1. Texas Childhood Obesity Prevention Policy
Evaluation (T-COPPE Project): Baseline data
from Safe Routes to School Policy evaluation
Co-Leads:
Deanna Hoelscher, PhD, RD, LD
Marcia G. Ory, PhD, MPH
2. Presentation Overview
¨ Why are we doing TCOPPE?
¨ Background and rationale
¨ What are we doing?
¨ Methods
¨ What did we find?
¨ Results
¨ What does it all mean?
¨ Discussion
3. The Rule of 2s
¨ Two policies to evaluate
¤ Both aspects of the energy equation: activity and
nutrition
• Safe Routes to Schools (SRTS)
• WIC
¤ Two environments
• School/home environment
• Grocery stores/home
¤ Two functional timelines
• School year
• Year round
4. Slide 2 of 2s
¨ Two state universities in a unique and effective
working relationship…
¤ The University of Texas School of Public Health
¤ Texas A&M (Health Science Center School of Rural
Public Health)
¨ Two rivalries……
5. Why Evaluating Childhood Obesity &
Prevention Policies?
National
¨ Robert Wood Johnson Foundation is
impact… committed to reducing childhood obesity
starts with by 2015
evidence of
¨ There are a number of national policies
local impact
identified as aimed at reducing
childhood obesity (i.e., SRTS)
¨ Which of these national policies are
actually shown to be effective in reducing
childhood obesity?
¨ What is the impact of implementing these
national policies locally?
6. How did we select our policies?
¨ Potential for evidence of effectiveness
¨ Political feasibility
¤ Potential for leadership engagement
¤ Champions identified in Texas
Legislature and State Government
¨ Public Acceptability
¤ Readinessand feasibility in implementation
¤ Documented history of obesity efforts during last decade
7. How did we select our policies? (cont.)
¨ Partnership support
¤ LiveSmart Texas coalition development/support
¤ Partnership for a Healthy Texas
¨ Policy sustainability
8. Standardized Mode of Transportation
Trips to School, 1965-2005
Source: McDonald, N.C. 2007. American Journal
of Preventive Medicine
9. Walking/Biking by Distance to School,
1965-2005
Source: McDonald, N.C. 2007. American Journal
of Preventive Medicine
10. Active Commuting to School
¨ Current childhood obesity epidemic
¨ Children are not meeting current recommendations for
physical activity
¨ One strategy to increase physical activity among children:
¤ Walking or biking to and from school (Active Commuting to
School or ACS)
¨ Currently, approximately less than 16% of children use
ACS
13. Safe Routes to School Policy
¨ SAFETEA-LU
¤ 2005 Federal Transportation Bill
¤ % of nation s total children K-8 to offer increased physical activity
through health alternatives to bus and car school transportation
¤ Texas received about $40 million in Safe Routes to School (SRTS) funding
between 2005 and 2009
¨ State support for SRTS
¤ In September 2007, The Texas Transportation Commission approved
$24.7 million for 244 projects in 66 communities throughout the state
¤ Supplemented by revenue from the God Bless Texas and God Bless
America specialty license plates
¤ Funds administered through grant process
Source: Tx DOT, 2008
14. Texas SRTS Policy
SRTS
¨ For Texas, two major types of grants
were awarded:
Education
¤ Construction
(Infrastructure) grants, which
Enforcement include brick and mortar type projects,
Encouragement such as construction of crosswalks,
sidewalks, etc.
Evaluation
n Schools need to have a SRTS plan in place
Engineering first
¤ Planning
grants, which include a SRTS
plan, which may or may not include
potential infrastructure changes or
implementation of the plan.
15. Methods
¨ Purpose
Measures: ¤ Todetermine the effects of differing
• Student (4th allocation methods of funding
grade) survey (construction versus planning grants)
• Parent survey
from the Texas Transportation
Commission on parent attitudes &
• ACS
behaviors.
• School Checklist
¨ Natural experiment
• Campus Policy
¤ Quasi-experimental
• School Audit
¨ Initial study assumptions
¤ Foractive commuting to schools (ACS),
construction (infrastructure) schools >
planning schools > comparison schools
16. Methods
Baseline ¨ Funded schools were selected for
measurement based on funding type,
data location (urban/rural), race/ethnicity, and
collected socioeconomic status (SES); comparison
in 2009 schools had similar characteristics but
received no funding.
