012) FIT for the purpose? Open space planning standards in Britain. J Policy Res Tour Leis Events
4(3):375-379.
Vella, EJ; Milligan, B; Bennett, JL. (2013) Participation in outdoor recreation program predicts improved
psychosocial well-being among veterans with post-traumatic stress disorder: A pilot study. Mil Med
178(3):254-260.
Ward Thompson, C; Roe, J; Aspinall, P; Mitchell, R; Clow, A; Miller, D. (2012) More green space is linked to less
stress in deprived communities: Evidence from salivary Cortisol patterns. Land Urban Plan
105(3):221-229. doi:10.1016/j.landurbplan.20il. 12.015.
Waverijn, G; Wolfe, MK; Mohnen S; Rijken, M; Spreeuwenberg, P; Groenewegen, P (2014) A prospective analysis
of the effect of neighbourhood and individual social capital on changes in self-rated health of people with
chronic illness. BMC Public Health 14:675-686. doi: 10.1186/1471-2458-14-675.
Weich, S; Blanchard, M; Prince, M; Burton, E; Erens, B; Sproston, K. (2002) Mental health and the built
environment: Cross-sectional survey of individual and contextual risk factors for depression. Br J
Psychiatry 180:428-433.
Weier, J; Herring, D. (2011) Measuring vegetation (NDVI & EVI).
http://earthobservatorv.nasa.gov/Features/MeasuringVegetation/ (last accessed August 18, 2015).
Wheeler, BW; Lovell, R; Higgins, SL; Depledge, MH. (2015) Beyond greenspace: An ecological study of
population general health and indicators of natural environment type and quality. Int J Health Geographies
14:1-17.
White, MP; Alcock, I; Wheeler, BW; Depledge, MH. (2013) Would you be happier living in a greener urban area?
A fixed-effects analysis of panel data. Psychol Sci 24:920-928.
Wilson, EO. (1984) Biophilia: The human bond with other species. Cambridge, MA: Harvard University Press.
Wolch, JR; Byrne, J; Newell, JP. (2014) Urban green space, public health, and environmental justice: The challenge
of making cities "just green enough." Land Urban Plan 125:234-244.
Wolfe, MK; Mennis, J. (2012) Does vegetation encourage or suppress urban crime? Evidence from Philadelphia,
PA. Land Urban Plan 108(2-4): 112-122.
WHO (World Health Organization). (2001) The world health report: 2001: Mental Health: New Understanding,
New Hope. Geneva: WHO.
WHO (World Health Organization) and CBD (Convention on Biological Diversity). (2015) Connecting global
priorities: biodiversity and human health: a state of knowledge review. Geneva: WHO.
Zhang, JW; Piff, PK; Iyer, R; Koleva, S; Keltner, D. (2014) An occasion for unselfing: Beautiful nature leads to
prosociality. J Environ Psychol 37:61-72.
6-8
-------
6.2 ADDITIONAL RELEVANT PUBLICATIONS
Africa, J; Logan, A; Mitchell, R; Korpela, K; Allen, D; Tyrvainen, L; Nisbet, E; Li, Q; Tsunetsugu, Y; Miyazaki, Y;
Spengler, J. (2014) The natural environments initiative: Illustrative review and workshop statement. Center
for Health and the Global Environment at the Harvard School of Public Health, Boston, MA.
http://www.chgeharvard.org/sites/default/files/resources/Paper-NaturalEnvironmentsInitiative O.pdf.
Allan, JD; Mclntyre, PB; Smith, SDP; Halpern, BS; Boyer, GL; Buchsbaum, A; Burton Jr, GA; Campbell, LM;
Chadderton, WL; Ciborowski, JJH; Doran, PJ; Eder, T; Infante, DM; Johnson, LB; Joseph, CA; Marino,
AL; Presevich, A; Read, JG; Rose, JB; Rutherford, ES; Sowa, SP; Steinman, AD. (2012) Joint analysis of
stressors and ecosystem services to enhance restoration effectiveness. Proc National Acad Sci
110(l):372-377.
Aspinall. P; Mavros, P; Coyne, R; Roe, JJ. (2013) The urban brain: Analysing outdoor physical activity with mobile
EEG. Br J Sports Med 49(4):272-276.
Berman, MG; Jonides, J; Kaplan, S. (2008) The cognitive benefits of interacting with nature. Psychol Sci
19(12):1207-1212.
