Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 5
2
Physical Activity and the
Built Environment
Key Points Noted in Presentations
• Physical activity is encouraged or discouraged by characteristics
of the physical and built environment (such as walkability or
the availability and condition of parks and recreational spaces),
none of which are under the control of the public health sector.
• Researchers use surveys, observations or audits, geographic in-
formation system (GIS)-based data, policy measures, and crime
data to assess physical activity environments and policies, but
a surveillance system that provides accurate, reliable, and com-
plete data currently does not exist.
• Non-health sectors need to be engaged in the collection of these
data because data related to the built environment are not rou-
tinely collected by the health sector.
• Clear national objectives need to be established for the collection
and use of data related to physical activity environments.
The reasons people become overweight or obese are multifaceted, but
the story begins with a mismatch between calories consumed and the body’s
capacity to burn that energy. Many factors influence the rate at which an
individual burns calories and the number of calories he or she consumes
5
OCR for page 6
6 MEASURING PROGRESS IN OBESITY PREVENTION
and burns in a day, but the environments in which people live can have a
profound impact on the amount of physical activity in which they engage.
Public policy plays a role: declines in the rates at which children walk to
school and adults use public transportation, for example, have coincided
with the obesity epidemic, and such trends in part reflect changes in zoning
and land use, funding for public transportation, and other policies. James
Sallis, professor of psychology at San Diego State University, and Christine
Hoehner, assistant professor in the Division of Public Health Sciences at
Washington University in St. Louis, discussed the environmental and policy
factors that likely affect physical activity levels, especially in young people;
the pros and cons of existing ways of measuring the effects of the environ-
ment; and possible ways to improve measurement in this area.
ENVIRONMENTAL AND POLICY INFLUENCES
ON PHYSICAL ACTIVITY
Presenter: James F. Sallis
An ecological model of health behavior, Sallis explained, is a way of
taking into account the impacts of society and culture, the physical envi-
ronment, and public policy on the behavior of individuals, who are also
influenced by biological and psychological factors and their own skills and
knowledge. Figure 2-1 illustrates the complexity of the ways in which the
environment affects levels of physical activity. The shaded circle represents
the four domains in which people can be active: at home, at work or
school, during recreation, and in moving from place to place. The other
circles depict the many factors that influence how active people are in each
domain. Thus, for example, the upper right quadrant shows how people’s
level of activity while commuting, doing errands, and making other trips
depends on their own characteristics; their perceptions of how convenient
and accessible different modes of transport might be; the characteristics
of the immediate environment (e.g., paths for biking and walking, traffic);
and policies such as zoning codes, traffic management, and investments in
public transportation.
Settings where it is possible to walk and bike to everyday destinations
and to engage in outdoor recreation (such as in parks and playground) sup-
port physical activity. What these settings have in common, Sallis noted,
is that none of them are under the control of the public health sector. He
explained that policies, whether formal or informal, issued by government
or the private sector, can affect physical activity in four ways. First, zoning
and building codes and the design of transportation and recreation facili-
ties all affect the built environment. Second, policies affect programs, such
as physical education requirements in schools and sports programs and
OCR for page 7
Policy Environment
Behavior Settings:
Access & Characteristics
Zoning codes
Development
Health care
regulations
policies/
Transport
incentives Behavior: Active Living Domains
investments
Neighborhood
Zoning codes
Traffic demand
Walkability
Development
management
Neighborhood Ped/bike facilities
regulations
Parking
Ped/bike facilities Parking
Transport
regulations
Aesthetics Transit
Perceived Environment Active
investments & Active Developer
