Background:
Studies have reported inconsistent findings for the relationship between active commuting to school (ACS) and physical activity or weight status/adiposity among youth. These mixed findings may be due to reliance on self-report measures of physical activity, samples drawn from specific populations, inadequate statistical power, or lack of data on dietary energy intake, which may confound the relationship. No reports have examined physical activity as a mediator of active commuting and weight status/adiposity.

Objectives:
To examine (1) the relationship between ACS and weight status/adiposity and (2) whether moderate-to-vigorous physical activity (MVPA) mediates the relationship between ACS and weight status/adiposity among a nationally representative sample of US youth.

Methods:
We conducted a cross-sectional analysis using multiple linear regression models on a nationally representative sample of participants aged 12-19 years from NHANES, 2003-2004 (n=827 unweighted, n= 15,946,216 weighted). The main exposure variable of ACS was assessed by recall of walking or bicycling as part of getting to and from school over the past 30 days. Anthropometric measurements (height and weight for BMI-z, waist circumference, skinfold thicknesses [sum of subscapular and triceps]) were obtained using standardized techniques and equipment. Physical activity was objectively measured by accelerometers, using 1-minute epochs. Besides overall daily MVPA, we also examined the subset of before- and after-school MVPA (Mon - Fri: 6:30-9 am and 2:30-4 pm). Dietary energy intake was a covariate and calculated from the mean of two 24-hour recalls. Other covariates included age, gender, race/ethnicity, and poverty-to-income ratio. We used the product-of-coefficients method to test for mediation.

Results:
32.9% of participants aged 12-19 years actively commuted to school at least two days per week. In multivariate models, controlling for age, gender, race/ethnicity, poverty-to-income ratio, and energy intake, ACS was inversely associated with BMI-z score (std. beta=-0.08, p=0.040, R²=0.06) and skinfold thicknesses (std. beta=-0.06, p=0.033, R²=0.13). ACS was positively associated with overall daily MVPA (std. beta=0.11, p=0.039, R²=0.28) and with before- and after-school MVPA (std. beta=0.20, p<0.001, R²=0.22). Before- and after-school MVPA significantly mediated the relationship between ACS and waist circumference (p=0.03) and skinfold thicknesses (p=0.03). Total minutes of MVPA mediated no relationships.

Conclusions:
Among a nationally representative sample of US youth, active commuting to school was associated with lower BMI-z and adiposity. Additionally, those who actively commuted to school achieved greater MVPA. Before- and after-school MVPA was a mediator between ACS and measures of obesity and body fatness. ACS appears to be an important behavior to target for population-based physical activity promotion and obesity prevention interventions.

Support:
US Department of Agriculture Current Research Information System