Development of a Method for Measuring Active Outdoor Play in Preschool-Aged Children: Integrating Accelerometer, GPS,...

Current physical activity measurement methods often fail to capture domain-specific contexts. This limitation hinders the understanding of active outdoor play in preschool-aged children (aged 3-5 years), a critical physical activity domain for healthy development. The primary objective of this study was to develop a method for measuring active outdoor play by integrating accelerometer, GPS, and time-use diary data. The secondary objective was to apply this method to quantify active outdoor play by movement intensity and examine its contribution to total physical activity.

This cross-sectional study included 92 preschool-aged children and their parents from Alberta, Canada. Data were collected in August, 2023-February, 2024 and analyzed in May, 2024-February, 2025. Children wore an accelerometer and a GPS monitor for 7 days, whereas parents completed a time-use diary documenting their child's daily activities. Data from these 3 sources were processed and integrated to derive active outdoor play variables. Active outdoor play was operationally defined as nontrip and nonorganized physical activity taking place outdoors. Accelerometer data classified physical activity versus stationary time. GPS data classified trip versus nontrip and indoor versus outdoor location. Time-use diary data classified organized versus nonorganized physical activity. Active outdoor play engagement was further classified into light intensity and moderate-to-vigorous-intensity using accelerometer data.

Applying this method to the present sample, children spent an average of 192.9 minutes per day in active outdoor play, with 143.1 minutes per day in light-intensity and 49.8 minutes per day in moderate-to-vigorous-intensity active outdoor play. On average, active outdoor play accounted for 60.6% of total physical activity.

This study introduced a multisource method for measuring active outdoor play in preschool-aged children, addressing key limitations of traditional single-source methods. Future research could explore this method's applicability to enhance domain-specific physical activity knowledge, beyond overall physical activity, across contexts.