31: Objectively Measuring Physical Activity in Early Childhood Using Accelerometers: Are Four Days Enough?

Studies of physical activity (PA) in young children have generally included only participants with at least four days of accelerometer wear-time. It is unclear whether four days characterizes usual PA in children 5 years and under. Investigating this protocol may inform future studies of early childhood obesity. To determine the minimal number of days, and what days, of accelerometer monitoring are needed to reliably estimate usual PA in young children. Ninety children (mean age=32 [range 4 to 70] months) were instructed to wear an Actical accelerometer for seven consecutive days and had at least four valid days. Intraclass correlation coefficients (ICC) and 95% CI were calculated for mean daily total PA and moderate-to-vigorous PA (MVPA) minutes derived from four days of monitoring. The Spearman-Brown prophecy formula was used to determine the required days of monitoring needed to achieve reliability estimates of 0.70, 0.80, and 0.90. A single day produced a reliability of ICC=0.70 (95% CI 0.58–0.79) and ICC=0.69 (95% CI 0.57–0.79) for total PA and MVPA, respectively. Spearman-Brown analyses indicated that two days are needed to achieve a reliability of 0.80. There were no differences between week-days and weekend days in mean total PA (205 (95% CI 187–222) vs. 210 (95% CI 192–228); P=0.70) or in mean MVPA minutes 26 (95% CI 20–32) vs. 29 (95% CI 23–36); P=0.45). Any two weekdays or weekend days of accelerometer monitoring can be used to assess usual PA in children 5 years and under. The use of a simpler protocol may improve compliance and feasibility for studies of PA in early childhood.