¨ Data were analyzed using mixed linear
regression and controlled for random and
fixed effects, and other independent
variables.
18. SRTS Baseline Data
School Survey Status Infrastructure Planning Control Total
Schools Schools Schools
Measured Spring 2009 11 13 13 37
Measured Fall 2009 14 9 21 44
Total 25 22 34 81
Survey Activity Total (to date)
Student Survey 3315
Parent Survey 2057
Parent/Student Survey 1653
Combination
Active Transport Count 12,167
Environmental Audit 79 *
* Two school environmental audits were not done due to safety concerns for the auditors
19. Demographic Variables by School
Condition (n = 81)
Variable Construction Schools Planning Schools (n Comparison Schools
(n = 25) = 22) (n = 34)
Student Male (%) 49.9 51.9 47.5
Parent race/ethnicity
White (%) 19.8 30.3 24.3
Other (%) 80.2 69.7 75.7
Economically 75.5 65.8 68.3
disadvantaged (%)
All family members 64.1 55.3 57.7
born in USA (%)
Mean time to school
<5 m (%) 27.0 28.5 20.6
5-10 m (%) 38.5 39.3 36.2
11-20 m (%) 20.9 20.3 25.5
>20 min (%) 13.6 11.9 17.6
20. Baseline Rates of Active Commuting to
School (ACS), n = 79
ACS in
1969 (48%)
Mean % ACS
*Means
are
significantly
different
from
comparison
schools
(p<0.05)
Ac>ve
Commu>ng
is
2-‐day
self-‐reported
walking
or
biking
to
or
from
school
Analyses
are
controlled
for
%
economically
disadvantaged,
%
white,
mean
precipita>on,
mean
heat,
mean
wind
speed.
21. Mean Active Commuting to School
Before School After School Mean ACS
n=79 n=79 n=79
10.4% 17.1% 13.8%
Analyses were conducted using Mixed Effects Linear Regression
22. School Environment
(Rural)
¨ School in rural area.
¨ Only 2 segments
indicated on audit tool.
¤ One was a one-lane
gravel road that
separated school
property from a corn
field.
26. Differences in Parent Attitudes &
Behaviors by School Types at Baseline
Variable
Construc0on
Schools
Planning
Schools
Comparison
Schools
Mean
(SD)*
Mean
(SD)
Mean
(SD)
n
=
25
n
=
22
n
=
34
Asking
Behavior
Scale
1.11
(.09)
1.29
(.10)a
.98
(.08)a
Parent
Self-‐Efficacy
Scale
18.6
(.4)a
20.4
(.5)ab
18.7
(.4)b
Parent
Outcome
13.9
(.2)
a
14.7
(.2)
ab
13.7
(.2)
b
Expecta>ons
Student
Self-‐Efficacy
Scale
27.2
(.5)
a
27.2
(.6)
b
24.7
(.5)
ab
Arrived
Walking
(%)
10.5
(1.5)
a
9.1
(1.7)
b
4.6
(1.3)
ab
Arrived
Biking
(%)
0.5
(0.6)
a
2.5
(0.6)
ab
0.5
(0.5)
b
Arrived
by
School
Bus
(%)
16.4
(4.1)
a
14.0
(4.5)
b
26.9
(0
ab
3.5)
TV
on
during
evening
meal
3.51
(.11)
a
3.14
(.12)
ab
3.58
(.09)
b
TV
>me
on
weekends
4.49
(.07)
4.32
(.08)
a
4.59
(.06)
a
School level analyses using weighted UNIANOVA
27. What Factors are Associated with
Walking or Biking to School (Child)?
Factors NOT ¨ Students who walked or biked to school
Associated
with ACS: were more likely to:
¤ Have a friend who walked or biked to
• Screen time
school
• Days PA
¤ Have self-confidence that they could walk
• Safety
to school
• Social
¤ Feel that they could ask their parents to
support
walk or bike to school
• Attitude
28. What Factors are Associated with
Walking or Biking to School (Parent)?
