Measurement effects of seasonal and monthly variability on pedometer-determined data

Background: The seasonal and monthly variability of pedometer-determined physical activity and its effects on accurate measurement have not been examined. The purpose of the study was to reduce measurement error in step-count data by controlling a) the length of the measurement period and b) the season or month of the year in which sampling was conducted. Methods: Twenty-three middle-aged adults were instructed to wear a Yamax SW-200 pedometer over 365 consecutive days. The step-count measurement periods of various lengths (eg, 2, 3, 4, 5, 6, 7 days, etc.) were randomly selected 10 times for each season and month. To determine accurate estimates of yearly step-count measurement, mean absolute percentage error (MAPE) and bias were calculated. The year-round average was considered as a criterion measure. A smaller MAPE and bias represent a better estimate. Results: Differences in MAPE and bias among seasons were trivial; however, they varied among different months. The months in which seasonal changes occur presented the highest MAPE and bias. Conclusions: Targeting the data collection during certain months (eg, May) may reduce pedometer measurement error and provide more accurate estimates of year-round averages.