Accelerometer Monitoring Duration for Reliable Estimates of Physical Activity, Sedentary Behavior, and Step Counts in Preschoolers.

Abstract

Purpose: This study aims to establish the reliability of accelerometer-based measurements for physical activity, sedentary behavior, and step counts in preschoolers and determine the optimal monitoring duration necessary for reliable data.

Methods: The study analyzed data from 434 preschoolers (ages 3-5) who wore waist-worn ActiGraph wGT3X-BT accelerometers for 7 days as part of the Sunrise Study in Spain. Two-way mixed effects intraclass correlation coefficients were calculated for various combinations of daily wear time and number of valid days to assess reliability. The Spearman-Brown prophecy formula was applied to determine the required days to achieve reliability scores of 0.7, 0.8, and 0.9.

Results: The findings showed that reliability improved with increased wear time, although intraclass correlation coefficients values were weaker around 7 to 9 hours per day. To achieve a reliability score of 0.7, moderate to vigorous physical activity required 3 to 4 days, light physical activity required 4 to 7 days, and sedentary behavior required 8 to 13 days. Steps per day required 7 to 9 days for a reliability score of 0.7, while other step outcomes needed 10 to 13 days.

Conclusion: The study provides evidence-based recommendations on the monitoring duration required for reliable estimates of physical activity, sedentary behavior, step counts, and sleep duration in preschoolers using accelerometers. These findings are crucial for researchers and practitioners aiming to accurately assess and promote physical activity in early childhood.