Daily‐Life, Sensor‐Derived Tremor Measures Are Sensitive to Progression in Early Parkinson's Disease
Annals of Neurology, EarlyView.
a Parkinson's disease feed — research, treatments, lived experience, in plain language
Annals of Neurology, EarlyView.
Researchers at Radboud University Medical Centre in the Netherlands asked 620 people with early-stage Parkinson's disease to wear a research smartwatch around the clock for two years — part of the large Personalized Parkinson Project. A custom algorithm analysed the raw wrist-movement data to extract two weekly measures: tremor time (how often the person was shaking) and tremor power (how intense the shaking was). These continuous, real-world readings were then compared against the standard clinical yardstick — the MDS-UPDRS (Movement Disorder Society–Unified Parkinson's Disease Rating Scale), a structured assessment done by a neurologist in clinic, typically once a year. This is an observational, longitudinal study — not a treatment trial — so it tells us what can be measured, not what can be treated.
The headline finding: among participants who had not yet started dopaminergic medication (drugs that replace or mimic dopamine), the smartwatch measures were substantially more sensitive to worsening tremor over the two-year period than annual clinical scores. Put simply, the watch picked up progression that a yearly hospital exam would miss or detect much later. In medicated participants the picture is more complex, because medication partially suppresses tremor and dose changes affect the sensor readings too — the researchers analysed these groups separately for exactly that reason. Crucially, the algorithm and its source code are freely available, and the method works on many standard research-grade wearables, not a proprietary device.
For people living with Parkinson's and their families, the immediate implication is indirect but meaningful: this kind of tool could make future clinical trials faster and smaller, because researchers could detect whether a new disease-modifying drug is working much sooner than with annual clinic visits. It could also, in the longer term, help neurologists personalise care — monitoring whether a medication dose is actually helping at home rather than only during a brief, potentially stressful clinic appointment. The technology is not yet a routine clinical tool; further work is needed to validate it across different devices and care settings. But it represents a genuine step forward in the field of digital biomarkers — objective, continuous measurements of how the disease is progressing in real life.