Berrigan, D; Hipp, JA; Hurvitz, PM; James, P; Jankowska, MM; Kerr, J; Laden, F; Leonard, T; McKinnon, RA;
Powell-Wiley, TM; Tarlov, E; Zenk, SN. (2015) Geospatial and contextual approaches to energy balance
and health. Annals of GIS 21(2):157-168.
Besenyi, GM; Kaczynski, AT; Stanis, SAW; Bergstrom, RD; Lightner, JS; Hipp, JA. (2014) Planning for health: A
community-based spatial analysis of park availability and chronic disease across the lifespan. Health Place
27:102-105.
Bratman, GN; Hamilton, JP; Daily, GC. (2012) The impacts of nature experience on human cognitive function and
mental health. Ann NY Acad Sci 1249(1): 118-136.
Cal/EPA (California Environmental Protection Agency). (2014) California communities environmental health
screening tool, Version 2.0. (CalEnviroScreen 2.0): Guidance and Screening Tool. Sacramento, CA: Office
of Environmental Health Hazzard Assessment.
http://oehlia.ca.gov/ei/pdf/05S20FinalReportUpdateOct2014.pdr.
Callahan, MA; Sexton, K. (2007) If cumulative risk assessment is the answer, what is the question? Environ Health
Perspect 115(5):799-806.
Caspersen, OH; Olafsson, AS. (2010) Recreational mapping and planning for enlargement of the green structure in
greater Copenhagen. Urban Forest Urban Green 9:101-112.
CDC (Centers for Disease Control and Prevention). (2011) Impact of the built environment on health. Healthy
community design fact sheet series. (June) Atlanta, GA: CDC National Center for Environmental Health,
Division of Emergency and Environmental health Services.
http://www.cdc.gov/nceh/publications/factsheets/impactofthebuiltenvironmentonhealth.pdf (last accessed
July 15, 2015).
Clougherty, JE; Levy, JI; Kubzansky, LD; Ryan, PB; Suglia, SF; Canner, MJ; Wright, RJ. (2007) Synergistic effects
of traffic-related air pollution and exposure to violence on urban asthma etiology. Environ Health Perspect
115(8): 1140—1146.
Comber, A; Brunsdon, C; Green, E. (2008) Using a GIS-based network analysis to determine urban greenspace
accessibility for different ethnic and religious groups. Land Urban Plan 86(1): 103-114.
Coombes, E; Jones, AP; Hillsdon, M. (2010) The relationship of physical activity and overweight to objectively
measured green space accessibility and use. Soc Sci Med 70(6):816-282.
Dige, G. (2011) Green infrastructure and territorial cohesion. The concept of green infrastructure and its integration
into policies using monitoring systems. Copenhagen: European Environment Agency, EEA Technical
Report No. 18/2011, ISSN 1725-2237. http://www.eea.europa.eii/publications/greeii-iiifrastmcture-and-
territorial-cohesion (last accessed July 15, 2015).
EC (European Commission). (2012) The multifunctionality of green infrastructure. Science for Environmental
Policy, DG Environment News Alert Service, In-depth Report (March).
http://ec.eiiropa.eii/environment/nature/ecosvstems/docs/Green Infrastructure.pdf (last accessed July 15,
2015).
EEA/JRC (European Environment Agency/Joint Research Centre). (2013) Environment and Human Health. EEA
Report No 5/2013, Report EUR 25933 EN, Joint Report of the European Environment Agency and
European Commission's Joint Research Centre. ISSN 1725-9177 (printed in Denmark).
http://www.eea.europa.eii/publications/environment-and-human-health (last accessed July 15, 2015).
6-9
-------
EU (European Union). (2012) The VALUE Project: Final report Valuing Attractive Landscapes in the Urban
Economy project funded through the EU Interreg IVB programme for North West Europe.
http://www.academia.edu/2480155/The VALUE Project Final Report September 2012 (last accessed
July 15, 2015).
Gollenberg, AL; Pekow, P; Bertone-Johnson, ER; Freedson, PS; Markenson, G; Chasan-Taber, LL. (2011) Physical
activity and risk of small-for-gestational-age birth among predominantly Puerto Rican women Matern
Child Health J 15:49-59.
GS (Greenspace Scotland). (2008) Health impact assessment of greenspace. A guide. Prepared jointly with Scottish
Natural Heritage and Institute of Occupational Medicine Health Scotland. Stirling, Scotland: Greenspace
Scotland (June). littp://greenspacescotland^orgjiMMtli4mpact-assessmentaspx (last accessed July 15,
2015).
