How Many Hours Is Enough? Optimal Accelerometer Wear Time To Reflect Daily Physical Activity

Abstract

Accelerometers are valuable objective tools used to assess physical activity (PA). Current research provides little consensus on how many hours of wear time is needed with minimum requirements ranging from 6 - 12 hours each day. PURPOSE: To identify the number of hours needed to have a valid measure of daily PA assessed by accelerometry. METHODS: Data from 119 participants (age = 41 ± 11 yrs; BMI = 27 ± 7 kg/m2) contributed approximately 2,000 days of ActiGraph GT1M accelerometer use. Five samples of 40 days were randomly selected containing exactly 10, 11, 12, 13, and 14 h/day of wear time. Utilizing a semi-simulation approach, the data pattern of the 10, 11, 12, and 13 h/day was applied in a one-to-one fashion to modify the 14 h/day data set, creating four new data sets (10, 11, 12, 13 h/day) for comparison. The 14 h/day sample was used as criterion to assess the impact of different monitor wear time. Absolute Percent Error (percent difference between two values; desirable <10%; APE) and repeated measures ANOVAs were used to assess differences in daily minutes of inactivity (<100 cts), light (100-1951 cts), moderate (1952-5724 cts), and vigorous (≥5725 cts) activity in each semi-simulated data set compared to 14 h/day. RESULTS: Mean ± SD values and APEs are presented in the table. APE ranged from 5.6 to 41.6 for the semi-simulated data sets when compared to the 14 h/day criterion. Repeated measures ANOVA results showed significant differences in daily minutes for inactivity, light, and moderate activity (all p <.05). No difference was found for vigorous activity, F (2.34,91.09) = 2.06, p =.13. CONCLUSIONS: These data showed only 13 h/day had APEs below 10%. This suggests a minimum accelerometer wear time of 13 hrs to provide a valid measure of daily PA.TABLE