Beyond the home: Effects of including more than one location on modeled relationships between environment and physical activity

Purpose: The dominant methodology for assessing the effects of environment on physical activity focuses on the home environment. However physical activity may occur throughout the day in other locations. These environments may play a role in influencing amount of physical activity conducted, and should be considered.

Methods: Eight studies using GPS and accelerometer devices were pooled together, representing a range of participants and lifestyles. The Personal Activity and Location Measurement System (PALMS) was used to identify locations for each participant’s day. Amount of time spent in location was derived from GPS traces, and the top three locations per day were extracted. Ordinary least squares regression compared a single location model to a multi-location model of walkability as a predictor of daily mean physical activity (>100 CPM) controlled for age, wear time, and sex.

Results: Preliminary results for 41 subjects with 273 days of data show an average of 3.3 locations per day. Average minutes spent in the primary location were 936±404 min. Secondary and tertiary locations had 113±109 min and 37±35 min of time spent respectively. Regression results with participants’ primary location indicate that higher walkability had a positive impact on minutes of physical activity (coeff(SE): 2.033(0.384), p < 0.01). When including three locations, the primary location retained significance (2.054(.0389), p < 0.01), however the secondary location was also significant (1.742(.0187), p<0.01).

Conclusions: Physical activity is occurring in more than the home location, and environmental factors of other locations may be important factors in driving physical activity.