Dog Walking as Physical Activity and Multi-Level Correlates of Dog Walking among Adolescents

PURPOSE: To assess the association of dog ownership and dog walking with moderate to vigorous physical activity (MVPA) and BMI among adolescents and examine correlates of dog walking. METHODS: Participants were adolescents (n=928) from the Baltimore, MD and Seattle, WA regions, ages 12-17 years. Measures included 7 days of accelerometer monitoring, GIS measures of the built environment, and surveys of psychosocial and environment characteristics. Minutes/day of MVPA and BMI were compared among adolescents (1) without a dog (n=441) and those with a dog who (2) do (n=300) or (3) do not (n=184) walk it. Among adolescents with a dog (n=484), models assessed relations between dog walking (days/week) and (1) demographic, (2) psychosocial (physical activity self-efficacy), (3) home environment (electronic device ownership), (4) perceived neighborhood environment, and (5) objective neighborhood walkability factors. All models were adjusted for participant demographic factors. RESULTS: Adolescent dog walkers obtained 4-5 more minutes/day of MVPA than those with dogs but were non-dog-walkers (p=.044) and those who lived in non-dog households (p=.025). Adolescents who walked their dog were more likely to have a more educated parent (p=.056), own more personal electronics (p=.047), have better perceived neighborhood aesthetics (p=.016), live in a more walkable neighborhood (p=.045), and live in a multifamily home (p =0.003) compared to adolescents with dogs who did not walk them. Adolescents’ age interacted with physical activity self-efficacy (p=.045) and perceived neighborhood aesthetics (p<.001), where self-efficacy and aesthetics were positively associated with dog walking among younger adolescents and not associated in older adolescents. Adolescent BMI was not associated with dog walking or dog ownership. CONCLUSIONS: Dog walking was an important contributor to overall MVPA, where dog walkers obtained 7-8% more minutes/day of MVPA than those who did not walk them, and correlates of dog walking were found at multiple levels of influence. Results suggest interventions to increase dog walking at the environmental and psychosocial levels are worthy of evaluation, particularly given that nearly 50% of US households own dogs.