Related Experiment Video
Updated: May 2, 2026
Visualization of Intensity Levels to Reduce the Gap Between Self-Reported and Directly Measured Physical Activity
Published on: March 7, 2019
Comparison of physical activity and sedentary time measured with the ActiGraph GT3X-BT and Actical accelerometers
Janine Clark1, Aidan Gribbon1, Mélie St-Laurent1
1Centre for Health Data Integration and Direct Measures, Statistics Canada.
Background:
Accurate and ongoing assessments of physical activity (PA) and sedentary time (SED) are needed to support public health surveillance, evaluate interventions, and advance the understanding of how movement behaviours relate to health. After six cycles of data collection (2007 to 2019) using the Actical (AC) accelerometer, the Canadian Health Measures Survey (CHMS) transitioned to the ActiGraph wGT3X-BT (AG). To understand how estimates from the AC accelerometer may compare with those from the AG in the context of the CHMS, this study compares AC and AG accelerometer estimates of PA, step counts, and SED using CHMS protocols.
Methods:
A convenience sample of 47 adults (aged 18 to 79 years) and 36 children and youth (aged 3 to 17 years) wore both AC and AG accelerometers on their waist for seven consecutive days. Estimates of PA and SED, step counts, and the percentage of those meeting PA recommendations were compared between the devices using descriptive, correlation, and agreement statistics.
Results:
Agreement ranged from poor to excellent, with variability across PA intensities and age groups. Significant absolute differences in SED and light PA (LPA) were observed across all age groups, and in step counts among children and youth. Agreement was good to excellent across most age groups for moderate-to-vigorous PA (MVPA), and among adults for step counts. While the percentage of those meeting PA recommendations was higher with the AG, results were not statistically different. Similar comparisons could be made with the AG device when using the normal and low frequency extension filters.
Interpretation:
The results of the present study provide data users and researchers with an indication of the expected differences between the devices across various movement behaviour outcomes in the context of the CHMS. Results suggest that comparisons between cycles 1 to 6 and Cycle 7 onward of the CHMS for MVPA are acceptable, but they should be carried out with caution. Comparisons of SED, LPA, vigorous PA, and step counts are not recommended.
