Supplemental material Supplemental Data

Figure S1 (related to Figure.1). Under tremor-frequency stimulation the frequency of the peripheral resting tremor self-modified and settled on a new frequency in 9 of our 12 patients from Experiment 1. This frequency remained fixed for the duration of the stimulation session. Diagram shows a sample spectrogram of the primary accelerometer channel as stimulation is imposed. Note the change in frequency from 4.8 Hz (dashed white line) to 4.2 Hz (solid white line). Figure S2 (related to Figure.2). (A) Phase-difference histograms between tremor and stimulation signals normalised as probability distributions. Labels correspond to patient number, followed by their derived phase-synchronisation index (PSI), where a PSI of 0 represents no entrainment and a 1 represents complete entrainment. Some patients display little to no phase-entrainment, whilst others show preponderance to certain phase-alignments. Note however that entrainment is modest and rarely approaches 0.1. Several patients show asymmetries in their phase-alignment histograms. (B) Regression analysis reveals that PSI is a reasonable predictor of stimulation efficacy (as quantified by the range of the drift profile), displaying a positive correlation (R2=0.34, p<0.05). However, entrainment capacity did not significantly impact on the frequency-shift observed during tremor-frequency stimulation (R2=0.25, p=NS; FDR corrected). (C) Group analysis relating the strength o