Gronlund, CJ; Berrocal, VJ; White-Newsome, JL; Conlon, KC, O'Neill, MS. (2014) Vulnerability to extreme heat
by socio-demographic characteristics and area green space among the elderly in Michigan, 1990-2007. Env
Res 136: 449-461.
Hagerhall, CM; Ode, A; Tveit, MS; Velarde, MD; Colfer, CJP; Saijala, TT. (2010) Forests, human health and well-
being in light of climate change and urbanisation In (ebook): Forests and society—responding to global
drivers of change (pp. 223-234) Vienna, Austria: IUFRO (International Union of Forest Research
Organizations), http://www.cifor.org/librarv/3170/forests-human-health-and-well-being-in-light-of-climate-
cliange-and-urbaiiisation/ (last accessed July 15, 2015).
Hipp, JA; Chalise, N. (2015) Spatial analysis and correlates of county-level diabetes prevalence, 2009-2010. Prev
Chronic Dis 12:140404. http://www.cdc.gov/pcd/issues/2015/14 0404.htm.
Holtan, MT; Dieterlen, SL; Sullivan, WC. (2014) Social life under cover: Tree canopy and social capital in
Baltimore, Maryland. Environ Behav 46(1): 1-24.
Hunter, MD; Eickhoff, SB; Pheasant, RJ; Douglas, MJ; Watts, GR; Farrow, TFD; Hyland, D; Kang, J; Wilkinson,
ID; Horoshenkov, KV; Woodruff, PWR. (2010) The state of tranquility: Subjective perception is shaped by
contextual modulation of auditory connectivity. Neurolmage 53(2):611-618.
James, P; Troped, PJ; Hart, JE; Joshu, CE; Colditz, GA; Brownson, RC; Ewing, R; Laden, F.F. (2013) Urban
sprawl, physical activity, and body mass index: Nurses' Health Study and Nurses' Health Study II. Am J
Pub Health 103(2):369-375.
Juster, R-P; McEwen, BS; Lupien, SJ. (2009) Allostatic load biomarkers of chronic stress and impact on health and
cognition. Neurosci Biobehav Rev 35(1):2—16.
Kaczynski, AT; Henderson, KA. (2007) Environmental correlates of physical activity: A review of evidence about
parks and recreation. Leisure Sci 29(4):315-354.
Karjalainen, E; Saijala, T; Raitio, H. (2010) Promoting human health through forests: Overview and major
challenges. Environ Health Prev Med 15:1-8.
Kondo, MC; Keene, D; Hohl, BC; MacDonald, JM; Branas, CC. (2015a) A difference-in-differences study of the
effects of a new abandoned building remediation strategy on safety. PloS One 10(7):e0129582.
Kondo, MC; Low, SC; Henning, J; Branas, CC. (2015b) The impact of green stormwater infrastructure installation
on surrounding health and safety. Am J Public Health 105(3):el 14-el21.
La Rosa, D. (2014) Accessibility to greenspaces: GIS based indicators for sustainable planning in a dense urban
context. Ecol Indicators 42:122-134.
Lederbogen, F; Kirsch, P; Haddad, L; Wheeler, BW; Cooper, AR. (2011) City living and urban upbringing affect
neural social stress processing in humans. Nature 474(7352):498-501.
Lewis, AS; Sax, SN; Wason, SC; Campleman, SL. (2011) Non-chemical stressors and cumulative risk assessment:
An overview of current initiatives and potential air pollutant interactions. Int J Environ Res Pub Health
8(6):2020-2073.
Maantay, J. (2002) Mapping environmental injustices: Pitfalls and potential of geographic information systems in
assessing environmental health and equity. Environ Health Perspect 110 (Suppl. 2):161—171.
Mennitt, D; Sherril, K; Fristup, K. (2014) A geospatial model of ambient sound pressure levels in the contiguous
United States. J Acoust Soc Am 135(5):2746-2674.
6-10
-------
NRC (National Research Council) (2013) Urban forestry: Toward an ecosystem services research agenda: A
workshop summary. (K. Thomas and L. Geller, Rapporteurs) Board on Atmospheric Sciences and Climate,
Division on Earth and Life Studies, National Academies Press, Washington, DC.
http://www.naD.edu/download.phD7record id= 18370 (last accessed July 15, 2015).
Park, J; O'Brien, L; Roe, J; Ward Thompson, C; Mitchell, R. (2011) The natural outdoors and health: Assessing the
value and potential contribution of secondary public data sets in the UK to current and future knowledge.
Health Place 17:269-279.
