Background: Accelerometry is rapidly becoming the instrument of choice for measuring physical activity in children. However, as limited data exist on the minimum number of days accelerometry required to provide a reliable estimate of habitual physical activity, we aimed to quantify the number of days of recording required to estimate both habitual physical activity and habitual sedentary behavior in primary school children.
Methods: We measured physical activity and sedentary behavior over 7 days in 291 6- to 8-year-olds using Actigraph accelerometers. Between-day intraclass reliability coefficients were calculated and averaged across all combinations of days.
Results: Although reliability increased with time, 3 days of recording provided reliabilities for volume of activity, moderate-vigorous intensity activity, and sedentary behavior of 68%, 71%, and 73%, respectively.
Conclusions: For our sample and setting, 3 days accelerometry provided reliable estimates of the main constructs of physical activity and sedentary behavior.