The Impact of Reducing the Number of Wearable Devices on Measuring Gait in Parkinson Disease: Noninterventional Exploratory Study.

The study was exploratory and noninterventional, so it did not test a treatment. All gait data came from a laboratory walk, which can be affected by the observer effect and heightened awareness. The wearable features were not validated against a gold-standard reference such as an instrumented walkway or motion capture. Healthy participants and participants with Parkinson disease were not age-matched, so accuracy and reliability findings in healthy adults may not apply to older adults. The single lower-back device did not reliably capture asymmetry or variability features. The single-device model did not significantly distinguish adjacent clinician gait scores such as 0 versus 1 or 2 versus 3. There were only 6 observations at MDS-UPDRS gait score 3. The paper suggests this may explain poor separation of higher scores. The On-to-Off directionality analysis involved 10 of 12 subjects. One participant with Parkinson disease who reported dyskinesia and one healthy volunteer were excluded from analysis.

Several authors are current employees of Pfizer Inc, one author is a former employee of Pfizer Inc and is currently employed by Biogen, and one author declared no competing interests. The healthy-participant study was conducted at Pfizer.

Do not conclude that one lower-back accelerometer replaces a six-device system for all gait measures. It did not reliably capture asymmetry or variability, and it did not significantly distinguish adjacent clinician gait scores such as 0 versus 1 or 2 versus 3.