Background Wrist-worn wearables like the Apple Watch are widely used to monitor physical activity in both clinical and home settings. But what happens when an middle-aged to older adults walks with a rolling walker or pushes an oxygen cart? This study shows that such devices dramatically impair step detection, leading to significant undercounting and potential clinical misinterpretation. This study aimed to evaluate the accuracy of a wrist-worn device (Apple Watch Series 8) and a waist-mounted device (Omron Walking Style IV) in middle-aged to older adults using common assistive walking aids. Methods In a cross-sectional design, 42 community-dwelling middle-aged to older adults (aged 51–80) performed walking tasks under four assisted conditions (forearm crutch, rolling walker, oxygen trolley, and unassisted) and three treadmill speeds (1.61, 3.22, and 4.83 km/h). Step counts and distances recorded by each wearable device were compared to manual reference measures. Wilcoxon signed-rank tests and Bland–Altman analyses assessed accuracy and agreement. Results The Apple Watch significantly underestimated step counts, particularly with a rolling walker (−36.4%, p < .001), and at low treadmill speed (−16.3% at 1.61 km/h). In contrast, the Omron pedometer demonstrated relatively consistent performance across most conditions, although it showed notable underestimation at the lowest walking speed (1.61 km/h). A weak correlation between arm length and Apple Watch error was observed ( r = − 0.38, p < .05). Conclusion Wrist-worn activity monitors may be unsuitable for middle-aged to older adults using assistive devices due to undercounting. Waist-mounted devices offer more reliable step and distance tracking, supporting their use in clinical practice and home rehabilitation. Device selection and placement should be carefully considered in mobility-impaired populations.
ÖZEL et al. (Wed,) studied this question.