| Summary |
Guidelines recommend that children and youths participate in > 60 minutes of physical activity per day. In order to reliably assess children's activity patterns, the measurement instrument and protocol must capture the intermittent nature of children's physical activity. Because accelerometers can assess different intensities and durations of physical activity, they are widely used in research for measuring children's physical activity. To assess the intermittent patterns of children's physical activity using accelerometers, it has been suggested that researchers use short time sampling intervals, or epochs. Because accelerometers sum activity counts over the selected epoch, short bursts of vigorous activity, which generate high activity counts, may not be detected when activity counts are summed over epochs typically used in current research (e.g., 30-seconds or 1-minute). In order to reliably assess young children's physical activity, the sampling interval effect needs further investigation. The purpose of this study was to investigate the effect of different epochs on the assessment of physical activity in pre-school children using Actigraph accelerometers. Seventy-two boys and girls (mean age = 3.9 ± 0.6 years) wore an Actigraph accelerometer set to record in 1-second epochs on their right or left hip (i.e., counterbalanced) for one day (mean of 9 hours) during pre-school and after-school hours. One-second epochs were reintegrated into 3-, 5-, 15-, 30- and 60-second epochs for comparison. Previously published cut-points were used to estimate minutes of moderate, moderate-to-vigorous, and vigorous intensity activity for each epoch setting. Repeated measures ANOVA and 99% confidence intervals were used to evaluate mean differences between estimated minutes of each outcome variable for each Actigraph epoch setting. Following ANOVA, Fisher's LSD tests were used to examine mean differences between each epoch setting. Effect size estimates were calculated with Cohen's delta. A significant epoch effect was present for time spent at moderate, moderate-to-vigorous, and vigorous intensity activity (p < .01). In general, minutes of activity differed significantly between the three smallest epoch settings (i.e., 1-, 3-, and 5-second) and the three largest epoch settings (i.e., 15-, 30-, and 60-second), although differences between 1- and 3-second and 1- and 5-second epochs for vigorous intensity activity were statistically significant (p < .01), with small to moderate effect sizes. In conclusion, statistically significant epoch effects are present when measuring children's physical activity of moderate intensity or greater. These findings indicate that the use of longer epochs would produce fewer minutes of moderate-to-vigorous physical activity and the use of epochs as short as 1-second may be necessary in order to reliably capture preschool children's physical activity. |
