Wrist-Based Accelerometer Cut-Points to Identify Sedentary Time in 5–11-Year-Old Children

Jessica Chandler, Michael Beets, Pedro Saint-Maurice, Robert Weaver, Dylan Cliff, Clemens Drenowatz, Justin B. Moore, Mei Sui, Keith Brazendale

The objective of this paper is to derive a wrist-placed cut-point threshold for distinguishing sedentary behaviors from light-intensity walking using the ActiGraph GT3X+ in children. This study employed a cross-sectional study design, typically used in measurement-related studies. A sample of 167 children, ages 5⁻11 years (mean ± SD: 8.0 ± 1.8 years), performed up to eight seated sedentary activities while wearing accelerometers on both wrists. Activities included: reading books, sorting cards, cutting and pasting, playing board games, eating snacks, playing with tablets, watching TV, and writing. Direct observation verified sedentary behavior from light activity. Receiver operator characteristic (ROC) analyses were used to determine optimal cut-point thresholds. Quantile regression models estimated differences between dominant and non-dominant placement. The optimal cut-point threshold for the non-dominant wrist was 203 counts/5 s with sensitivity, specificity, and area under the curve (AUC) of 71.56, 70.83, and 0.72, respectively. A 10-fold cross-validation revealed an average AUC of 0.70. Statistically significant (p ≤ 0.05) differences in median counts ranging from 7 to 46 counts/5 s were found between dominant and non-dominant placement in five out of eight sedentary activities, with the dominant wrist eliciting higher counts/5 s. Results from this study support the recommendation to place accelerometers on the non-dominant wrist to minimize "noise" during seated sedentary behaviors.