Validity of Habitual Physical Activity: A Study of 365 Days of Pedometer Monitoring

Abstract

PURPOSE: To determine how many days of pedometer monitoring were necessary to achieve valid estimates of yearly physical activity (PA) in adults based on consecutive and random days of data collection. METHODS: Twenty-three participants (7 men, 16 women; age = 38 ± 9.9 years; body mass index = 27.7 ± 6.2 kg/m2) wore a Yamax SW 200 pedometer and recorded their step counts for 365 consecutive days. Prior to data analysis, 2% of missing values were identified and replaced by the average of remaining values of the person. Nine step- count measurement periods of various lengths (i.e., 2, 3, 4, 5, 6, 7, 14, 21 and 30 days) were randomly selected 10 times each for consecutive vs. random days from the larger data base. For consecutive days, the starting day was randomly selected from the 365 days of data for each measurement period, and for random days, all the values were randomly selected using SPSS subcommand "select cases". The year-round average was considered as a criterion measure of the "true" habitual PA. Absolute percentage error (APE) was calculated as: ((| selected measurement period step-count average -365 day step-count average |) / 365 day average) * 100. The APE mean and standard deviation for the 10 samplings were computed for both consecutive and random days. A smaller APE represents a better estimate of the "true" habitual PA; an APE of 10% was interpreted as minimally indicative of acceptable validity. RESULTS: As the length of the measurement period increased, the APE scores decreased. Overall, the APE averages for the consecutive days were larger than those of the random days. The APE averages for consecutive days ranged from 20.47% for 2 days to 9.57% for 30 days, with an average APE less than 10% at 30 days (9.57% ± 2.13%). The APE averages for random days ranged from 21.65% for 2 days to 5.40% for 30 days, with an average APE less than 10% at 14 days (8.25% ± 1.14%). CONCLUSIONS: A minimum of 30 consecutive days or 14 random days of data collection are necessary to achieve an APE lower than 10% in adults' pedometer data when a year-round average is considered the criterion measure. These findings provide researchers and practitioners useful information to determine appropriate measurement length and the method of data selection based on a desired level of validity.