Pretty, JN; Barton, J; Colbeck, I; Hine, R; Mourato, S; MacKerron, G; Wood, CC. (2011) Health values from
ecosystems (Chapter 23) In: The UK national ecosystem assessment: Understanding nature's value to
society: Technical report, (pp. 1153-1182). Cambridge, UK: UNEP-WCMC.
ht)^://www,_cbdiittffiiiaiiaa]/values/uiiitedMn&dom-h£aMLBdf-
Rider, CV; Dourson, ML; Hertzberg, RC; Mumtaz, MM; Price, PS; Simmons, JE. (2012) Incorporating nonchemical
stressors into cumulative risk assessments. Tox Sci 127(1): 10-17.
Rosenthal, JK; Kinney, PL; Metzger, KB. (2015) Intra-urban vulnerability to heat-related mortality in New York
City, 1997-2006. Health Place 30:45-60.
Schacham, E; Hipp, JA; Scheuermann, M; Onen, N; Overton, ETET. (2013) Parks as a tool for HIV management. J
Int Assoc Providers AIDS Care 14(1):8—11.
Sexton, K; Lindner, SH. (2011) Cumulative risk assessment for combined health effects from chemical and
nonchemical stressors. Am J Public Health 101(Suppl. 1):S81-S88.
Thomas, K; Geller, L. (2013) Urban forestry: Toward an ecosystem services research agenda: A workshop
summary. Washington, DC: National Academies Press. http://www.nap.edu/catalog/18370/uiban-forestrv-
toward-an-ecosvstem-services-research-agenda-a-workshop (last accessed May 1, 2015).
Townshend, T; Lake, AA. (2009) Obesogenic urban form: Theory, policy and practice. Health Place 15:909-916.
Ulrich, RS; Simons, RF; Losito, BD; Fiorito, E; Miles, MA; Zelson, MM. (1991) Stress recovery during exposure to
natural and urban environments. J Environ Psychol 11:201-230.
U.S. EPA (U.S. Environmental Protection Agency) (2011) Integrated Risk Information System (IRIS) Glossary.
http://ofmpub.epa.gov/sor intemet/registrv/termreg/searchandretrieve/glossariesandkevwordlists/searcli.do
?details=&glossarvName=IRIS Glossary (Vocabulary Catalog, last update Aug. 31, 2011; last accessed
May 1, 2015).
U.S. EPA (U.S. Environmental Protection Agency) (2014a) Urban environmental program in New England, Region
1. http://www.epa.gov/regionl/ecoAiep/ (last updated May 6, 2014; last accessed May 1, 2015).
U.S. EPA (U.S. Environmental Protection Agency (2014b). EnviroAtlas eco-health relationship browser (access to
literature compilation).
llttp:Z/enviroatlas.epa.gov/enviroatlas/Tools/EcoHealtli RelatioiishipBrowser/introduction.htiM.
6-11
-------
APPENDIX A: PARTICIPANT BIOSKETCHES
A-l
-------
(This page intentionally left blank)
A-2
-------
APPENDIX A:
PARTICIPANT BIOSKETCHES
A.l. JULIA AFRICA
Julia Kane Africa leads the ecological infrastructure, biophilic design, and restorative
landscape areas of the Nature, Health, and the Built Environment program at the Harvard Center
for Health and the Global Environment. In this role, she examines the ways in which nature
(parks and greenspaces) and natural design cues (natural features in built environment settings)
support psychological and physiological health and resilience. The program produced an
illustrative review of greenspace and health found here: www.cheeharvard.org/NEI Paper. She
has completed graduate coursework in environmental health, exposure assessment, and
sustainable design at the Harvard T.H. Chan School of Public Health and the Harvard Graduate
School of Design (MDesS).
A.2. MATILDA ANNERSTEDT VAN DEN BOSCH
Dr. Matilda Annerstedt van den Bosch is a medical doctor working on interdisciplinary
projects to study associations between various natural environments and public health with
epidemiological and experimental methods. Her main focus is health opportunities provided by
greenspaces to various populations, but she also investigates environmental threats like pollen
exposure. She has several publications and is coeditor of the Oxford textbook on Nature and
Public Health. Among other tasks, she collaborates with the World Health Organization to
develop urban health indicators based on geographical and population distribution data. She is
president for the Swedish Society of Behavioural Medicine and directing board member of the
International Society of Doctors for the Environment.
