Aperiodic subthalamic activity predicts motor severity and stimulation response in Parkinson disease

Abstract

IntroductionRhythmic beta activity in the subthalamic nucleus (STN) local field potential (LFP) is associated with Parkinson disease (PD) severity, though not all studies have found this relationship. We investigated whether aperiodic ‘noise’ elements of LFP, specifically slope of the 1/f broadband, predict PD motor symptoms and outcomes of STN-DBS. MethodsWe studied micro-LFP from 19 PD patients undergoing STN-DBS, relating the aperiodic 1/f slope and the periodic beta oscillation components to motor severity using the UPDRS-III and improvement with DBS at 1 year. ResultsBeta power, not 1/f slope, independently predicted baseline UPDRS-III (r = 0.425, p = 0.020; r = −0.434, p = 0.032, respectively), but multiple regression using both predicted better (F (2, 16) = 6.621, p = 0.008, R2 = 0.453). Only multiple regression using both slope and beta power predicted improvement in UPDRS-III at 1 year post-operatively (F (2, 15) = 6.049, R2 = 0.446, p = 0.012). ConclusionsBoth beta synchronization and slope of the 1/f broadband are informative of motor symptoms in PD and predict response to STN-DBS.