Traffic safety Traffic
regulations Transport
Recreation incentives
Safety Accessibility
Public
recreation
Info during transport
Intrapersonal
investments
Safety signage
Park policies
Demographics
Attractiveness Radio ads & news
Billboards
Biological
Recreation Environment
Psychological
Home PA equipment
Convenience
Family Situation
Parks, trails, programs
Workplace Environment
Private recreation
Comfort Neighborhood walkability
facilities
Parking
Community organizations
Zoning codes
Transit access
Sports - amateur, pro
Fire codes
Trail access
Sedentary options
Building codes
Perceived crime Building design
Occupational
Household Parking regulations
Stair design
Activities
Activities Transportation
PA facilities & programs
Home Environment investments
Health care policies
PA equipment
Interpersonal School Environment
Gardens
modeling, Neighborhood walkability
Stairs
social support,
Subsidized equipment Ped/bike facilities
Electronic entertainment
partners for social
Health care policies Facilities
Labor-saving devices
activities
Zoning codes PE program
School siting policies
Home prices Walk to School
Health care - counseling,
PE policies & funding
Housing-jobs balance program
Weather
info
Facility access policies
Topography
Social climate,
Mass media - news, ads
Facilities budgets
Open space
safety, crime, clubs, teams,
Sports
Safe Routes to School funding
Air quality
programs, norms, culture,
Informal discussions
social capital
Media regulations
Transport policies
Health sector policies Advocacy by
Land use policies
Business practices individuals &
organizations
Information Natural
Environment Environment
Social Cultural
Environment
FIGURE 2-1 Ecological model of four domains of active living.
NOTES: PA = physical activity; PE = physical education; ped = pedestrian.
Figure 2-1.eps
SOURCE: Sallis et al., 2006. Annual Review of Public Health Copyright 2005 by ANNUAL REVIEWS, INC. Reproduced with
7
permission of ANNUAL REVIEWS, INC. in the format Other book via Copyright Clearance Center.
OCR for page 8
8 MEASURING PROGRESS IN OBESITY PREVENTION
leagues in parks and recreation centers. Policies can also offer incentives,
such as insurance discounts, subsidies for parking or commuting by bicycle,
or cash in lieu of parking subsidies for workers who commute without cars.
Finally, whatever the approach, secure funding for policies and programs
is a critical element.
Research
Researchers have examined the relationships among these factors and
physical activity systematically, and Sallis summarized their findings on
attributes of the built environment (Sallis and Kerr, 2006). One example
is walkability. Numerous studies have shown that when it is easy to walk
to school, work, and local businesses, people do walk, as well as ride
bicycles, more. A more modest number of studies have shown that people
are more likely to walk where there are sidewalks, although the relation-
ship here is less consistent. People who live near parks, private health
clubs, playgrounds, and other recreational facilities engage in more physi-
cal activity, researchers have found, but the aesthetic characteristics of the
facilities make a difference. Thus, not surprisingly, a park that is run down
and not kept clean and safe is less likely to attract those who live nearby
to engage in physical activity. In answer to a question, Sallis noted that a
small amount of research has begun to explore the associations between
the built environment and levels of fitness, although one would expect the
higher activity levels in supportive environments to lead to increased fitness.
Sallis presented preliminary results from a review he and colleagues
were conducting of several hundred studies of the specific associations
between attributes of the built environment and physical activity levels in
youth (Figures 2-2a and 2-2b). The results are in line with the general find-
ings from studies of adults. However, studies of youth appear to show less
consistent associations between neighborhood environments and physical
activity. The bars represent the percentages of studies that demonstrate
the associations one would expect for particular features, and they point
to features that might be most rewarding from a policy perspective, such
as walkability. For youth, the most consistent associations are with mixed
land use (a component of walkability) and access to parks and recreational
facilities. Findings differed according to whether environmental attributes
and physical activity were measured objectively or by self-report.
