Figure 4.

Relationship between tremor amplitude modulation during stimulation at near postural tremor frequency (f_T) and the degree of tremor entrainment. Although tremor suppression was inversely proportional to tremor entrainment, tremor amplification did not show any dependency, possibly because tremor amplitude was already at ceiling values and could not be amplified further. Linear regression fits (thick grey lines) and their 95% confidence limits (thin grey lines) are shown in each panel. (A) Lower panel (in blue): tremor suppression showed a trend towards dependency on tremor entrainment at the tremor axis that showed maximal change in tremor amplitude (F-statistic P = 0.0803, R² = 0.3336). Upper panel (in red): in contrast, tremor amplification did not show any dependency on tremor entrainment at the same axis (F-statistic P = 0.9155, R² = 0.0015). (B) This difference between suppression and amplification was confirmed when changes across all three axes of the tri-axial accelerometer were considered. Lower panel: tremor suppression was inversely proportional to tremor entrainment (F-statistic P = 1 × 10⁻⁴, significant following FDR correction, R² = 0.4171), confirming the relationship between amplitude suppression and entrainment. Upper panel: tremor amplification did not show dependency on tremor entrainment (F-statistic P = 0.4089, R² = 0.0245).