Assessment of accelerometer non-wear time in children

Abstract

Different definitions for determining non-wear time are currently used. The aim of this study was to determine the effect of the definition of ActiGraph and activPAL non-wear time on data retention and the classification of sedentary and sitting time in children. Methods: Accelerometer and inclinometer data were drawn from three studies conducted in Victoria, Australia, using comparable protocols. Boys’ and girls’ (n=626) sedentary time was assessed using ActiGraph accelerometers for 8 consecutive days. Sitting time was also assessed using activPAL inclinometers. Four definitions were applied to identify non-wear time: 10, 20, 30 and 60 minutes of consecutive zeros. Descriptive analyses were initially conducted to examine the effects on sample size retention, wear time, and sedentary or sitting time estimates. Univariate ANOVAs were used to determine whether significant differences were apparent in sedentary or sitting time estimates depending on non-wear definitions. Results: Using the 10 minute non-wear definition, fewer children were retained compared to the 60 minute definition for both the ActiGraph (487 vs. 509 children) and the activPAL (379 vs. 416 children). Increasing wear time, sedentary time and sitting time were observed with increasing non-wear definitions; significant differences were apparent between all definitions except between the 20 and 30 minute definition. Conclusions: Non-wear definitions had a significant impact sample retention, wear time and average daily sedentary time. Researchers should be aware of the influence of non-wear criteria on sedentary and sitting time estimates. Future studies are needed to validate non-wear time definitions in children.