For reducing childhood obesity, Sallis added, characteristics of schools
are key. He noted that the quantity and quality of physical education time,
recess, classroom breaks, after-school programs and joint-use agreements
that allow community access after school hours, the nature of the school
grounds, and the distance of schools from students’ homes all are related
to activity levels. Measuring the degree to which these features influence
OCR for page 9
9
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
54 Associations in
Vegetation 36
expected
29
Crime directions (%)
14
42
Pedestrian safety structure 16
50
Traffic speed/volume
*
50
Walking/biking facilities 22
100
Walkability 14
45
Street connectivity 21
60
Residential density 33
66
Land-use mix 50
54
Recreational facilities 40
Reported PA
49
Parks 49 Objectively measured PA
FIGURE 2-2a Associations between objectively measured attributes of the built
2-2a.eps
environment and physical activity among youth.
NOTE: Data presented were preliminary; final results were separated by children
aged 3-12 and adolescents aged 13-18 and are presented in Ding and colleagues
(2011). PA = physical activity. * Data not reported for objectively measured PA for
this attribute.
SOURCE: Sallis, 2011.
21
Vegetation Associations in
*
expected
52
Incivilities/Disorders
* directions (%)
19
Unspecified safety 11
19
Crime 8
40
General traffic safety
*
83
Pedestrian safety structures
*
34
Traffic speed/volume 0
42
Walking/biking facilities 25
37
Street connectivity
*
36
Residential density *
23
Land-use mix 13
51
Home equipment 56
31
Recreational facilities Reported PA
18
24
Parks Objectively measured PA
19
FIGURE 2-2b Associations between perceived attributes of the built environment
and physical activity among youth. 2-2b.eps
NOTE: Data presented were preliminary; final results were separated by children
aged 3-12 and adolescents aged 13-18 and are presented in Ding and colleagues
(2011). PA = physical activity. * Data not reported for objectively measured PA for
this attribute.
SOURCE: Sallis, 2011.
OCR for page 10
10 MEASURING PROGRESS IN OBESITY PREVENTION
physical activity would require another, more complicated, type of analysis,
he noted in response to a question. He concluded from these preliminary
results that there is strong support for the value of a few features, whereas
others have not been extensively studied.
Sallis summarized the role of policy in Table 2-1. The table shows the
relationships among attributes that can influence physical activity, the poli-
cies that shape those attributes, and the decision makers who can imple-
ment the policies.
TABLE 2-1 Relationships Among Selected Influences on Physical Activity
Environment Attribute Policy Determinant Decision Makers
Mixed land use Zoning Local governments (as
informed by planning officials)
Street connectivity Guidelines, standards Institute of Transportation
Engineers; developers; local
governments
Residential density Zoning Local governments (as
informed by planning officials)
Pedestrian/bicycle facilities Transport/complete Transportation departments;
streets* state and local governments
Traffic volume/speed Transport Transportation departments;
governments
Transit access Transport Transportation departments;
governments
Parks, trails Park and recreation Developers of national
standards and funding standards; local governments
Private recreation facilities Marketplace Business owners
Aesthetics, vegetation Various Multiple
School grounds, siting Standards, joint-use State education departments
agreements and governments
*A complete street is a road designed and operated to be safe for all users, including drivers,
bicyclists, transit vehicles and users, and pedestrians of all ages and abilities.
SOURCE: Sallis, 2011.
OCR for page 11
11
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
Measurement Tools
Given this portrait of the possible ways of boosting physical activity
levels, what would be the objectives of measuring progress? First, Sallis
observed, one would want to measure the simple presence or absence of an
environmental attribute or policy, as well as its characteristics and quality.
For example, does the park have a playground? Second, the quality of the
attribute should be evaluated. For example, is the equipment in good shape?
Is the physical education program required or recommended? Third, are
there disparities in access? In using data on these attributes and policies,
one must also consider the geographic scale of the measures, Sallis added.
For example, if a county had 5 acres of park land for every 1,000 people,
one would still want important to know how the parks were distributed by
neighborhood, so local-level data are important. Sallis outlined a number
of options for measuring environmental attributes and policies.
