Tremor analysis in two normal cohorts

Abstract

Objective: Quantitative tremor analyses using almost identical methods were compared between two independent large normal cohorts, to separate robust measures that may readily be used diagnostically from more critical ones needing lab-specific normalization. Methods: Hand accelerometry and surface EMG from forearm flexors and extensors were recorded with (500 and 1000 g) and without weight loading under postural conditions in 117 and 67 normal volunteers in two different specialty centers for movement disorders in Germany. Results: Tremor amplitude (total power) and frequency fell within a similar range but differed significantly. A significant reduction of tremor frequency under 1000 g weight load (>1 Hz), and a lack of rhythmic EMG activity at the tremor frequency in around 85–90% of the recordings were robust findings in both centers. Conclusions: The differences in frequency and total power indicate that these measures critically depend on the details of the recording conditions being slightly different between the two centers. Thus each lab needs to establish its own normative data. We estimate that at least 25 normal subjects have to be recorded to obtain normal values. The reduction of tremor frequency under load and lacking tremor-related EMG activity were well reproducible allowing a differentiation of physiological from low amplitude pathological tremor. Significance: This study provides a framework for more standardized tremor analyses in clinical neurophysiology.