Real-world gait study of Parkinson's disease using wearable sensors: a systematic review.

The search covered only a small set of databases and excluded grey literature, so relevant work may have been missed. The studies differed so much in design, sensors and outcome definitions that no pooled estimate could be produced, which is why the review reports themes rather than an effect. Most included studies were small, and the most common quality problems were inadequate power and unclear external validity, so the participants may not represent the wider Parkinson's population. More than half the studies (54%, 34 of 63) did not report any check that participants actually wore the sensors during home recording, which can leave unreliable or incomplete data. No consensus emerged on sensor type, body placement, sampling frequency or how walking bouts were defined, so recordings from different studies are not comparable. The review is limited to articles published between September 2013 and September 2023 in the databases searched.

The paper names its funders: the Agence de l'innovation de Defense, Sysnav, Ecole Normale Superieure Paris Saclay, and the French Defense Health Service (SSA). No separate conflict-of-interest declaration appears in the text supplied.

Do not read this as evidence that wearable gait monitoring is ready for routine clinical use. The review could not combine the 63 studies because they used different devices, sensor placements and analysis methods, and no agreed way of computing the measures exists.