Tremor Amplitude as a Digital Biomarker

Tremor is involuntary rhythmic oscillation of a body part, and amplitude is how large that oscillation is. Clinically it has always been graded by observation on ordinal rating scales, where a clinician watches a patient hold a posture or perform a task and assigns a score from zero to four. Motion sensors replace that judgement with a physical quantity: displacement or angular velocity, recorded continuously, alongside the frequency of the oscillation and the proportion of time tremor is present.

The difference is not only precision. An ordinal scale has four or five usable steps, which puts a floor on how small a treatment effect a trial can detect. A kinematic measurement is continuous, so it can resolve change that a rating scale rounds away, and it can be collected while the person goes about a normal day rather than only while they are being watched.

Amplitude is derived from accelerometer and gyroscope signals, usually worn on the wrist or the affected limb. The signal is band pass filtered around the tremor frequency band, conventionally three to twelve hertz depending on tremor type, to separate oscillation from voluntary movement, and amplitude is then computed within that band. Because tremor type matters, classification guidance from the International Parkinson and Movement Disorder Society distinguishes rest, postural and kinetic tremor, and each is elicited by a different condition, which is why structured tasks are still used alongside passive recording.

Validation has largely been done against clinical rating scales. Sensor derived amplitude correlates well with those scales while resolving finer gradations, and dedicated systems for structured in clinic and at home tremor assessment have been evaluated in essential tremor and Parkinson's disease.

The main uses are treatment response and progression. In essential tremor and Parkinson's disease, tremor amplitude quantifies the effect of medication, deep brain stimulation and focused ultrasound with sensitivity a rating scale cannot match. In drug trials, continuous measurement also captures how amplitude varies across a dosing cycle rather than only at the moment of assessment.

Passive monitoring adds something separate: it describes how much tremor a person actually experiences during ordinary activity, which is closer to the burden they report than a scored task is. These measures are reported alongside disease specific instruments such as the Hoehn and Yahr stage and the Parkinson's Disease Questionnaire, since the sensor measures the movement while the questionnaire measures what that movement costs the person.

Amplitude depends heavily on the task and posture. A resting tremor and a postural tremor recorded in the same person produce different numbers, so passive recordings must be interpreted with context or they conflate conditions that clinicians deliberately separate.

Sensor placement changes the value, and wrist measurement does not represent tremor elsewhere in the body. Voluntary movement can be misclassified as tremor and the reverse, particularly during activity. Most importantly, the amount of change in amplitude that matters to a person has not been established for most contexts, so a statistically detectable reduction is not automatically a clinically meaningful one. The tremor rating scales most trials still report as their primary endpoint are not in this library, which is a real gap in the instrument bridge for this measure.

• Bhatia KP, et al. Consensus statement on the classification of tremors, from the task force on tremor of the International Parkinson and Movement Disorder Society. Mov Disord. 2018. pubmed.ncbi.nlm.nih.gov
• Elble R, et al. Task force report: scales for screening and evaluating tremor: critique and recommendations. Mov Disord. 2013. pubmed.ncbi.nlm.nih.gov
• Mostile G, et al. Correlation between Kinesia system assessments and clinical tremor scores in patients with essential tremor. Mov Disord. 2010. pubmed.ncbi.nlm.nih.gov
• Elble RJ, et al. Tremor rating scales and laboratory tools for assessing tremor. J Neurol Sci. 2022. pubmed.ncbi.nlm.nih.gov