Related Experiment Video
Updated: May 13, 2026
Physical Activity Measurement in Children Accepting Table Tennis Training
Published on: July 27, 2022
Validity of the ActivPAL™ and the ActiGraph monitors in preschoolers
Ellen De Decker1, Marieke De Craemer, Alejandro Santos-Lozano
11Department of Movement and Sport Sciences, Ghent University, BELGIUM; and 2Faculty of Health Science, University of León, Ponferrada, SPAIN.
Insights
Objective measures like the ActiGraph and ActivPAL show poor accuracy for tracking preschooler sedentary behavior. The ActivPAL also caused skin irritation, reducing its practical utility for this age group.
Area of Science:
- Pediatric health research
- Behavioral science
- Biomedical engineering
Background:
- Sedentary behavior in preschoolers is a growing public health concern.
- Accurate measurement of sedentary behavior is crucial for intervention development.
- Objective measures are preferred over subjective methods for assessing physical activity and sedentary time.
Purpose of the Study:
- To compare the criterion validity of the GT1M ActiGraph and ActivPAL™ against direct observation for measuring sedentary behavior in preschoolers.
- To assess the convergent validity between the ActiGraph and ActivPAL™ sedentary behavior estimates.
- To evaluate the practical utility of the ActivPAL™ in a preschool setting.
Main Methods:
- Fifty-two preschoolers (4–6 years) wore ActiGraph and ActivPAL™ devices for five days.
- Direct observation was used as the criterion measure for sedentary behavior during classroom activities.
- Receiver operating characteristic (ROC) curve analysis assessed criterion validity; repeated-measures ANOVA assessed convergent validity.
Main Results:
- Both ActiGraph and ActivPAL™ demonstrated poor classification accuracy (Area Under the Curve ≈ 0.6) compared to direct observation.
- ActivPAL™ reported lower sedentary time than ActiGraph, with wide limits of agreement, indicating poor convergence.
- Parental reports revealed 38% of children experienced skin irritation from the ActivPAL™.
Conclusions:
- Current objective measures (ActiGraph, ActivPAL™) have limited accuracy for assessing sedentary behavior in preschoolers.
- The lack of systematic bias and wide agreement limits prevents correction factors for device estimates.
- The ActivPAL™ has lower practical utility due to poor accuracy and participant discomfort (skin irritation).
Purpose:
This study aimed to compare three objective measures (GT1M ActiGraph, ActivPAL™, and direct observation) of sedentary behavior in preschoolers.
Methods:
Fifty-two 4- to 6-yr-old preschoolers wore an ActivPAL™ and a GT1M ActiGraph for five consecutive days and were videotaped for 1 h during classroom activities at preschool. The area under the receiver operating characteristic curve was calculated to assess the criterion validity of the ActivPAL™ (sitting/lying, with and without standing still) and the GT1M ActiGraph (<100 counts per minute) to estimate sedentary behavior (directly observed sitting behaviors). A two-way repeated-measures ANOVA was used to define the convergent validity of the ActivPAL™ and the GT1M ActiGraph sedentary behavior estimates across the measurement days. The practical utility of the ActivPAL™ was tested in the same sample by asking the parents how their child perceived wearing the ActivPAL™.
Results:
Results indicated a poor classification accuracy for both devices (area under the receiver operating characteristic curve, 0.6) to measure sedentary behavior based on the direct observation, with and without the inclusion of standing. Time defined as sedentary behavior (sitting/lying) was lower for the ActivPAL™ compared with the GT1M ActiGraph (mean bias, 7.7%; limits of agreement, -29.01% to 13.6%). According to the parental reports, 38% of the preschoolers had skin irritation due to wearing the ActivPAL™ for consecutive days.
Conclusions:
Low classification accuracy was found for the ActivPAL™ and the GT1M ActiGraph to measure sedentary behavior in preschoolers. No correction factor can be suggested to make the sedentary estimates of the GT1M ActiGraph and the ActivPAL™ convergent as no systematic bias and wide limits of agreement were found. Furthermore, the practical utility of the ActivPAL™ was perceived to be lower compared with the ActiGraph accelerometer in preschoolers.
