Hawai‘i's Opportunity for Active Living Advancement (HO‘ĀLA): Addressing Childhood Obesity through Safe Routes to School

Abstract

Increasing active transportation to and from school may reduce childhood obesity rates in Hawai‘i. A community partnership was formed to address this issue in Hawai‘i's Opportunity for Active Living Advancement (HO‘ĀLA), a quasi-experimental study of active transportation in Hawai‘i County. The purpose of this study was to determine baseline rates for active transportation rates to and from school and to track changes related to macro-level (statewide) policy, locally-based Safe Routes to School (SRTS) programs and bicycle and pedestrian planning initiatives expected to improve the safety, comfort and ease of walking and bicycling to and from school. Measures included parent surveys, student travel tallies, traffic counts and safety observations. Assessments of the walking and biking environment around each school were made using the Pedestrian Environment Data Scan. Complete Streets and SRTS policy implementation was tracked through the activities of a state transportation-led Task Force and an advocacy-led coalition, respectively. Planning initiatives were tracked through citizen-based advisory committees. Thirteen volunteer schools participated as the intervention (n=8) or comparison (n=5) schools. The majority of students were Asian, Native Hawaiian, and Pacific Islander in schools located in under-resourced communities. Overall, few children walked or biked to school. The majority of children were driven to and from school by their parents. With the influence of HO‘ĀLA staff members, two intervention schools were obligated SRTS project funding from the state, schools were identified as key areas in the pedestrian master plan, and one intervention school was slated for a bike plan priority project. As the SRTS programs are implemented in the next phase of the project, post-test data will be collected to ascertain if changes in active transportation rates occur.

Introduction

Active travel modes to and from school contribute significantly to physical activity rates as well as lower obesity rates among school children.^1–3 Conversely, children taking motorized transportation to and from school have shown a two to three pound per year weight gain.⁴ Over time - in accordance with rising obesity rates (especially for children in low-income households⁵) - the percentage of United States youth walking or bicycling to school has drastically declined.⁶ Among the many reasons for this shift are a lack of safe facilities and a perceived lack of safety among parents.^7,8 In response, communities across the United States have adopted Safe Routes to School (SRTS) programs which take a comprehensive approach to improving bicycling and walking to school.⁹

Another emerging trend in transportation policy reform is Complete Streets (CS) policies. These policies require roads to accommodate all users. It is often pointed out that CS policies can augment SRTS programming. And, research shows up to a three-fold increase in active transportation to school among children after the addition or improvement of bicycle lanes, traffic signals, crosswalks and sidewalks.¹⁰ What is not clear is whether statewide CS and SRTS policies help to increase physical activity rates and decrease obesity rates among children by requiring (via CS policy) and accelerating (via SRTS policy) both the necessary engineering improvements and Hawai‘i's Opportunity for Active Living Advancement (HO‘ĀLA): Addressing Childhood Obesity through Safe Routes to School the equally important education, encouragement and enforcement steps to get more children walking and bicycling to school.

To address this question of macro-level policy influence on childhood obesity, this study capitalized on two recently passed statewide policies for CS (Act 54¹¹) and SRTS (Act 100¹²) as well as accompanying bicycle and pedestrian planning initiatives and new SRTS programs expected to have impacts over the next 3–5 years on the built environment around school zones. The study was coordinated by an advocacy-academic partnership between Peoples Advocacy for Trails Hawai‘i (PATH) and the University of Hawai‘i at Manoa, with strong support from the County of Hawai‘i. The purpose of this manuscript is to describe the progress to date and baseline results for the first 6 months of the project.

Methods

Participants

Hawai‘i's Opportunity for Active Living Advancement (HO‘ĀLA) was funded for a 12-month period by the Robert Wood Johnson Foundation through the Active Living Research Program. In this quasi-experimental study we recruited 13 schools on the island of Hawai‘i (33% of all schools on the island) to participate as either an intervention school (receiving SRTS assessments and SRTS programming) or comparison school (receiving only the SRTS assessments). Schools in under-resourced communities were targeted using several steps. Initially, all eligible schools were sent a postcard notifying them of the study. Eligibility criteria included: (1) having ≥35% of students who qualified for free and reduced school lunch; (2) were willing to fulfill study requirements; and (3) were rural (<20% of student lived within 1-mile) or neighborhood (>60% of students lived within 1-mile). Next, the schools were emailed and mailed informational packets explaining study details and offering $1000 mini-grants to those that would like to participate as intervention schools (comparison schools received the assessments for free). A second follow-up e-mail was sent and interested schools responded by e-mail. Finally, phone calls were made to interested schools to set up an in-person meeting to finalize the recruitment process. School administrators chose whether their school would be an intervention (n=8) or comparison (n=5) school.

