Comparison of accelerometers and a self-report record

The health benefits of physical activity (PA) have been well documented. Perhaps equally important but much less studied are the health-related problems associated with sedentary behavior. The purpose of this study was to examine the convergent validity of the ActiGraph and activPAL Professional accelerometers with the Bouchard Activity Record (BAR). Sixteen males and sixteen females wore the two accelerometers and completed the BAR concurrently for 1 day. Descriptive statistics and delta values were reported for the ActiGraph, activPAL and the BAR. Aggregate time spent in sedentary behavior and walking between all instruments was analyzed using Repeated Measures Analysis of Variance. Dependent t-tests were used to analyze time in standing between activPAL and BAR and moderate-vigorous PA (MVPA) between ActiGraph and BAR. Epoch-by-epoch data were compared using mean percent agreement. There was a significant difference found in time spent in sedentary behavior with post hoc pairwise differences occurring between the activPAL and ActiGraph as well as between the ActiGraph and BAR. There was also a significant difference detected in time spent in walking. The post hoc analysis showed the differences existing between the activPAL and ActiGraph as well as between the ActiGraph and the BAR. In the epoch-by-epoch analysis, mean percent agreement ranged from 54.0 (for walking as detected by the activPAL and the ActiGraph) to 86.7 (for MVPA as detected by the ActiGraph and the BAR). The activPAL and BAR detected significantly different time spent standing and ActiGraph and BAR detected significantly different time spent in MVPA. The overall findings were that activPAL and BAR showed convergence on both an aggregate level and epoch-by-epoch in both sedentary behavior and PA. The comparative discordance between the activPAL and BAR with the ActiGraph was likely a function of the cutpoints used to distinguish sedentary behavior from PA.