A.3. GLENNON BERESIN
Glennon Beresin is an environmental health fellowship participant with the Association
of Schools and Programs of Public Health (ASPPH), hosted by EPA Office of Research and
Development (ORD) National Center for Environmental Assessment (NCEA) in Cincinnati, OH.
She is working in cumulative risk assessment under mentors Drs. Michael Wright and Glenn
Rice, with a research focus on health impacts of industrial livestock production. Her work is
informed by One Health-oriented environmental health research, which integrates human,
animal, and ecosystem health. Ms. Beresin earned her Master of Science (MS) and her Master of
Public Health (MPH) degrees within the Tufts Friedman School's Agriculture, Food, and
Environment program, and Tufts School of Medicine's Public Health and Professional Degrees
program, respectively.
A-3
-------
A.4. GEOFFREY DONOVAN
Dr. Geoffrey Donovan is an economist with the U.S. Department of Agriculture, Forest
Service, Pacific Northwest Research Station in Portland, OR. He has quantified a wide range of
urban-tree benefits, ranging from intuitive benefits—reduced summertime cooling costs, for
example—to less intuitive benefits such as crime reduction. More recently, he has focused on
the relationship between trees and public health. He found that mothers with trees around their
homes are less likely to have underweight babies, and when trees are killed by an invasive pest,
more people die from cardiovascular and lower respiratory disease. He has a number of ongoing
projects, including a collaboration with the women's health initiative as well as studies using
bio-indicators to quantify human exposure to polycyclic aromatic hydrocarbons and heavy
metals.
A.5. REBECCA GERNES
Rebecca Gernes is an environmental health fellowship participant with the ASPPH,
hosted at EPA ORD NCEA in Cincinnati, OH. Her research focuses on intersections between
the built and natural environment, social and economic development, and human behavior in
relation to health. Ms. Gernes is currently working on incorporating greenspace exposure
assessment into the Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS) as part of
her work on cumulative risk assessment with her mentors Drs. Glenn Rice and Michael Wright.
She has a dual Masters in Public Health and Social Work (MPH, MSW) from the Brown School
at Washington University in St. Louis.
A.6. RICHARD HERTZBERG
Dr. Richard Hertzberg is an adjunct professor of environmental health at Emory
University, special-term appointment at Argonne National Laboratory, and a private consultant.
He retired from EPA ORD NCEA in 2006 after 25 years, mostly leading the research on mixture
risk methods. He is primary author of the EPA 1986 Guidelines for the Health Risk Assessment
of Chemical Mixtures and 2000 Supplementary Guidance for the Health Risk Assessment of
Chemical Mixtures. He has served on cumulative risk groups for the EPA Office of Pesticide
Programs and Risk Assessment Forum, and external advisory groups on mixture risk for Agency
for Toxic Substances and Disease Registry, National Institute for Occupational Safety and
Health, and the Dutch Health Council. His current work includes modeling mixture
dose-response and interaction effects of pesticide combinations. He received a PhD in
biomathematics from the University of Washington and a Bachelor of Science (BS) in
mathematics from Harvey Mudd College.
A-4
-------
A.7. J. AARON HIPP
Dr. J. Aaron Hipp is an associate professor at North Carolina State University, currently
working on a variety of projects investigating the built environment and health behaviors. One
of Dr. Hipp's projects uses public, outdoor, online webcams across the United States to measure
physical activity across built environments including parks, beaches, plazas, and streets. In
addition, he works on several accelerometer and global positioning system (GPS) studies to
better understand where populations engage in physical activity. Dr. Hipp instructs courses in
geographic information system (GIS) and Built Environments and Community Health, and he
serves on the national board of the Open Streets Network of Champions.
A.8. PERRY HYSTAD
Dr. Perry Hystad is an assistant professor within the College of Public Health and Human
Sciences at Oregon State University. He is an environmental epidemiologist focused on
understanding the health impacts related to place (i.e., where we live, work, and play). A large
portion of his research uses spatial exposure assessment methods to determine the chronic health
effects associated with exposure to air pollution, including cardiovascular, respiratory, and
reproductive outcomes. Recently he conducted analyses of residential greenness and adverse
birth outcomes and cardiovascular disease. Given the spatially correlated nature of different
environmental (and social) exposures, he is developing methods to incorporate multiple
exposures related to place into epidemiological analyses.
A.9. LAURA JACKSON
Dr. Laura Jackson is a biologist with the EPA ORD; she is a principal investigator in the
Sustainable and Healthy Communities Research Program. Her work focuses on the hazard
buffering and health promotional aspects of urban ecosystems. Current studies explore linkages
among physical and mental health metrics, near-road tree cover, and neighborhood greenspace.
