Related Experiment Video
Updated: Aug 8, 2026

A Method for Quantifying Upper Limb Performance in Daily Life Using Accelerometers
Published on: April 21, 2017
Comparing daily step counts between a hip and wrist-worn device in adults with Parkinson's disease
Madison S Smith1, Kimberly J Waddell2,3, Stephanie Wood2,4
1Department of Neurobiology and Anatomy, Drexel University, Philadelphia, PA, United States of America.
Introduction:
Wearable devices are increasingly used for measuring physical and physiological activity in daily life. Common appendicular symptoms of Parkinson's disease (PD), such as tremor and rigidity, could impact the accuracy of step counts when the device is worn in different body locations. Few studies have compared wrist- and hip-worn activity monitoring devices worn simultaneously in the home environment among adults with PD. The purpose of this study was to examine the agreement in daily step counts between two identical devices worn simultaneously at the wrist and hip in patients with PD.
Methods:
Twenty-one individuals with PD wore Fitbit Inspire 2 devices simultaneously on the wrist and hip during waking hours for 7 days. Total step counts from both devices were obtained for each day. The agreement between devices was examined using an intraclass correlation coefficient. A Bland-Altman plot was used to visually examine agreement.
Results:
The mean (SD) step count collected by the wrist-worn devices was 5221 (4345) and 5371 (3936) for the hip-worn devices, with an average difference of 150 steps. There was good overall agreement between the two devices with an intraclass correlation coefficient (ICC) of 0.84 (95% CI 0.78, 0.88), which was visualized with the Bland Altman plot.
Conclusions:
In an independently ambulatory PD population, wrist- and hip-worn Fitbit devices showed good agreement when measuring daily steps, with the strength of agreement varying by overall step activity. These findings support the use of wrist worn activity devices in PD and their utility for capturing physiological metrics beyond step count, such as heart rate, activity intensity, and sleep quality, which cannot be captured with a hip-worn device.

