Abstract
The ActiPed (FitLinxx) is a uniaxial accelerometer, which objectively measures physical
activity, uploads the data wirelessly to a website, allowing participants and researchers
to view activity levels remotely. The aim was to validate ActiPed’s step count, distance
travelled and activity time against direct observation. Further, to compare against
pedometer (YAMAX), accelerometer (ActiGraph) and manufacturer’s guidelines. 22 participants,
aged 28±7 years, undertook 4 protocols, including walking on different surfaces and
incremental running protocol (from 2 mph to 8 mph). Bland-Altman plots allowed comparison
of direct observation against ActiPed estimates. For step count, the ActiPed showed
a low % bias in all protocols: walking on a treadmill ( − 1.30%), incremental treadmill
protocol ( − 1.98%), walking over grass ( − 1.67%), and walking over concrete ( − 0.93%).
When differentiating between walking and running step count the ActiPed showed a %
bias of 4.10% and − 6.30%, respectively. The ActiPed showed >95% accuracy for distance
and duration estimations overall, although underestimated distance (p<0.01) for walking
over grass and concrete. Overall, the ActiPed showed acceptable levels of accuracy
comparable to previous validated pedometers and accelerometers. The accuracy combined
with the simple and informative remote gathering of data, suggests that the ActiPed
could be a useful tool in objective physical activity monitoring.
Key words
accelerometer - motion sensor - accuracy - steps