Surveys
In surveys, people are asked to report on the attributes of their neigh-
borhoods. There are validated measures of attributes of neighborhood
environments, and the Centers for Disease Control and Prevention (CDC),
for example, has funded a brief survey (Sallis et al., 2010) on neighbor-
hood environments that provides a validated measure. However, Sallis said,
validated, self-reported surveys of parks, trails, and school environments
are lacking. He also said that measures need to be adapted for racial and
socioeconomic status subgroups, and that measures for rural environments
are still in development. Sallis believes a survey approach could be used for
national surveillance, but research is needed to refine such an approach.
Surveys would be the lowest-cost option and could be deployed nationwide
relatively quickly.
Observation or Audits
Trained data collectors can count streets, parks, trails, and other fea-
tures, and these counts provide useful data. However, Sallis explained, these
data are expensive to collect. They can yield large amounts of information,
but there is as yet no accepted method of scoring and summarizing the find-
ings. Researchers also have not yet devised ways to connect these counts
to people’s behaviors or to place them in context. Ideally, Sallis suggested,
more concise instruments and improved scoring procedures would allow
community or advocacy groups to assess their own neighborhoods. In addi-
tion, web-based programs such as Google could be a lower-cost alternative
to in-person audits.
OCR for page 12
12 MEASURING PROGRESS IN OBESITY PREVENTION
Geographic Information System (GIS)
A GIS has been defined as the “integration of software, hardware, and
data for capturing, storing, analyzing, and displaying all forms of geo-
graphically referenced information.”1 This sort of mapping can provide
highly detailed data, for example, about the land devoted to different pur-
poses in a local area and, Sallis explained, has the potential for managing
and displaying national-level data as well. At present, however, data are
collected by a variety of local and national entities, including tax asses-
sors, departments of parks and transportation, and private companies. The
quality and currency of the data are inconsistent, Sallis said, and there is
little standardization in what is collected or how, or in what is accessible to
researchers. For example, a transportation department might have detailed
information about local roads but collect nothing on sidewalks or bicycle
facilities because they are not a priority. Thus, Sallis explained, if agencies
not typically concerned with public health issues could be persuaded to
include health-related data in their collection efforts, and if consensus could
be developed on variables, a GIS could be much more useful for collecting
data related to physical activity.
Policy Measures
Enumerating and rating policies at the local level that may have an
impact on physical activity is another approach. However, Sallis explained,
it is difficult to collect information on and monitor local policies, and there
is significant variation in their nature and purpose. An online system for
tracking such policies would be useful, he suggested, as would increased
standardization of ways to describe policies and their specific attributes.
Further work on ways to assess variations in the ways policies are imple-
mented would also be valuable.
Crime Measures
Crime data are useful for understanding local environments, and a
national system for collecting such data would be useful. Unfortunately,
Sallis explained, the way information is coded in this area also lacks stan-
dardization. Connecting data to specific geographic points can be difficult,
and the associations between crime levels and physical activity have not
been clearly identified in research. Public health researchers would benefit
from a systematic research agenda and from work with the Department of
Justice to develop standardized accessible data.
1 See http://www.gis.com/content/what-gis (accessed August 30, 2011).
OCR for page 13
13
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
Collaboration with Non-health Sectors
Sallis concluded by noting that, while health researchers and agencies
need more and better data to understand clearly the influence of the physi-
cal environment on obesity, improved data systems to meet this need will
depend on collaboration with other groups outside of the health field. Such
groups have different priorities from health research entities and are just
as likely to face tight budgets. Nevertheless, much of the data on land use,
transportation, education, crime, and commerce already being collected
could be useful for health research, and could be more useful if coordinated
and expanded in even modest ways. Such data would be extremely helpful
to local and state health departments, but the existing state of affairs is
difficult for them to sort out. “Health agencies will need to invest in these
systems,” said Sallis. “Who will take the lead?”
SURVEILLANCE
Presenter: Christine Hoehner
Surveillance is important for public health researchers in any field,
and each of the measures Sallis described may be used in surveillance,
explained Hoehner. Surveillance serves numerous purposes. Using surveil-
lance, researchers can assess the magnitude of a problem and how it is dis-
tributed geographically, as well as monitor changes over time in its patterns.
