Long-term consistency of aperiodic and periodic physiomarkers in subthalamic local field potentials in Parkinson’s disease

Abstract

Beta oscillations (±13–35 Hz) and aperiodic spectral features extracted from local field potential (LFP) recordings have been identified as promising physiomarkers for adaptive deep brain stimulation (aDBS) in Parkinson’s disease. However, the long-term consistency of these signal features across behavioural and clinical conditions remains unclear. Bilateral subthalamic nucleus LFPs were recorded from twelve patients with an average inter-recording interval of 137 days, during rest, a finger-to-nose task and speech, with stimulation switched off and on. Intra-class correlation coefficients indicated moderate between-visit consistency for aperiodic offset and exponent but good to excellent consistency of beta peak power. Most aperiodic and power changes induced by task execution and stimulation were statistically comparable across visits. Results remained inconclusive regarding the properties of beta peaks exhibiting the strongest power suppression post-stimulation. Our findings support the potential of beta peak power as primary physiomarker for aDBS, with aperiodic components as possibly suitable supplementary markers.