Are newer versions of ActiGraph accelerometers (GT3X+) comparable with older models (7164) in young people, and does the low frequency extension make a difference?

Purpose: Data comparability is an important question in longitudinal studies using different models of accelerometers to measure physical activity. Although studies have been done with adults to compare 7164 and GT3X+ Actigraph accelerometers, comparability in youth has not been demonstrated. The purpose of this study was to determine whether the 7164 is comparable with the GT3X+ using the normal (N) and low frequency extension (LFE).

Methods: 57 youth (54.4% female) between ages 5-17 years (M=10.66; SD=3.7) wore a 7164 and a GT3X+ accelerometer simultaneously for three consecutive days. Accelerometers were side-by-side, attached to a belt, and collected data in 30-second epochs. GT3X+ were downloaded using normal and LFE extension. Average daily step counts and minutes per day in nonwear, sedentary, light, moderate, vigorous, and MPVA were calculated.

Results: Means for 7164 and GT3X+N were significantly different for all 7 comparisons (p<.05). Largest differences were in sedentary and light activity, with GT3X+N showing more sedentary and less light than the 7164 (+17.9 and -19.4 min/day, respectively). The GT3X+N also recorded 9.1 fewer MVPA min/day and 2721 fewer steps per day. Applying LFE to GT3X reduced differences in light and sedentary to about 10 min/day (p=0.14 and p=0.00, respectively) and in MVPA to 5.5 min/day (p=0.00). However, LFE recorded 5273 more steps per day compared to the 7164.

Conclusions: To reduce the differences found between old and new generations of ActiGraph, studies should employ the LFE for more comparable results. Comparing step counts across models and filters should be done with caution.