Measures

Standardized measures developed by the National Center for SRTS were used to assess travel modes to and from school for both students and parents (http://www.saferoutesinfo.org/resources/index.cfm). The Parent Survey about Walking and Biking to School was distributed to all students to take home to their parents. The parent survey was estimated to take 5–10 minutes and gathered information about factors that affected whether or not parents would let their children walk or bike to school. Parents indicated their child's grade and gender and the total number of children they had in grades K-8. They were asked to indicate how far their child lived from school and how “on most days” their child arrived and departed from school (i.e., walking, biking, school bus, family vehicle, carpool, transit, or other). In addition, parents were asked to indicate how long it normally took their child to get to and from school, if their child had asked for permission to walk or bike to or from school, and their opinions on issues affecting their decision to allow their child to walk or bike to or from school. Finally, parents indicated their own level of education and could provide additional comments.

The Student Travel Tally was used to assess how children traveled to and from school. Teachers in first and fourth grade classrooms at each school were asked to conduct the travel tally with their class. Students were asked “How did you arrive at school today?” and “How do you plan to leave for home after school?” Students then raised their hands to indicate the travel mode (identical to the parent survey categories) the used for each trip. Weather conditions, class size, and the number of students present at the time of the tally were recorded.

Traffic Counts and Safety Observations were taken using methodology from the PATH Hawai‘i SRTS Toolkit.¹³ Trained observers were stationed at standardized locations around each school during the 1 1/2 hours before school began and after school ended. For the traffic counts, observers indicated the number of people that passed through their observation zone using one of the following transportation modes: car, public transportation, bike with a helmet, bike without a helmet, pedestrian and other (e.g., skateboard). For the safety observations, observers counted the occurrence of seven safety hazards (e.g., number of motorists failing to yield to pedestrians, number of cars speeding, jaywalking).

The Pedestrian Environment Data Scan (PEDS)¹⁴ was used to track the physical conditions of all street segments that comprised the main travel routes within 1/2 mile of each school. The PEDS provided a comprehensive, objective assessment of the built environment for walkability and bikeability. This one-page paper instrument consisted of four major sections: a) environment, b) pedestrian facilities, c) road attributes, and d) walking/cycling environment. Raters indicated the absence or presence of each item and counted items as appropriate. In addition, raters were asked to subjectively rate the attractiveness and safety of the segment for both walking and cycling on a 3-point scale from 1=strongly agree to 4=strongly disagree. Individual items of the PEDS have shown high inter-rater reliability, with 89% of items having 80% agreement or higher.¹⁴

The SRTS programmatic components that were tracked at each school included the formation of a SRTS team; completion of a SRTS action plan; completion of various educational, encouragement, or enforcement strategies as identified in the action plan; technical assistance provided for additional funding; and SRTS funding awards.

HO‘ĀLA project staff members attended and participated in all relevant planning or task force meetings to track the progress on the two policies and the bicycle and pedestrian initiatives. In addition, we took notes and saved copies of relevant documents.

Procedures

All study procedures were approved by the University of Hawai‘i Institutional Review Board. Data collectors were trained on the use of each data collection instrument, including a five-hour training session specifically for the PEDS instrument. Teams of data collectors were dispersed on “assessment days” at each school.

The initial year of the project was focused on assessing existing active transportation behaviors and physical infrastructure around Hawai‘i County schools while monitoring the first efforts to implement macro-level policies and plans as well as locally-based SRTS programs. All schools agreed to fulfill study requirements and both comparison and intervention schools received baseline SRTS reports about street segment conditions and how students traveled to and from school. Each intervention school agreed to host a SRTS workshop. Intervention schools then began PATH's “Three-Steps to Success” SRTS implementation model, resulting in the development of their own custom SRTS plan. Besides the $1,000 mini-grants, intervention schools received technical assistance from PATH to aid in the planning and implementation of their SRTS programs. Schools will now be implementing their SRTS programs and follow-up assessments are scheduled to determine the impact of the intervention.

All school data were analyzed with PASW Statistics 18 (Chicago, IL). Summary statistics were created to provide baseline characteristics for each school. Comparisons were made to determine if statistically significant differences existed at baseline between intervention and comparison schools. Progress on the policies and planning initiatives were summarized qualitatively.

Results

As shown in Table 1, five schools were in neighborhoods and eight were in rural settings. Enrollment ranged from 128 to 850 students. Asians, Native Hawaiians, and Pacific Islanders made up the ethnic majority for all but one of the schools. Percentages of free and reduced school lunches ranged from 35.9% to 93.5%. Baseline data were available from 8 intervention and 4 comparison schools.

