Accelerometer and GPS Analysis of Trail Use and Associations with Physical Activity and Sedentary Time

Abstract

PURPOSE: The combined use of accelerometers and global positioning system (GPS) devices may provide insights into how trails can support physical activity (PA) and an alternative approach for objectively measuring specific activities on trails, such as cycling. The aims of this study were to examine associations between trail use and PA and sedentary behavior (SB) and to quantify on-trail PA with accelerometer activity counts only and then with a combination of accelerometer and GPS data. METHODS: Participants (N=141, 53% female, 19-78 y), who were recruited on five trails in Massachusetts, wore an Actigraph 7164 accelerometer and a GeoStats GPS data logger concurrently for 1-4 days. Total PA (daily mean counts·min-1), and daily minutes of light (LPA=100-759 counts), moderate (MPA=760-5724), and vigorous PA (VPA≥5725), and SB (0-99) were derived from activity counts. A trail use day was defined as a day on which a participant engaged in a minimum of 2 consecutive minutes of activity on a trail as determined by GPS coordinates. Mixed linear models were used to examine whether trail use was associated with PA outcomes and SB, controlling for demographics, trail site, weekday/weekend trail use, SB for PA outcomes, and PA (all intensities) for SB outcome. Intensity of PA on trails was quantified using counts only and a combination of counts and GPS speed. Ainsworth’s compendium was used to classify intensity based on the metabolic equivalent levels for different cycling speeds. RESULTS: Trail use had positive associations with LPA (β=7.7 min·d-1, p=0.04), MPA (β=28.3 min·d-1, p<.0001), and total PA (β=522 counts·min-1, p<.0001). Trail use was not associated with VPA or SB. Minutes of activity on trails classified as VPA increased by 346% when counts and GPS speed were used to define intensity, compared to using activity counts only. On-trail LPA, MPA, and SB decreased by 15%, 91%, and 85%, respectively. CONCLUSION: Adults accumulated more LPA, MPA, and total PA on days when they used trails, indicating the importance of these resources for supporting regular PA. The combination of GPS and accelerometer data may be useful for classifying PA intensity on trails where individuals are likely to be cycling. Supported by funding from the Active Living Research Program, the Robert Wood Johnson Foundation.