Finding the Goldilocks Zone for Toddler Accelerometry: How Many Days Are Needed for a Reliable Estimate of Physical Activity and Sedentary Time Using Machine Learning?

PURPOSE: To understand how many days and hours per day of accelerometer wear are needed for a reliable estimation of sedentary time and physical activity (PA) using machine learning (ML). METHOD: One hundred and nine toddlers wore an accelerometer on the right hip at home for 7 days. Time in sedentary time, light PA, moderate to vigorous PA, and total PA were assessed using a validated ML model for toddlers. Single-day intraclass coefficients were calculated for each minimum hours per day of wear time and each outcome. These intraclass coefficients were passed to the Spearman-Brown prophecy equation to determine the reliability of each hour per day and days combination (3-12 h, 1-10 d). RESULTS: Predicted reliabilities ranged from 0.32 to 0.98, increasing as both numbers of hours per day and number of days increased. Four days of 6 hours of wear balances a 0.7 reliability threshold while retaining 94% of participants. CONCLUSION: Our findings support the recommended 6 hours per day of wear for at least 4 days, as it balances acceptable reliability with participant retention. This recommendation is valid for this ML method, and we anticipate that it can be used to further explore sedentary time and PA in toddlers using ML advances.