Correlation coefficients with motor impairment were relatively low (maximum value was found to be 0.33). The UPDRS scores assessed months before surgery are unlikely to have been completely representative of the clinical state at the time of recording due to the temporary post-operative improvement caused by the stun effect. Tykocki et al. (2013) examined the magnitude of this ‘microlesion effect’ in a large cohort of patients (n = 74), and found that total UPDRS III scores assessed within 48 h post-operatively decreased by 17.9% on average (standard deviation of 15.7%). The relatively long time period between the acquisition of the UPDRS scores and the actual recordings may have introduced additional variability in the relation with power and PAC measures. For these reasons, the actual relation between PAC and motor impairment might be stronger than we observed. On the other hand, low coefficients may indicate that constraining effects of the low-beta rhythm on HFO activity patterns might be only one of several mechanisms in which motor impairments could arise. It is also thinkable that PAC and/or power are epiphenomenal to a
