period (T4 to T5), no significant PAC increase was found in either controls or patients (Fig. 5E). Notably, the absolute PAC values did not differ between patients and controls in any of the 5 periods (Wilcoxon rank-sum, all P > 0.2). This finding showed less PAC modulation in patients during selected periods of the slow tapping cycle. We subsequently tested the hypothesis that the magnitude of the PAC change around movement onset determines the ability to recruit muscles engaged in the tapping rapidly and thus may contribute to the motor impairment in slow tapping performance in patients. Although the EMG slope was correlated with the PAC change between T1 and T2 in patients, this correlation was lost when the computation of PAC was corrected for the shorter averaged duration of T2 across subjects. EMG slope and PAC change were not significantly correlated in controls (ρ = −0.12, P = 0.601).
