Calibration of accelerometer output for children.

Area of Science:

• Pediatric Exercise Science
• Biomedical Engineering
• Kinesiology

Background:

• Accurate assessment of physical activity in children and youth is vital for designing effective interventions.
• Objective methods like accelerometers are preferred over self-report due to recall bias in younger populations.
• Existing calibration methods for accelerometers in children show considerable variability.

Purpose of the Study:

• To review the calibration of commonly used accelerometers (ActiGraph, Actical, Actiwatch, RT3) in children and youth.
• To examine regression modeling approaches for calibrating accelerometers against measured energy expenditure.
• To present findings on energy expenditure estimates and activity intensity cut-points across studies.

Main Methods:

• Systematic review of studies calibrating accelerometers using direct energy expenditure measurements.
• Analysis of regression modeling techniques employed in accelerometer calibration.
• Compilation of energy expenditure point estimates and count ranges for different activity intensities.

Main Results:

• Significant variation exists in accelerometer count cut-points defining activity intensities (e.g., 3 and 6 METs) across studies.
• Inconsistencies persist despite similar testing protocols including walking, running, and free-living activities.
• Calibration approaches differ, impacting the comparability of physical activity and sedentary behavior data.

Conclusions:

• Standardized best practices for accelerometer calibration in pediatric populations are needed.
• Alternative data processing using raw acceleration signals may offer a more robust method for activity characterization.
• Further research should focus on refining calibration techniques to improve the accuracy and reliability of objective physical activity measures in children.