ARS | Publication request: Impact of a walking school bus program on children's active commuting to school and physical activity

Submitted to: Pediatric Academic Society
Publication Type: Abstract Only
Publication Acceptance Date: January 1, 2010
Publication Date: May 1, 2010
Citation: Mendoza, J.A., Watson, K., Baranowski, T., Nicklas, T., Uscanga, D.K., Hanfling, M.J. 2010. Impact of a walking school bus program on children's active commuting to school and physical activity [abstract]. Pediatric Academic Societies Annual Meeting, May 1-4, 2010, Vancouver, Canada. 2010 CDROM.

Technical Abstract: Walking School Buses (WSB) are groups of students, led to and from school by adults, in which students are picked up at designated "bus stops". The impact of WSB programs on students' active commuting to school (walking/cycling to school, ACS), and moderate-to vigorous-physical activity (MVPA) has not been studied in a group randomized controlled trial (RCT). To conduct a pilot evaluation of a WSB program's impact on students' ACS and MVPA. We conducted a group RCT in 8 low socioeconomic status (SES), elementary schools in Houston, TX. The students (n=149) were ethnically diverse 4th graders. The intervention was a WSB program led by study staff 5-days/week. Schools were matched by race/ethnicity and % qualifying for free/reduced lunch, and then randomized to intervention or control conditions. Outcomes were measured pre- and 5-weeks post-intervention. The main outcome measure was the weekly rate of ACS. Daily MVPA, measured by accelerometers, was a secondary outcome (n=75). Potential covariates included child characteristics (age, gender, BMI z-score, and walking distance to school), and psycho-social variables (child and parent self-efficacy and parent outcome expectations). We used mixed-model regression analyses, including students' baseline characteristics and psycho-social variables, to account for clustering by school. We included two-way interactions with group to identify moderators. We used a stepwise procedure with backward elimination of non-significant interactions and covariates to identify significant moderators and predictors. Intervention students had 26.4% +/- 30.8% higher rates of ACS and had 9.2 +/- 5.4 more minutes of daily MVPA than control students at post assessments. Baseline ACS and parent outcome expectations were significant moderators of ACS at post assessment. As expected, baseline ACS was more strongly related to post ACS in the control schools, indicating that control students were likely to do ACS at post assessment only if they did ACS at baseline. In the intervention group, students whose parents had high outcome expectations at baseline were more likely to do ACS at post assessment whereas the relationship was weaker among the controls. In conclusion, a WSB program appears promising to improve ACS and daily MVPA among urban, low SES, elementary students. Future WSB programs should address parents' outcome expectations.