Surveillance can also help in defining a problem and generating hypoth-
eses or research targets and in evaluating the effectiveness of intervention
strategies, as well as support planning decisions and policy development.
Most important, Hoehner observed (quoting a publication by Institute of
Medicine committee member Jamie Chriqui and colleagues), is that “what
gets measured, gets changed” (Chriqui et al., 2011). She explored whether
current measures related to physical activity and the built environment are
“accurate; reliable; feasible to collect across diverse communities; sensitive
in detecting change to the environment and policies associated with physical
activity; and responsive to the data needs of advocates, decision makers,
and planners at the local level.”
When surveillance is effective, the result is an ongoing information
loop, Hoehner explained, that works as shown in Figure 2-3. In practice,
however, the surveillance loop for a given public health issue is usually
incomplete because connections may not have been forged, and data may be
unavailable. In these cases, decisions are made without the necessary data.
The loop also illustrates that the physical environment and activity levels
are influenced by many sectors of public life and levels of government and
that policies involve many stages, said Hoehner.
OCR for page 14
14 MEASURING PROGRESS IN OBESITY PREVENTION
Program
Data
Planning
Interpretation
Information
Data Program
Dissemination
Analysis Implementation
Data Program
Collection Evaluation
FIGURE 2-3 Public health surveillance loop.
SOURCE: Remington and Goodman, 1998. Reprinted with permission from the
2-3.eps
Sheridan Press: [American Public Health Association] Remington, P. L., and R. A.
Goodman. 1998. Chronic disease surveillance. In Chronic disease epidemiology
and control. 2nd ed., edited by R. C. Brownson, P. L. Remington, and J. R. Davis.
Washington, DC: American Public Health Association.
National Objectives for Physical Activity Environments and Policies
Ideally, surveillance is guided by national objectives. Hoehner explained
that in the case of physical activity, the National Physical Activity Plan—
developed under the sponsorship of numerous organizations, including the
YMCA, the American Cancer Society, and the American Heart Association,
and launched in 2010—describes policies and programs designed to pro-
mote physical activity “to improve health, prevent disease and disability,
and enhance quality of life.”2 The plan offers recommendations directed
to eight sectors with potential influence: business and industry; educa-
tion; health care; mass media; parks, recreation, fitness, and sport; public
health; transportation, land use, and community design; and volunteer and
nonprofit organizations. These recommendations include surveillance and
reporting of data.
Another initiative, Healthy People 2020, outlines a set of specific
objectives for boosting physical activity (selected objectives are shown in
Box 2-1) and emphasizes the importance of nationally representative data.
CDC also includes among its objectives for reducing obesity three that are
specific to physical activity, along with suggested measures. All of these
plans highlight the fact that the necessary data are not consistently avail-
able, particularly at the local level, Hoehner pointed out.
2 See http://www.physicalactivityplan.org/.
OCR for page 15
15
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
BOX 2-1
Physical Activity Environment/Policy Objectives for
Healthy People 2020
1. Increase the proportion of the nation’s public and private schools that require
daily physical education for all students.
2. Increase regularly scheduled elementary school recess in the United States.
3. Increase the proportion of school districts that require or recommend elemen-
tary school recess for an appropriate period of time.
4. Increase the proportion of the nation’s public and private schools that provide
access to their PA [physical activity] spaces and facilities for all persons outside
of normal school hours.
5. Increase the number of states with licensing regulations for PA provided in
child care.
6. (Developmental)* Increase the proportion of employed adults who have access
to and participate in employer-based exercise facilities and exercise programs.
7. (Developmental)* Increase legislative policies for the built environment that
enhance access to and availability of PA opportunities.
