Beyond task activation : a consistent vigour network in Parkinson’s disease

Motor vigour refers to the speed and force with which an action is initiated and performed. It can vary from trial to trial, even when the same person performs the same action. In Parkinson’s disease (PD), dopaminergic loss in basal-ganglia circuits is related to bradykinesia and less consistent motor output. PD therefore provides a useful model for studying motor vigour variability. However, behavioural measures alone cannot show whether this variability reflects an unstable neural vigour-related command, or whether a more stable command becomes variable as it is translated into movement. Testing these possibilities requires measurement of trial-wise neural activity. To address this question, we analysed single-trial fMRI data from $18$ participants with PD who performed a ballistic hand-squeeze task in OFF- and ON-medication sessions during subthreshold galvanic vestibular stimulation (GVS). Trial-wise BOLD responses were estimated using GLMsingle and entered into a constrained optimisation framework. This framework was designed to identify a "vigour network" that was task-engaged, behaviourally linked, spatially coherent, and stable across consecutive trials. The activity of this network was related to behavioural vigour, but varied less across consecutive trials than behaviour. The network included motor, striatal, cerebellar, thalamic, parietal, temporal-limbic, and orbitofrontal regions. Somewhat surprisingly, the vigour network showed limited overlap with the conventional task-activation map, i.e., voxels defined as task-activated. The vigour network was also sensitive to medication and GVS: medication increased local connectivity within the network, and the same GVS waveform that changed behaviour also altered the vigour-network projection -- features not seen in the task-related network. In summary, these findings suggest that some behavioural variability may arise after the vigour-related network initiates movement.