Date of Award


Document Type

Open Access Dissertation


Exercise Science


The Norman J. Arnold School of Public Health

First Advisor

Michael W. Beets


The purpose of this dissertation was twofold: 1) to establish best practices for wrist-based accelerometry for 5-11 year old children and 2) to explore the contribution of activity structure to children’s physical activity (PA) levels.

The purpose of study 1 aimed to determine differences in counts/5second epoch produced by the dominant and non-dominant wrist during seated, sedentary activities in 5-11 year old children and ultimately create a cutpoint threshold to distinguish seated sedentary behavior from light physical activity. 167 children, ages 5-11 years, performed up to 8 sedentary activities for 5 minutes while wearing ActiGraph GT3X+ accelerometers on both wrists. Participants walked at a normal pace to elicit light physical activity. The optimal cutpoint threshold for the non-dominant wrist was 203 counts/5s with sensitivity, specificity, and an 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.

The purpose of the study 2 was to develop an equating system to translate estimates of moderate-to-vigorous physical activity (MVPA) collected at the wrist to those previously published collected at the hip. 185 children ages 5-11 years wore an ActiGraph GT3X+ on the hip and the non-dominant wrist for up to three hours. Data was distilled into minutes of MVPA using six commonly used hip-based cutpoints (i.e. Evenson, Pate, Puyau, Mattocks, VanCauwenberghe, and Freedson) and one non-dominant wrist-based set of cutpoints (i.e. Chandler). Among the six developed regression equations, the proportion of variance ranged from Freedson 0.29 to Puyau 0.81 and the absolute error ranged from Pate 20.7% to Puyau 42.9% at the individual level. When cross-validated at the group level, % differences in actual versus predicted values ranged from Puyau -1.8% to Mattocks 56.3%.

The purpose of Study 3 was to determine which structure of PA elicits the most MVPA: 1.) free play (FP), 2.) organized, adult-led activities (ORG) or 3.) a mixed condition in which both FP and ORG were offered during the same opportunity (MIX). Participants included 197 unique children that were 53% male, 55% Caucasian, and averaged 7.7 years. Statistically significant differences were observed in the percent of time boys spent in MVPA during FP and MIX compared to ORG sessions (35.8% and 34.8% vs. 29.4%). No significant difference was observed in the percent of time girls spent in MVPA during FP compared to ORG or MIX (27.2% and 26.1% vs 26.1%). Both boys and girls experienced ~10% less time sedentary during FP compared to ORG and MIX.

This dissertation was the first to assess the contribution of activity structure to children’s PA levels. It was also the first to our knowledge to address the timely matter of the use of wrist-based accelerometry to assess PA levels and how this change impacts accurate measurement and comparisons of PA in youth. In conclusion, this dissertation represents a novel approach in the analysis of children’s physical levels through measurement and activity structure.