Abstract
BACKGROUND: Smartwatches are increasingly used to monitor and motivate physical activity. Patients with cardiovascular disease (CVD) and peripheral artery disease (PAD) often do not meet national physical activity recommendations. They may, thus, benefit from a physical activity program using smartwatches. The Life Plus smartwatch is designed to facilitate activity monitoring by counting steps, but its validity needs to be determined, particularly in patients who may not have a normal gait, such as those with cardiovascular pathology. OBJECTIVE: This study evaluates the accuracy of the Life Plus smartwatch (versions 2 and 3) in healthy adults, patients with CVD, and patients with PAD at different walking speeds (1.8, 2.5, 3.2, and 4 km/h) and different localizations (wrists, hips, and ankles) to determine best accuracy. METHODS: In total, 34 participants, comprising healthy individuals (n=10), patients with CVD (n=14), and patients with PAD (n=10), wore 6 Life Plus watches simultaneously (3 of version 2 and 3 of version 3), located on wrists, hips, and ankles. Participants walked on a treadmill for 3-minute sessions at speeds of 1.8, 2.5, 3.2, and 4 km/h; they then performed a 10-minute free walking on the ground and again walked for 3-minute sessions on a treadmill at the same speeds. Actual step counts were recorded through video footage. RESULTS: When worn at the wrist, no significant difference between the actual number of steps and step count by version 2 watches was found in each group independently (healthy group: P=.25; CVD group: P=.50; and PAD group: P=.37). Significant differences were found with the version 2 watches at the wrist in the healthy group at 3.2 (-5.26%; P=.01) and 4 km/h (-6.13%; P=.008) and in the CVD group at 2.5 (-5.94%; P=.008), 3.2(-13.1%; P=.008), and 4 km/h (-13.96%; P=.004). When worn at the wrist, no significant difference between actual number of steps and step count by version 3 watches was found in the healthy group (P=.02) and the PAD group (P=.90). Significant differences were found with the version 3 watches at the wrist in the CVD group at 3.2 (-6.43%; P=.001) and 4 km/h (-7.3%; P=.01) and in the PAD group at 4 km/h (-5.77%; P=.04). CONCLUSIONS: For optimized counting when worn at the wrist, healthy individuals should prefer version 2 watches for slow walking (1.8 and 2.5 km/h) and version 3 for higher speeds (3.2 and 4 km/h). Patients (CVD and PAD) should prefer version 2 watches at 1.8 km/h and version 3 watches for higher speeds (2.5, 3.2, and 4 km/h).