Table 1.

School Characteristics

School #	Enrollment	Intervention or Comparison	Neighborhood or Rural	% Asian, Native Hawaiian, Pacific Islander	% Free and Reduced School Lunch
1	128	C	R	56.3	93.5
2	551	I	N	50.6	49.8
3	157	C	R	89.3	59.6
4	630	I	R	66.8	82.6
5	273	I	R	64.0	50.4
6	686	I	N	40.1	37.3
7	640	I	R	70.7	51.9
8	500	I	N	51.1	35.9
9	383	C	N	55.9	68.3
10	850	I	N	68.7	51.0
11	331	C	N	90.6	82.0
12	141	C	N	72.3	66.0
13	245	I	N	70.3	52.7

Parent Surveys

Survey response rates ranged from 13.1% to 58.4% per school with 1191 surveys completed at intervention and 457 surveys completed at comparison schools. Although parents of children from all grades were asked to complete a survey, about one-third of the responses were from parents of 1st and 4th graders (students targeted by the intervention). Children's genders were almost equally represented and most parents had 2 children in grades K-8. See Table 2.

Table 2.

Parent Survey Demographics

Characteristic	Intervention n (%)	Comparison n (%)
Grade of Child
1st	229 (19.0)	71 (15.7)
4th	196 (16.5)	67 (14.7)
Gender of Child
Female	491 (50.9)	212 (51.2)
Male	474 (49.1)	202 (48.8)
Average Number of Children in grades K-8 in Household	mean = 1.72, sd = 0.88	mean = 1.92, sd = 1.00
Parent had graduated college	298 (25.0)*	63 (13.8)
Total Number of Surveys	1191	457

^*p<.001

Statistically significant differences existed between intervention and comparison schools for household distance from school, χ²=121.1, p<.001, with comparison school parents tending to live closer to school. Almost half of the intervention (49.5%) but only 34.1% of the comparison school parents reported living more than 2 miles from school. In contrast, 25.4% of comparison and only 8.1% of intervention school parents reported living within 1/4 mile of the school. Table 3 displays the households' distances from school, modes of arrival, and modes of departure from school.

Table 3.

Distance, Mode of Arrival, and Mode of Departure from School

Variable	Intervention n (%)	Comparison n (%)
Household distance from school
< 1/4 mile	96 (8.1)	116 (25.4)
1/4 to 1/2 mile	75 (6.3)	53 (11.6)
1/2 to 1 mile	158 (13.3)	52 (11.4)
1–2 miles	199 (16.7)	42 (9.2)
> 2 miles	590 (49.5)	156 (34.1)
How Child Arrives at School
Walk	41 (7.4)	34 (7.4)
Bike	3 (0.3)	2 (0.4)
School bus	273 (22.9)	30 (6.6)
Family vehicle	814 (68.3)	353 (77.2)
Carpool	33 (2.8)	10 (2.2)
Transit
Other	0	1 (0.2), 2 (0.4)
How Child Departs School
Walk	56 (4.7)	50 (10.9)
Bike	3 (0.3)	2 (0.4)
School bus	295 (24.8)	33 (7.2)
Family vehicle	708 (59.4)	308 (67.4)
Carpool	22 (1.8)	9 (2.0)
Transit	17 (1.4)	1 (0.2)
Other	1 (0.1)	3 (0.7)

Statistically significant differences were also found between intervention and comparison schools for how children arrived at school (χ²=72.7, p<.001) and how they departed school (χ²=84.8, p<.001). As shown in Table 3, the main difference was that children at intervention schools were more likely to ride the bus to and from school if they were not driven to school by their parents.

Only 16% of students from intervention schools had asked their parents if they could walk or bike to school, while 34.1% of students from comparison schools had done so. However, the majority of parents at both school types (67.8% at intervention and 54.9% at comparison schools) reported that they would never let their child walk or bike to or from school without adult supervision.

The top six factors influencing parents' decisions whether to let their children walk or bike to school included distance, speed of traffic, amount of traffic, safety of intersections and crossings, weather or climate, and the conditions of sidewalks or pathways. As shown in Table 4, significantly more parents at intervention schools, as compared to comparison schools, reported these factors (except weather or climate).

Table 4.