¨ Parents more likely to let their children commute to
school:
¤ Had higher self-efficacy (self-confidence) in letting their
child walk to school
¤ Perceived better walkability around their house and their
child s school
¤ Were more likely to let their children walk to other places
from school
¤ Reported better outcomes associated
with walking to school (e.g., children
would be healthier)
¤ Reported fewer barriers to commuting
29. Does weather influence ACS?
¨ Students decrease ACS in the morning due to
precipitation (marginally significant, p-
value=0.099)
¨ When the morning temperature was warmer, ACS
was higher (p-value=0.019)
¤ Morning
temperature range = 10.4-89.6 degrees
Fahrenheit
¨ Analysis
¤ Covariates
in the Mixed Effects Linear Regression
Modeling of ACS
30. Implications
¨ Number of children
walking or riding a
bike to school was low
¨ We need policies that
promote environments
that are conducive to
walking and biking
¨ We need safety and
perception of safety
31. Policy Implications
¨ Many parent-related variables were consistent with
ACS
¤ Parents are highly motivated to participate and be
engaged
¤ Parents made a point to contact both PI and Project
Director to express interest and ask how else to be
involved
¨ Need programs that focus on parent education
¨ Need programs that make neighborhoods safer (e.g.,
benefits of complete streets)
32. Conclusions
¨ Significant differences were seen in ACS between
planning/construction and comparison schools
¤ Outcome expectations, self-efficacy, TV
¤ Grant application process encouraged schools to collect
pilot data
n Smaller
grants (planning grants) may be as effective in getting
ACS as larger grants (construction grants) initially
n Grant processèAwarenessèMore likely to engage in ACS?
n More likely to have a program champion?
n Planning schools had greater percentage of children who biked
¨ Allocation of resources may be given to schools who are
already working on SRTS
¤ How do we reach other schools?
33. Conclusions
¨ Child behaviors associated with walking & biking to
school included asking behaviors & having friends
commute
¨ Programs like SRTS increase walking and bike
riding
35. Why School Audits?
¨ Important role of the built environment in promoting
WTS.
¨ Recognition of the many micro-scale and modifiable
barriers at/around schools.
¨ Importance of the context-specific and detailed
environmental features in changing school travel
behaviors
à Lack/shortage of instruments designed to capture
school environments systematically and
comprehensively
37. Audit Components and Items
FORMAT:
Letter-size sheets with checklist, rating, closed-end choices, a
nd mapping items
COMPONENTS:
A. STREET AUDIT
B. SCHOOL SITE AUDIT
C. MAP AUDIT – sidewalk, bike lane, drainage ditch, buffer, tr
ail, crosswalk, and bus stop
§ Land Uses
§ Street/traffic/parking conditions
§ Lighting , other amenities, and sigs
§ Unattractive items
§ Perceptual rating items (surveillance, maintenance, cleanliness, vis
ual quality, safety and attractiveness)
38. Street Segment Audit
• Audit info. • Segment Image
- Auditor info. - Indicating each
- Date, weather segment
- Start/end time - North up
- Street name
• Perceptions
• Audit Items
- For objective
observations
• Map Audit Indic
ators
- If related items
present, go to
Map Audit(s)
39. School Site Audit
• School Site Ima
ge
- Indicating
School site and
• Frontage property line
- Maine entry
- Street facing
- Vehicular and
pedestrian entries
• On-site facilities
- Physical features
• Main entry
- Amenities, etc.
- Amenities around
main entry
• D/P Area
- Location, types,
and capacity
40. Map Audit Example
Map audit A : sidewalk &
informal path
• Exact locations
• Detailed conditions
slope, shade, width, holes & cracks
, bumps & uneven surface , weed
s , litter , drainage problems, etc.
• Obstructions
poles , parked cars, mail boxes,
etc.
• Connections
42. Descriptive Findings from 79
TCOPPE Schools audited across Texas
Street & Map Audit Elements
Requiring Improvements:
¤ Bike lanes (98% lacked)
¤ Benches and trash cans (96%)
¤ Traffic calming devices (85%)
¤ Unattractive items/social disorder
(75% with 1+)
¤ Street lights (25% lacking)
¤ Sidewalk obstructions (many with
poles, parked cars, mail boxes, etc.)
