Evaluation Of The Actigraph’s ‘Low-Frequency Extension’ In The Free-living Environment

The ActiGraphTM is the most commonly used research grade physical activity monitor. However, the effect of the ‘low-frequency extension’ (LFE) on step counts in a free-living environment has not been widely studied.

Purpose We sought to determine the accuracy of the ActiGraph’s LFE during freeliving activities in two versions of the device (GT1M and GT3X).

Methods Twenty-four participants wore the StepWatch (SW) and both versions of the ActiGraph with (GT1M-LFE and GT3X-LFE) and without (GT1M-N and GT3X-N) the LFE activated on different days. The SW served as the criterion method and all statistics were performed on the percent differences between the devices [(Measured-Actual)/Actual].

Results Repeated measures ANOVA showed that both devices with the LFE ‘on’ (GT1M-LFE and GT3X-LFE) overestimated steps taken (30% ± 40% and 34% ± 31%, respectively; P < 0.001); however, without the LFE turned ‘on’ (GT1M-N and GT3X-N) actual steps were significantly underestimated (-27% ± 20% and -23% ± 14%, respectively; P < 0.001).

Conclusions Our results revealed that the increased sensitivity provided by the ActiGraph’s LFE overestimates steps taken throughout the day. Meanwhile, failure to turn the LFE ‘on’ underestimates the total number of steps compared to a criterion method. Researchers using the ActiGraphTM should be aware of these limitations when measuring activity levels in the free-living environment