Introduction. Parkinson's disease (PD) requires objective and continuous monitoring of motor symptoms. Wearable sensors are a promising tool for improving diagnostic accuracy and monitoring disease dynamics. However, they are underutilized in clinical practice due to the lack of uniform standards and limited data reproducibility. The presented study fills this gap. The objectives of the research are to analyze current approaches to the use of wearable systems for monitoring motor symptoms of PD, identify limitations (including those related to validation standards), and determine ways to overcome them for the efficient use of sensors in clinical practice.

Materials and Methods. Using the Prisma 2020 methodology, a literature search was conducted for the years 2020–2025 in PubMed, Scopus, Web of Science, and Elibrary.ru. Peer-reviewed studies on the development, validation, and application of wearable devices for assessing gait, tremor, bradykinesia, and dyskinesia were examined. Nine key terms in digital medicine and neurodiagnostics in Russian and English were used for the search: “Parkinson's disease”, “digital biomarkers”, “wearable devices”, and others. The final sample of 48 studies was dominated by meta-analytics (31%) and clinical studies (29%). Nineteen percent of the sources discussed the development of monitoring systems, 15% were longitudinal studies, and 6% were systematic reviews.

Results. Descriptions of nine wearable devices for monitoring motor performance in patients with PD were compared. The types of metrics, clinical scenarios, and tasks were considered. Two concepts of the devices under study were outlined:

These two cases required different labeling standards, analysis windows, and clinical significance criteria. To improve the comparability of results, a “minimum validation set” specific to the class of problems was needed. Conditions for overcoming these contradictions were:

Discussion. The widespread use of wearable devices for analyzing motor symptoms in patients with PD is hampered by a lack of analytical and clinical validation standards and economic ambiguity of implementation. In general, five of the devices reviewed show promise. However, clinical data on their efficiency and impact on quality of life are insufficient, as research is primarily focused on the potential of the concepts (accuracy of the algorithms) rather than the practical value and readiness for everyday use of real devices. There is little research on external (multicenter) transferability, unified endpoints, and clinical utility.

Conclusion. Current data on the capabilities and limitations of wearable sensors for Parkinson's disease has been systematized. Widespread adoption of such devices is impossible without standardization and unified criteria for efficiency, safety, and economic viability. Addressing these identified challenges will transform approaches to diagnosis and treatment, making wearable systems a key tool in personalized medicine.