Past research has explored the landscape ecology of urbanizing areas and the effects of the built
environment on ecological and public health. Dr. Jackson has developed and led studies in
cross-disciplinary research topics and helped to plan and manage environmental research
programs at EPA since 1990.
A.10. MICHELLE KONDO
Dr. Michelle Kondo is a research scientist with the U.S. Forest Service, stationed in
Philadelphia, PA. Dr. Kondo's research investigates the relationship between environments,
public health, and safety. She has conducted multiple community-based air pollution exposure
A-5
-------
assessments. Her recent work evaluates the effects of urban sustainability and stabilization
initiatives, as well as invasive pests, on human health and crime outcomes. Some of her recent
work established a relative reduction in crime (narcotics possession) around green stormwater
infrastructure installations in Philadelphia, and larger and more significant reductions in crimes
surrounding community-initiated greened vacant lots in comparison to city-run
cleaned-and-greened lots in Youngstown, OH. She has also recently published a study which
measured stress-response to greened versus blighted vacant spaces using mobile biosensors. She
has training in civil engineering, urban planning, spatial epidemiology, and environmental
health.
A.11. MARGARET MACDONELL
Dr. Margaret MacDonell is a principal environmental systems engineer in Argonne
National Laboratory's Environmental Science Division and adjunct professor at Northwestern
University. She conducts risk analyses for federal agencies with a focus on cumulative risk
assessment. Margaret was a contributing author to the 2007 EPA NCEA cumulative risk
resource document and is a member of three National Research Council committees addressing
toxicity and exposure guidelines. She has a PhD in civil engineering/environmental health
engineering from Northwestern University, an MS in the same from Notre Dame, and a BS in
biology from Notre Dame.
A.12. YVONNE MICHAEL
Dr. Yvonne Michael is an epidemiologist known for research on multilevel influences on
population health. She has led research projects on the impact of neighborhood environments on
health, the role of psychosocial factors in health, healthy aging, and women's health. She
developed an audit instrument for research evaluating neighborhood walkability (Senior Walking
Environmental Assessment Tool) and has developed modified versions for use with community
members. She is the Associate Dean for Academic and Faculty Affairs and an associate
professor of epidemiology at the Drexel School of Public Health. She completed doctoral
degrees in epidemiology and health and social behavior at Harvard T.H. Chan School of Public
Health and a postdoctoral research fellowship in the epidemiology of aging at Johns Hopkins
Bloomberg School of Public Health.
A.13. TRAVIS MILLER
Travis Miller is the regional planning manager for the Ohio-Kentucky-Indiana (OKI)
Regional Council of Governments with 20 years of land use, economic development, and
environmental planning experience. Travis heads OKI's Water Quality and Greenspace Office
A-6
-------
and is directly involved in the region's Water Quality Management Plan. He has been
instrumental in the launch and continued growth of the region's Taking Root campaign and has
recently led efforts to inform regional stakeholders about solar energy opportunities through
participation in the U.S. Department of Energy SunShot Initiative and Solar Ready project. He
holds an MS in community planning (University of Cincinnati) and a BS in landscape
architecture (The Ohio State University).
A.14. RICHARD MITCHELL
Dr. Richard Mitchell is a professor of health and environment, and head of the Public
Health Group at the Institute for Health and Wellbeing, University of Glasgow. He is also a
codirector of the Centre for Research on Environment, Society, and Health (CRESH,
http://cresh.org.uk). an interdisciplinary and interinstitute center, focused on exploring how
physical and social environments can influence population health, for better and for worse. Dr.
Mitchell is an epidemiologist and geographer. Earlier in his career, he focused on monitoring
and exploring socioeconomic and geographic inequalities in health. Today, his focus is on the
potential for environments, and natural environments in particular, to positively influence
population health and health inequalities.
A.15. MARK NIEUWENHUIJSEN
Dr. Mark Nieuwenhuij sen is an expert in environmental exposure assessment,
epidemiology, and health risk/impact assessment. He has experience and expertise in areas of
respiratory and cardiovascular morbidity and mortality, mental health, cognitive function, cancer
and reproductive health, and exposure measurement and modelling of indoor and outdoor air
pollution, pesticides, greenspace, ultraviolet exposure, chlorination by-products in drinking
water, and heavy metals, using new technology such as GIS, smartphones, and remote sensing.