43. Descriptive Findings from 79
TCOPPE Schools audited across Texas
School Site Audit Elements Requiring Improvements:
¤ Designated drop-off/pick-up area (21 lacked)
¤ Adjacency to vacant/abandoned/undeveloped areas (19
schools)
¤ Lack of walkway connections to school buildings (14 lacked)
¤ Trails/paths within campus (73 lacked)
44. Frontage Street Audit Items
Correlated with % Walkers
Variables B Sig.
Presence of sidewalks 10.996 0.001
Presence of street parking 7.143 0.012
Presence of vacant areas -6.999 0.022
Presence of unattended/stray -8.358 0.050
dogs
Presence of drainage ditches -6.853 0.047
Surveillance* 2.030 0.058
Safety in walking* 3.033 0.013
Safety in bicycling* 3.453 0.008
Attractiveness in walking* 2.459 0.048
Attractiveness in bicycling* 2.451 0.047
*likert-type scale item (1 being poor to 5 being excellent)
45. Other Street Audit Items
Correlated with % Walkers
Variable B Sig.
Number of 4 – 10 7.090 0.055
intersections* 11+ 6.194 0.064
Number of street 1-3 6.854 0.037
lights** 4+ 4.202 0.247
Presence of street parking 4.628 0.094
Presence of street calming -7.178 0.019
devices
Presence of safety/child 7.943 0.006
crossing sign
Presence of landscaped buffer 7.642 0.008
Presence of drainage ditch -5.094 0.096
Presence of crosswalk 6.308 0.081
* The reference category is 0-3 driveways.
** The reference category is 0 street light.
46. School Site Audit Items
Correlated with % Walkers
Variable B Photograph by Yang Mi Kim
Sig.
Number of school bus only entry & exit -2.717 0.051
Number of pedestrian only entries & exits 1.562 0.028
Presence of vacant area -7.179 0.029
Presence of sidewalk/walkway connection 9.234 0.016
Presence of private car area -7.163 0.050
Presence of basketball/tennis/volleyball court 7.147 0.006
Presence of baseball/football/soccer field -6.616 0.016
Presence of outdoor swimming pool 5.427 0.056
Presence of bench / seating 6.411 0.019
Presence of picnic table 7.604 0.015
* The reference category is none of evergreen tree.
47. Conclusion
¨ This School Audit Instrument is a tool that can provide
effective and efficient assessments of street and school
site environments, focusing on those attributes related to
children s active transportation to school.
¨ The instrument s three components help objectively
identify many easily modifiable elements, facilitating
policy development toward creating safe and walkable
communities.
¨ With proper training, this audit can be used for education,
research, intervention, and policy-support purposes.
¨ The instrument can be shortened and customized, once
more data are collected from diverse communities.
49. Multi-pronged Support and
Dissemination System…
¨ Partnership with Texas Health Institute part of initial funding proposal with
expectation of
¤ Legislative policy forums in years 1, 2, 3 and 5 of the grant
¤ Sharing activities and findings in real time
¨ Support and advisement from Texas Obesity Policy Research Advisory
Council (TOPRAC) whose mission is
¤ To provide health policy research, translation, evaluation, and dissemination
support to TCOPPE and Live Smart Texas
¨ Regular feedback to Live Smart Texas (LST) collaboration
¤ Texas coalition working collaboratively on obesity prevention efforts and the
development of resources to fund it
¤ TCOPPE is LST s first major research project
¨ Respond to opportunities as they arise and are appropriate
50. Additional opportunities…
¨ Testimony to the Institute of Medicine on Childhood
Obesity Prevention workshop in Texas, February 2009
¨ Annual participation in the Texas Obesity Awareness
Week events at the Texas Capital
¨ Invited testimony to Texas legislative committees on the
state of obesity in Texas
¨ Development of a Strategic Communications Plan
