Validation of different stepping counters during treadmill and over ground walking.

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Sports Science

Background:

• Physical activity trackers are increasingly used in clinical and research settings.
• The expanding market necessitates ongoing validation of these popular devices.
• Existing validation studies often lack real-world field testing.

Purpose of the Study:

• To compare the step-counting validity of activity trackers between treadmill and field test protocols.
• To determine if different activity tracker types exhibit varying accuracy in different testing environments.

Main Methods:

• Thirty healthy adults participated in treadmill tests (2, 3, 4 km/h) and a 982m field test.
• Step counts were video-recorded and compared against four devices: Polar M200, Polar A300, Dunlop pedometer, and a Samsung Galaxy S9 smartphone.
• Validity and agreement were assessed using Bland-Altman plots and Spearman's correlation.

Main Results:

• Field tests showed higher validity than treadmill tests at 4 km/h for all devices.
• The smartphone exhibited the best accuracy, validity, and agreement in both treadmill and field tests.
• Device performance was poor at 2 km/h on the treadmill; only the smartphone performed well at 3 km/h.

Conclusions:

• Step-counting validity differs between treadmill protocols and field tests.
• Future validation studies should incorporate field testing for activity trackers.
• Smartphones offer superior accuracy and lower bias in step-counting validation, particularly in field settings.