*There is currently no national baseline data for developmental objectives, but they should
have a confirmed nationally representative data source that will ultimately provide baseline
data and at least one tracking point. Developmental objectives address areas of national
importance for which investments should be made over the next decade to measure their
change (http://healthypeople.gov/2020/about/aboutdata.aspx [accessed August 30, 2011]).
SOURCE: http://healthypeople.gov/2020/topicsobjectives2020.
Measurement Tools and Gaps
At present, surveys, GIS, and auditing and observational tools are the
primary sources of data for physical activity and the built environment, as
Sallis described. Hoehner provided additional detail on how these three
approaches are used to measure physical activity.
Surveys
Surveys generally are conducted by mail or telephone and provide
specific information about the characteristics of an environment, depend-
ing on the number of questions asked (which may vary significantly). Two
instruments used for surveillance of the built environment are the Environ-
mental Supports for Physical Activity survey (SIP 4-99 Research Group,
2002) and the Physical Activity Neighborhood Environment Scale (PANES)
(Sallis et al., 2010). While these instruments provide valuable data at low
OCR for page 16
16 MEASURING PROGRESS IN OBESITY PREVENTION
cost and cover many regions where observational data are unavailable, it
is important to note, Hoehner explained, that agreement between the per-
ceptions captured by surveys and objective measures is relatively low. Dif-
ferences may be accounted for by differing definitions of a neighborhood’s
boundaries or other relevant factors, variation in people’s expectations and
perceptions, or measurement error. Hoehner stressed that, although the
d ifference should not be overlooked, the perceived environment can
b e measured reliably and is associated with physical activity. She noted
that it would be useful to collect more of this type of data because the tools
needed to do so are available at the national level.
Geographic Information Systems
GIS data provide the only feasible objective measures of the built envi-
ronment across large areas, Hoehner explained. Although much of this
type of data currently is collected by non-health sectors, she continued, it
provides information about parks, indoor recreation facilities, land use,
streets and public transit, vegetation, traffic accidents, and neighborhood
deprivation. The sources and scope of these data are shown in Table 2-2.
Although valuable, GIS-based data can be costly to collect, and they
currently vary considerably in terms of regions covered, geographic scale,
and type of data. Locally collected data may not be readily available to
researchers, for example. Quality may vary as well, and in some cases the
origin of the data is unknown. There is relatively little standardization
in what is included in the data, and in many cases, details about features
particularly relevant to physical activity are lacking. For example, Hoehner
described her experience with data on parks in the Dallas-Fort Worth area
that were difficult to use for research. From an initial count of 2,800 parks,
she and her colleagues eventually arrived at a count of approximately 2,000
because records from multiple sources may have included features that were
not actually parks, such as medians, cemeteries, or mobile home parks. At
the same time, some parks evident in aerial photographs or other sources
had to be added.
Commercial databases are another source of information about the
locations of food and physical activity establishments that can be integrated
with GIS, Hoehner noted, and two studies have examined their validity.
Boone and colleagues (2008) compared commercial data with data from
a field census and found moderate agreement: 39 percent for nonurban
areas and 46 percent for urban areas. The agreement varied by facility
type. Hoehner and Schootman (2010) conducted a similar comparison—of
measures by InfoUSA and Dun & Bradstreet (two major commercial data
sources) versus independent measures of census tracts in the St. Louis area.