He leads the European Commission-funded PHENOTYPE (www.phenotype.eu) study,
examining the relations between greenspace and health. He is a coinvestigator in other
programs, notably C1T1SENSE (http://citi-sense.eu/). which aims to empower citizens using
smartphone technology; HELIX (http://www.proiectfa.elix.eu/). which examines the early life
exposome and childhood diseases; EXPOsOMICs (http://www.exposomicsproiect.eu/). which
examines the air pollution and water exposome and health; and PASTA
(http://www.pastaproiect.eu). which promotes active transportation through sustainable transport.
A.16. GLENN RICE
Dr. Glenn Rice has served as an environmental health scientist at EPA NCEA since 1990.
His research interests focus on developing human health risk assessment methods for chemical
A-7
-------
mixtures and cumulative risk scenarios. He is one of the primary authors of the EPA's
Supplementary Guidance for the Health Risk Assessment of Chemical Mixtures and the EPA's
Mercury Study: Report to Congress. He holds a ScD in environmental health and health policy
management from the Harvard School of Public Health, an MS in microbiology from Miami
University, as well as undergraduate degrees in biology and chemistry from Thomas More
College.
A.17. PATRICK RYAN
Dr. Patrick Ryan is an associate professor of pediatrics and environmental health at
Cincinnati Children's Hospital Medical Center and the University of Cincinnati. Dr. Ryan is an
environmental epidemiologist with research interests in the fields of air pollution epidemiology
and exposure assessment. He is the principal investigator on multiple National Institutes of
Health-funded studies of air pollution and respiratory and neurobehavioral development in
childhood, the use of sensor technology to characterize personal exposure to ultrafine particles,
and the impact of traffic-related air pollution at schools. Other research interests include studies
of indoor pollutants and mold, environmental exposure to naturally occurring asbestos, and the
elemental composition of fine particulate matter (PM2.5).
A.18. WILLIAM SULLIVAN
Dr. William Sullivan works to create healthier, more sustainable communities. He is
Professor of Landscape Architecture at the University of Illinois where he, his students, and
collaborators examine the health benefits that come from having regular exposure to urban
landscapes containing green infrastructure. Together, they have found that regular contact with
urban green infrastructure—places with trees, grass, rain gardens, and the like—has profound,
positive impacts for individuals and communities. These urban greenspaces need not be large or
pristine to convey a variety of broad-ranging outcomes. They must, however, be easily
accessible from a person's home or workplace. He is a senior fellow at the National Council for
Science and the Environment and is an active member of the University's Education Justice
Project. Sullivan holds a PhD in Natural Resources with a concentration in Environment and
Behavior from the University of Michigan. (For more about his work, see http://willsull.net.)
A.19. J. MICHAEL WRIGHT
Dr. J. Michael Wright has served as an epidemiologist with EPA NCEA for 14 years. He
has conducted epidemiologic studies on the relationship between waterborne contaminants and
adverse reproductive outcomes and neurodegenerative disorders. In addition, he conducts
exposure assessment research, some of which quantifies the magnitude of bias due to exposure
A-8
-------
misclassification in epidemiologic settings. Dr. Wright has served on several advisory
committees and technical panels on various topics including drinking water quality,
epidemiology, and cumulative risk assessment. He earned his Doctor of Science degree in
Environmental Health from the Environmental Epidemiology Program at the Harvard School of
Public Health.
A-9
-------
(This page intentionally left blank.)
A-10
-------
APPENDIX B: MEETING AGENDA AND TECHNICAL PRESENTATIONS
B-l
-------
{This page intentionally left blank)
B-2
-------
APPENDIX C: WORKING DRAFT GLOSSARY
C-l
-------
(This page intentionally left blank.)
C-2
-------
APPENDIX C:
WORKING DRAFT GLOSSARY
The working draft glossary provided to participants at the meeting as a preliminary draft,
subject to change, is presented in Section C. 1. Sources of the definitions reflected in the glossary
(as indicated by the superscript following the definition) are identified in Section C.2. Other
than greenspace, the individual terms and definitions were not discussed at the meeting. (See
Section 1.2 for the definition of greenspace used in this report.)
