Multiple sclerosis (MS) is a disease of the central nervous system affecting 2.8 million people worldwide. It is characterized by the development of focal white matter (WM) lesions and neurodegeneration, which lead to a progressive accumulation of permanent disability, both in motor and cognitive functions (Walton et al., 2020). One of the most common and disabling consequences of MS is gait impairment (LaRocca, 2011), which is present since the early stages and worsens over time (Baird and al., 2019). Difficulties in walking may arise due to a multitude of symptoms, including muscle weakness, sensory disturbances, spasticity and fatigability (Spain et al., 2012). Given the relevance of deep grey matter (GM) nuclei in gait control and their extensive involvement in MS, it is important to better understand the relation between their integrity and walking dysfunction in pwMS. In this perspective, a study that has examined the association between gait impairment and subcortical GM volumes in pwMS found that the volume of the globus pallidus (GP) was the strongest predictor of walking performance (Motl et al., 2015). Further evidence of involvement of the GP in MS was recently highlighted by another study that reported increased T1/T2 ratio values, a measure sensitive to myelin and iron concentration, in this nucleus in pwMS compared to healthy controls (HC), which were associated to more severe disability and longer disease duration (Margoni et al., 2022). Also, exercise treatment has been shown to elicit morphological adaptations of the GP both in pwMS (Feys et al., 2019) and in healthy older adults (Niemann et al., 2014).

As part of the basal ganglia network, the GP is not only involved in gait control, but it also regulates cognitive and emotional behaviours (Lanciego et al., 2012). It is crossed by a large number of myelinated axons, which account for its typical pale appearance, and it is divided by the medial medullary lamina (MML) into two portions, the external (GPe) and internal GP (GPi), which are involved differently in the functional pathways – direct and indirect – of the basal ganglia circuit. While both nuclei receive inputs from similar structures, mainly the dorsal striatum and the subthalamic nucleus, the GPe has been classically viewed as an intrinsic nucleus, which relays afferent signals incoming from other subcortical areas, whereas the GPi has been considered as the final output station of inhibitory projections towards the thalamus (Calabresi et al., 2014). However, recent studies have suggested that the GPe is likely to have a more central role than previously thought, as it contributes to communication between the two pathways and is highly connected to several cortical areas (Grewal et al., 2018; Courtney et al., 2023).

Advanced magnetic resonance imaging (MRI) techniques allow not only to identify the GP, but also to visualize the MML (O'Gorman et al., 2011; Deistung et al., 2013; Maruyama et al., 2019) and separate the GP into GPi and GPe (Ewert et al., 2018; Pauli et al., 2018; Solomon and al., 2021). Several studies have explored this possibility in the field of Parkinson's disease, mainly due to the importance of the GPi as a target for deep brain stimulation (Maruyama et al., 2019).

We believe that exploring the relation between damage of the individual GP components and measures of clinical disability, motor and cognitive function in MS is relevant, as it would improve our understanding of the factors underlying subcortical network dysfunction and its eventual collapse. To this aim, we 1) assessed whether there are unique structural alterations of the GPe and GPi, in terms of volume, iron content and microstructural integrity, in pwMS compared to HC; 2) evaluated resting state (RS) functional connectivity (FC) differences of the GPe and GPi in pwMS compared to HC; and 3) analyzed clinical correlates of GP MRI abnormalities in pwMS.