They found “mostly fair” agreement among databases: 32 percent. These
OCR for page 17
17
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
TABLE 2-2 GIS-Based Data Collected by Non-health Sectors
Type Source Scale
Parks and open space Park and recreation departments Local
National/state parks National/state
Indoor recreation facilities Commercial databases (InfoUSA, National
Dun & Bradstreet)
Land use Parcel databases Local
Density U.S. Census/American National
Community Survey
Streets Census TIGER National
ESRI Streetmap National
Transportation or planning Local
agency
Sidewalks and bicycle facilities Transportation or planning Local
agency
Vegetation U.S. Geological Survey’s National
Landsat*
USDA National Agriculture National
Imagery Program
Crime FBI Uniform Crime Reports National
(city/county)
Local police departments Local
Traffic accidents Motor vehicle accident reports State or local
Neighborhood deprivation U.S. Census/American National
Community Survey observations
NOTES: ESRI = Environmental Systems Research Institute; ESRI develops GIS to address
social, economic, business, and environmental concerns at the local, regional, national, and
global scales. FBI = Federal Bureau of Investigation. TIGER = Topologically Integrated Geo-
graphic Encoding and Referencing; Census TIGER is the name for the system and digital
database developed at the U.S. Census Bureau to support its mapping needs for the decennial
census and other Bureau programs. USDA = U.S. Department of Agriculture.
*Landsat is a global land-imaging project that provides space-based images of the earth’s
land surface, coastal shallows, and coral reefs.
SOURCE: Hoehner, 2011.
databases contain a great deal of error, Hoehner observed, so researchers
should exercise caution when using them.
In short, GIS-based measures are valuable, but the lack of standard-
ization in the underlying spatial data limits their usefulness, and there is a
paucity of studies including measures of sidewalks, crime, park qualities,
OCR for page 18
18 MEASURING PROGRESS IN OBESITY PREVENTION
and vegetation. Additionally, GIS-based measures provide limited infor-
mation about many features, particularly those related to the interactions
between the built and social environments, such as crime, disorder, and
public access.
Audits and Observational Tools
These tools provide measures of urban design features that are not
observable remotely as with GIS or aerial photographs, such as sidewalk
quality, incivility, or lighting.3 They are more useful for community assess-
ment and advocacy than for surveillance. Hoehner and her colleagues
(Brownson et al., 2009) reviewed 20 tools used for audits and observations
and found that they covered a range of domains. Most covered land use,
streets and traffic, sidewalks, bicycling facilities, public space/amenities,
building characteristics, parking, maintenance, and safety. Few, however,
covered such issues as the presence of dogs, noise, or signage. These tools
have high rates of interrater reliability, in the range of 0.6 to 0.8; the rates
are highest for measures of physical features, such as land use and trans-
portation environments, and lower for other attributes, such as social and
physical disorder. Challenges in the use of these tools are similar to those
for GIS-based data: cost, lack of scoring protocols, and difficulty of analyz-
ing complex data.
Analysis, Interpretation, and Dissemination of Data
Given the available measures described above, Hoehner noted, a variety
of questions about analyzing, interpreting, and disseminating the data must
be considered. For example, if improving GIS measures in order to develop
an improved national sample were identified as a goal, one would need to
consider who would be responsible for collecting, analyzing, and dissemi-
nating the data (for example, CDC, state or local health departments, uni-
versities); which geographic areas should be covered; what resources would
be required; what measures would be reported; and how the data would be
made accessible and useful to advocacy groups and decision makers.
Several current initiatives, Hoehner observed, involve collecting stan-
dardized data across jurisdictions. Communities Putting Prevention to
Work, a project of the U.S. Department of Health and Human Services,
has engaged approximately 50 communities in working to reduce the rate
of diseases related to obesity and tobacco consumption using environmen-
3 Many tools for audits and observations are listed on the Active Living Research website,
maintained by the Robert Wood Johnson Foundation (www.activelivingresearch.org [accessed
July 2011]).
OCR for page 19
19
PHYSICAL ACTIVITY AND THE BUILT ENVIRONMENT
tal policy strategies. Communities are encouraged to use an evaluation
tool—designed to identify existing community activities and develop action
plans—that Hoehner said has been deemed useful for surveillance. Bridging
the Gap, a program funded by the Robert Wood Johnson Foundation, is
examining policy and environmental measures in a nationally representa-
tive sample of 150 to 200 communities defined by school catchment areas.