Percentage of parents whose decisions to allow child to walk/bike were affected by each factor

Factor	Intervention Schools	Comparison Schools
Distance	68.0**	56.2
Speed of traffic along route	67.6**	51.2
Amount of traffic along route	65.2**	49.9
Safety of intersections/crossings	59.6**	50.3
Weather/climate	56.9	52.5
Sidewalks or Pathways	54.4**	40.7
Violence/crime	45.3	41.6
Time	39.0	35.4
Crossing Guards	32.9	32.2
Child's before/after school activities	27.6	27.6
Adults to walk or bike with	24.3	23.9
Convenience of Driving	23.4	29.3*

^*p<.05,

^**p<.001

The top six factors that, if they would improve, parents would be more likely to let their child walk or bike to school included the condition of sidewalks or pathways, safety of intersections and crossings, crossing guards, speed of traffic, amount of traffic, and adults to walk or bike with. As shown in Table 5 parents from comparison schools were more likely to report that they would let their child walk or bike if each factor improved.

Table 5.

Percentage of parents who would let child walk/bike if the factor improved

Factor	Intervention Schools	Comparison Schools
Sidewalks or Pathways	36.4	38.6
Safety of intersections/crossings	34.2	38.5
Crossing Guards	32.9	35.6*
Speed of traffic along route	29.1	34.1
Amount of traffic along route	28.9	33.7
Adults to walk or bike with	28.2	36.1*
Weather/climate	23.2	31.7*
Distance	22.9	29.4**
Violence/crime	22.4	25.4**
Time	20.3	29.3*
Child's before/after school activities	13.6	29.0***
Convenience of Driving	13.1	23.8**

^*p<.05,

^**p<.01,

^***p<.001

Student Travel Tally

Student travel tallies were completed in 40 classrooms at six intervention schools and 13 classrooms at five comparison schools. The average class enrollment was slightly higher at intervention schools (mean=22.65, sd=4.74 students) than at comparison schools (mean=18.62, sd=2.10 students). Rates of walking were comparable between intervention and comparison schools, averaging 2 or fewer students per classroom, as were rates of bicycling, averaging 1 or fewer students per classroom. In agreement with the parent surveys, more students from intervention schools reported riding the bus, averaging 4.6 students per classroom, as compared to 2.4 students per classroom at comparison schools. The majority of students reported riding in their family vehicle; averaging 14.8 students per classroom at intervention and 12.8 at comparison schools.

Pedestrian Environment

Using the PEDS, a total of 242 segments were assessed at intervention and 129 segments were assessed at comparison schools. As shown in Table 6, similarities were found for the frequency of intersections, pedestrian facilities, paved trails and sidewalks, walkway obstructions, road conditions, and a modal speed limit of 25mph, as well as for the lack of crosswalks, bicycle facilities, or amenities. Main differences consisted of more intervention school street segments having a slight hill (44.2%) and more comparison school street segments being flat (57.4%); fewer intervention school street segments had pathways in good condition, complete walkways within the segment, traffic control devices, or crossing aids. However, intervention school street segments were more likely to have buffers and speed limits ranging up to 55mph.

Table 6.

Street Segment Characteristics

Item	Intervention (242)	Comparison (129)
Grade of hill	44.2% had slight hill (n=107)	57.4% were flat (n = 74)
Intersections	83.5% had an intersection (n=202)	88.4% had an intersection (n=114)
Pedestrian facilities	64.5% yes (n=156)	66.7% yes (n=86)
Paved trail/sidewalk	Yes 61.6% (98/159)	Yes 68.2% (58/85)
Path condition	46.1% good (70/152)	63.5% good (54/85)
Walkway obstructions	42.4% of segments with a walkway (n=67/158)	41.2% of segments with a walkway (n=35/85)
Buffers	30.3% (47/155)	20.9% (18/86)
Is walkway complete?	58.7% yes (54/92)	78.8% (41/52)
Condition of road	75.3% good (177/235)	72.1% good (93/128)
Speed limit	Range 10–55, mode 25mph	Range 10–35, mode 25mph
Any traffic control devices	39.4% yes (93/236)	57.4% yes (74/129)
Any crossing aids	17% yes (40/235)	25.6% yes (33/129)
Crosswalks	66.5% had none	62.8% had none
Bicycle facilities	Only for 10 segments 4.1%	Only for 6 segments 4.7%
Any amenities in segment	Only for 7 segments 2.9%	Only for 5 segments 3.9%

Traffic counts confirmed the data reported from the parent surveys and student travel tallies. For the trip to school, cars comprised the majority of the traffic counts (82.1%, n = 3,641) up to 485 per school (total = 3,641). Buses comprised 1.7% of the total counts (n = 76), transporting multiple children on each bus. Slightly more pedestrians were counted than indicated by parent and student data (14.7%, n = 651). Few bicyclists were counted (0.8%, n = 37), although one intervention school had 15. Up to 9 other forms of transportation were counted per school (0.7%, n = 30). For the trip from school, travel modes were similar: 75.2% of children were transported in cars (n = 2478), 2.2% in buses (n = 73), 20.9% were pedestrians (n = 689), 0.7% were bicyclists (n = 24), and 0.9% used other forms of transportation (n = 31). The main problems reported during safety observations included cars not yielding to crossing pedestrians and speeding cars.