¤ Intensive communications workshop provided by RWJF to
selected individuals
¤ To provide focus and benchmarks for monitoring success and
outlining timely policy forum opportunities
51. Conclusions
¨ Close collaboration and communication with
stakeholders at multiple levels
¨ Dissemination throughout the project
¨ Establish team of credible experts who can inform
and educate legislators—the go to team
¨ Policy makers knowledgeable about
issue before research conclusions
are made/available
52. It takes more than a Village to
do this Texas-sized project…
It takes a TEXAS-sized team…
¤ Roy Allen ¤ Jingang Miao
¤ Heather Atteberry ¤ Lisako McKyer
¤ Chester Bryant ¤ Hyung Jin Kim
¤ Arthur Castro ¤ Deb Kellstedt
¤ Yichen Cheng ¤ Tiffni Menendez
¤ Diane Dowdy ¤ Marcia Ory
¤ Sandra Evans ¤ Courtney Peterson
¤ Kyna Farmer ¤ Mike Pomeroy
¤ Selina Guerra ¤ Donna Nichols
¤ Emily Hines ¤ John Reilly
¤ Deanna Hoelscher ¤ Tina Simms
¤ Leah Kolar ¤ Carolyn Smith
¤ Pat Koym ¤ Christine Tisone
¤ Chanam Lee ¤ Suojin Wang
¤ Kris Lykins ¤ Pete Walton
¤ Jay Mendoza ¤ Jerri Ward
¤ Ann Mesaros ¤ Cheryl Brien-Warren
53. Contact Information
Live Smart Texas
¤ Tiffni Menendez, MPH
512-391-2512
Tiffni.Menendez@uth.tmc.edu
T-COPPE
¤ Diane Dowdy, PhD
979-458-4249
Dowdy@srph.tamhsc.edu
54. Acknowledgements
This work was partially supported by three Robert Wood
Johnson Foundation grants (64634, 63755, 65539).
We would like to thank:
n Arthur Casto for his help with the audits.
n Jun Hyun Kim, Carolyn Smith, Ashley Wilson, and Chelsea
Mounce for the valuable inputs during the instrument
development phases.
n Dr. Woosung Lee for his help with the data analyses.
To request a copy of the instrument & manual, please
contact Diane Dowdy, PhD, TCOPPE Project Director:
Dowdy@srph.tamhsc.edu
55. Current Stats: The Walking School Bus and Children's
Physical Activity Study
¨ Objective: Evaluated a walking school bus program on active commute and moderate to
vigorous physical activity (MVPA)
¨ Intervention: Walking school bus (a group of children led by an adult to and from school )
¤ Intervention group: n=4 schools; 70 4th graders
¤ Control group: n=4 schools; 79 4th graders
¤ 76% of total students from low-income families (<= $30,000)
¤ 91% of students Hispanic; 47% of students Black
¨ Measures: self-questionnaire and accelerometry at Time 1 and Time 2
¨ Results:
¤ Intervention group increased daily minutes of MVPA from 46.6 +/- 4.5 at Time 1 to 48.8
+/- 4.5 at Time 2
¤ Control group decreased daily minutes of MVPA from 46.1 +/- 4.3 at Time 1 to 41.3 +/-
4.3 at Time 2
Source: Mendoza JA, et al. Pediatrics, 2011.
56. The Walking School Bus and Children's Physical
Activity Study: continued
¨ Objectives:
¤ Evaluate the feasibility of a protocol to measure changes to children s
pedestrian safety behaviors
¤ Evaluate the potential influence of the WSB program, neighborhood safety, and
intersection characteristics on children s pedestrian safety behaviors at the
school-level
¨ Intervention group: Taught and modeled safe pedestrian behaviors during walks
from trained staff members
¨ Control group: Received usual information from school district about school
transportation
¨ Results: impact on pedestrian behaviors is unknown
¤ Child pedestrians at the intervention schools had a five- fold higher odds of
crossing at the corner or crosswalk compared to pedestrians at control schools
¤ Child pedestrians at the intervention school also had five-fold lower odds of
stopping at the curb versus control schools
Source: Mendoza JA, et al. Health Place, 2012.
57. Methods
¨ Baseline data collected for T-COPPE Study
th
¨ 4 grade students and parents were recruited
through 81 schools
¨ Active Transport Survey/Counts
¤ Collected in classrooms
¤ 2-day self-report
¤ Validated instrument