Maximizing the detection of youth sedentary behavior with accelerometer data

Purpose: Accelerometers are commonly used as an objective measure of sedentary behavior, but the method of defining nonwear time directly affects sedentary estimates. The present study examined nonwear definitions and their impact on detection of sedentary time in youth using different ActiGraph models and filters.

Methods: 60 youth (34 children and 26 adolescents; ages 5-17; 54.4% female) were asked to wear a 7164 and a GT3X+ ActiGraph side-by-side on a belt during a 90-minute structured sedentary activity (sitting on a couch watching a movie). Research assistants observed participants to ensure they remained seated. Data from the GT3X+ were downloaded using the normal (N) and LFE extension (LFE). Five nonwear definitions were applied to the 7164, GT3X+N and GT3X+LFE data, and sedentary estimates were computed.

Results: The 7164 was most sensitive to movement, requiring shorter nonwear definitions to detect 100% of observed sedentary time (i.e., 30 minutes of consecutive zero counts for children and 60 minutes for adolescents) compared to the GT3X+N, which was least sensitive and required longer definitions (i.e., 60 minutes for children and 90 minutes for adolescents). The GT3X+LFE detected 100% of sedentary time with a 60 minute definition for both children and adolescents.

Conclusions: Different nonwear definitions are needed for children and adolescents, and for different Actigraph models and filters. The recommendation for children is 30 minutes for the 7164 and 60 minutes for the GT3X+N and GT3X+LFE. For adolescents, 60 minutes is recommended for the 7164 and GT3X+LFE and 90 minutes for the GT3X+N.