SRTS Progress at Intervention Schools

Each school received a baseline report, summarizing the SRTS and PEDS assessments. Reports were then disseminated to the greater community through an interactive Town Hall process. Schools formed SRTS teams and worked with those teams to develop SRTS Action Plans that identified key strategies that the school and community agreed upon. Schools have begun to implement their SRTS programs and follow up assessments were scheduled.

Policy Tracking

The Executive Director of PATH served on the Statewide Complete Streets Task Force, with the final policy document finalized on September 15, 2010. Next, the State of Hawai‘i and the counties will develop their policies using the model guidance created by the task force. In addition, PATH led a statewide coalition of over 30 organizations focused on implementation of the SRTS law passed in 2009, resulting in an obligation of $1.2 million for SRTS projects. An additional $700,000 was anticipated to be obligated by 2011. This represented the first obligation of SRTS funds in the state. Of these funds, approximately $600,000 should be awarded to two of the intervention schools in the HO'ĀLA project.

Planning Initiatives

HO‘ĀLA study staff members participated in all of the transportation planning initiatives underway in the state and ensured that SRTS solutions were considered in projects, with special attention paid to projects that would impact intervention schools. PATH was selected to serve on the 20 member statewide Pedestrian Master Plan Citizens Advisory Committee. This resulted in an emphasis on schools as a key area of concern in the plan. In addition, PATH was successful in identifying four high priority projects in the existing Bike Plan, one of which would directly impact an intervention school. PATH also hosted a Pedestrian Safety Action Plan Workshop that created a Pedestrian Plan for Hawai‘i County through the Strategic Highway Safety Plan Bicycle and Pedestrian Emphasis Area. The Workshop focused on SRTS as a key area of concern for Hawai‘i County and developed strategies for enhancing and maintaining pedestrian safety through SRTS.

Discussion

In the first six months of the HO‘ĀLA project, thirteen schools were recruited to participate and baseline observations were completed at twelve. Similar to other states, few students walked or biked to school in this ethnically diverse population in Hawai‘i.⁶ In almost all schools, the majority of children were driven to and from school by their parents. Real safety concerns, including missing infrastructure, existed around each school that were both noted by parents and by observers.^7,8 Although most parents said they would never let their child walk or bicycle to school, others indicated that if specific problems were addressed, they would be more likely to do so.

Project successes to date included the number of schools that signed up to participate (one in three schools on the island of Hawai‘i). Another area of success was the ability to help influence the Complete Streets policy language to specifically point toward the need to accommodate children in the planning and design of roadways. In addition, through the influence of HO‘ĀLA project staff, schools were included as one of the primary criteria in the methodology for determining project priorities in the Statewide Pedestrian Plan. Another area of success was in the collaboration between the University of Hawai‘i at Manoa and PATH, working together to accomplish rigorous evaluation with meaningful impact on the school communities.

Primary project challenges were in the formation of SRTS teams at some of the intervention schools. The rapid nature of the project did not allow for a lot of time to develop strong teams and this resulted in some lack of initiative and leadership in a few schools. This challenge was overcome by working closely with the schools to help them find ways to incorporate the SRTS effort into existing groups such as the Parent-Teacher Organization. Another challenge included the discontinuation of the Balanced Transportation Coordinator position in the Hawai‘i County Planning Department which led to a change in our initial project team structure. This was overcome by working closely with the Hawai‘i County Public Works Department and the Data Systems Department. In the end, these two challenges strengthened the project considerably, by allowing for closer management and guidance of the schools.

As SRTS programs are implemented in the next phase of this project, it is important to address traffic problems and missing infrastructure in order to increase walking and biking to school by low-income, Asian/Native Hawaiian/Pacific Islander children in Hawai‘i County. Project collaborations will help utilize these results to move forward necessary changes in programming, policies, and the physical environment.

Disclosures

This study was conducted while KMH was at the University of Hawai‘i at Manoa. The study was designed by KMH and LD. Data were gathered by LD, MP, and ACP. Data were entered and cleaned by DA and analyzed by KMH and DA. This manuscript was primarily written by KMH and LD with some assistance by DA. The authors have no financial relationships or conflicts of interest to disclose for this manuscript.