Effect of ActiGraph's low frequency extension for estimating steps and physical activity intensity.
Yuri FeitoLyndsey M HornbuckleLauren A ReidScott E CrouterPublished in: PloS one (2017)
This study examined the effects of the ActiGraph's (AG) low-frequency extension (LFE) filter on steps and physical activity classification in the free-living environment. Thirty-four African-American women (age, 24.5±5.2 years; BMI, 24.9±4.5 kg/m2) had daily activity measured simultaneously with an AG-GT3X+ accelerometer and a New Lifestyles NL-800 pedometer for seven days. Steps per day (steps/day) and time (minutes/day) spent in sedentary, light, and moderate-to-vigorous physical activity (MVPA) were examined with and without the LFE filter (AG-LFE and AG-N, respectively). The AG-LFE recorded more total steps (13,723±4,983 steps/day) compared to AG-N and NL-800 (6,172±2,838 and 5,817±3,037 steps/day, respectively; p<0.001). Compared to the AG-N, the AG-LFE estimated less time in sedentary behaviors (518.7±92.1 vs. 504.2±105.4 min/day, respectively; p<0.001), and more time in light (247.7±70.4 vs. 279.1±74.7 min/day, respectively; p<0.001) and MVPA (18.9±16.9 vs. 21.5±18.2 min/day, respectively; p<0.001), respectively. These data suggest that steps and physical activity classifications will be affected when using the ActiGraph with and without the LFE filter. Future research should investigate the accuracy of these measures using the LFE filter.