C.l. PRELIMINARY GLOSSARY FOR DISCUSSION
Biophysical services: Ecosystem services provided by the physical environment (water, soil, air,
etc.) and the biological activity within it (plants, animals, etc.).1
Built environment: All the physical (human-made) parts of where people live, work, and play
(e.g., homes, buildings, streets, open spaces, and infrastructure)2
Buffer: A factor that reduces risk associated with a stressor(s).3
Cultural ecosystem services: Nonmaterial benefits people obtain from ecosystems, such as
cultural diversity, spiritual and religious values, knowledge systems, educational values,
inspiration, aesthetic values, social relations, sense of place, cultural heritage values,
recreation, and ecotourism.1
Dose-response assessment: A determination of the relationship between the magnitude of an
administered, applied, or internal dose and a specific biological response.4
Dose-response relationship: The relationship between a quantified exposure and the proportion
of subjects demonstrating specific biologically significant changes in incidence and/or in
degree of change (response).4
Disservices: Negative or unintended consequences.1
Ecosystem services: Life-sustaining benefits humans receive from nature, such as clean air and
water, fertile soil, pollination, and flood control.1
Effect measure modification: Occurs when the magnitude of the effect of the primary exposure
on an outcome (i.e., the association) differs depending on the level of a third variable.6
Gray infrastructure: Traditional practices (or systems) for stormwater management and
wastewater treatment, such as pipes and sewers.1
Green infrastructure: A variety of natural elements (trees, grasses, gardens) designed and
landscaped to manage water naturally.1
C-3
-------
Greenspace: Open land partly or completely covered by vegetation.6
Hyperfunctional or hyperfunctionality (referring to systems of managed landscapes,
infrastructure): Because cities can only afford to allocate limited space to infrastructure
and land, each unit needs to be hyperefficient to achieve its goal (e.g., reductions in
pollution, runoff, temperature, etc.).1
Receptor: The individual or population group actually or potentially exposed to a chemical
(receptors can be real or hypothetical). For contaminated sites, various receptors are
typically hypothesized to evaluate potential risks under likely future uses to help guide
risk management decisions. In cases where real people might be incurring exposures
(e.g., including cleanup workers), these should clearly be assessed.4
Response: Response can be expressed as measured or observed incidence or change in level of
response in a population over a specified period of time, or change in level of response,
percentage response in groups of subjects (or populations), or the probability of
occurrence or change in level of response within a population.4
Street tree: Trees located on a strip of land between a roadway and a sidewalk.1
Urban forest: A collection of trees (including any woody plants) that grows within a city, town,
or suburb.1
Urban forestry: The care and management of urban forests.1
Urban heat island: A phenomenon where air temperatures in urban areas are 2-10°F hotter than
surrounding rural areas due to the high concentrations of buildings and pavement in
urban areas.1
Urban metabolism: Quantification of the total resource inputs, outputs, and transformations in a
city stemming from urban socioeconomic activities and regional and global
biogeochemical processes.1
C.2. SOURCES
Note that some of the definitions in Section C. 1 reflect slight refinements from those
indicated in these sources.
1. Thomas, K; Geller, L. (2013) Urban forestry: Toward an ecosystem services research
agenda: A workshop summary. Washington, DC: National Academies Press.
http://www.nap.edu/catalog/18370/urban-forestrv-toward-an-ecosvstem.-services-
research-agenda-a-workshop (last accessed May 1, 2015).
C-4
-------
2. CDC (Centers for Disease Control and Prevention). (2011) Impact of the built
environment on health. Healthy community design fact sheet series. (June) Atlanta, GA:
CDC National Center for Environmental Health, Division of Emergency and
Environmental Health Services (June)
http://www.cdc.eov/nceh/publications/factsheets/impactofthebuiltenvironmentonhealth.p
df (last accessed July 15, 2015).
3. Sexton, K; Lindner, SH. (2011) Cumulative risk assessment for combined health effects
from chemical and nonchemical stressors. Am J Public Health 101(Suppl. 1):S81—S88.
4. U.S. EPA (U.S. Environmental Protection Agency). (2011) Integrated Risk Information
System (IRIS) Glossary.
http://ofmpub.epa.gov/sor interoet/registrv/temireg/searchandretrieve/giossariesandkeyw
ordlists/search.do?details=&glossaryName=IRIS Glossary (Vocabulary Catalog, last
update Aug. 31, 2011; last accessed May 1, 2015).
5. BUSPH (Boston University School of Public Health). (Undated) Confounding and effect
measure modification (module used for both BS704 and EP713);
http://www.cdc.gOv/nceh/publications/factsheets/impactofthebuiltenvironmentonhealth.p
df (last accessed October 28. 2015).
6. U.S. EPA (U.S. Environmental Protection Agency). (2014) Urban environmental
program in New England, Region 1. http://www.epa.gov/regionl/eco/uep/ (last updated
May 6, 2014; last accessed May 1, 2015).
C-5
------- |