Other initiatives include Healthy Kids; Healthy Communities, which uses
online tracking to explore community partnerships; and the Childhood
Obesity GIS System, an online tool for mapping many types of data. Finally,
the National Collaborative on Childhood Obesity has developed a web-
based registry of valid and reliable measures4 (discussed in greater detail
in Chapter 4), which Hoehner explained is expected to help in identifying
gaps in measurement.
Recommendations
Hoehner closed by offering some recommendations, many of which
echoed Sallis’s comments. Working with those outside the health sector who
do or can collect valuable data, she believes, will be critical to improving
the information base. At the same time, it will be important to give prior-
ity to measures that address national objectives and strategies. Collecting
data periodically will enable the assessment of trends over time. Hoehner
also argued that methods and measures that can be developed most easily
and quickly to support the development and modification of policies for
the built environment should be an early priority. Finally, both the estab-
lishment of priorities and the lessons to be learned from existing initiatives
should support sound decisions about the collection, analysis, and inter-
pretation of these data.
REFERENCES
Boone, J. E., P. Gordon-Larsen, J. D. Stewart, and B. M. Popkin. 2008. Validation of a GIS
facilities database: Quantification and implications of error. Annals of Epidemiology
18(5):371-377.
Brownson, R. C., C. M. Hoehner, K. Day, A. Forsyth, and J. F. Sallis. 2009. Measuring the
built environment for physical activity. State of the science. American Journal of Preven-
tive Medicine 36(Suppl. 4).
Chriqui, J. F., J. C. O’Connor, and F. J. Chaloupka. 2011. What gets measured, gets changed:
Evaluating law and policy for maximum impact. Journal of Law, Medicine and Ethics
39(Suppl. 1):21-26.
Ding, D., J. F. Sallis, J. Kerr, S. Lee, and D. E. Rosenberg. 2011. Neighborhood environment
and physical activity among youth: A review. American Journal of Preventive Medicine
41(4):442-455.
4 See www.nccor.org/measures (accessed September 14, 2011).
OCR for page 20
20 MEASURING PROGRESS IN OBESITY PREVENTION
Hoehner, C. M. 2011. The physical inactivity, inactivity, and built environments: Current and
potential sources of measures for assessing progress in obesity prevention. Presented at
the Workshop on Measurement Strategies for Accelerating Progress in Obesity Preven-
tion, March 23, Irvine, CA.
Hoehner, C. M., and M. Schootman. 2010. Concordance of commercial data sources for
neighborhood-effects studies. Journal of Urban Health 87(4):713-725.
Remington, P. L., and R. A. Goodman. 1998. Chronic disease surveillance. In Chronic disease
epidemiology and control, 2nd ed., edited by R. C. Brownson, P. L. Remington, and J. R.
Davis. Washington, DC: American Public Health Association.
Sallis, J. 2011. Measuring environmental and policy exposure: Which ones? Presented at the
Workshop on Measurement Strategies for Accelerating Progress in Obesity Prevention,
March 23, Irvine, CA.
Sallis, J., and J. Kerr. 2006. Physical activity and the built environment. President’s Council
on Physical Fitness and Sports Research Digest 7(4).
Sallis, J. F., R. B. Cervero, W. Ascher, K. A. Henderson, M. K. Kraft, and J. Kerr. 2006. An
ecological approach to creating active living communities. Annual Review of Public
Health 27:297-322.
Sallis, J. F., J. Kerr, J. A. Carlson, G. J. Norman, B. E. Saelens, N. Durant, and B. E. Ainsworth.
2010. Evaluating a brief self-report measure of neighborhood environments for physical
activity research and surveillance: Physical Activity Neighborhood Environment Scale
(PANES). Journal of Physical Activity and Health 7(4):533-540.
SIP 4-99 Research Group. 2002. Environmental supports for physical activity questionnaire.
Columbia, SC: University of South Carolina, Norman J. Arnold School of Public Health,
Prevention Research Center. http://prevention.sph.sc.edu/tools/Env_Supports_for_PA.pdf
(accessed August 17, 2011).
