Cognitive outcomes in multiple sclerosis are shaped by divergent functional connectivity trajectories

Cognitive impairment in people with multiple sclerosis (pwMS) is highly heterogeneous, highlighting the need to better understand the underlying pathophysiological processes of cognitive decline and the brain's mechanisms for adapting to MS damage. This longitudinal study explores functional connectivity (FC) changes and their relationship with cognitive trajectories over seven years in pwMS. We aimed to determine whether individuals with cognitive decline exhibited different FC patterns compared to those with stable cognitive performance. For this purpose, we analysed data from 58 pwMS, including cognitive assessments using the Rao's battery and functional MRI at two-time points with an interval of seven years. Cognitive worsening was defined as 25% decline in global cognitive scores. Graph-based networks metrics, including global and node-based strength, efficiency and clustering coefficient, alongside regional normalized grey matter (GM) volumes, were computed using MRI. We used mixed effect regression models with random subject-specific intercepts to explore FC and GM volume differences and the association between FC and cognition. The cohort was predominantly female (78%), with a mean age of 46.8 years and a median disease duration of 11.6 years. We found a significant group-by-time interaction, patients with cognitive decline showed reductions in node strength (21.1% of nodes), local efficiency (64.4%), and clustering coefficient (85.5%) particularly in the deep GM and parietal cortex at follow-up, and reduced global graph metrics. In contrast, the cognitively stable group exhibited increased node strength (15.8%) and local efficiency (5.3%), mainly in the temporal and prefrontal cortices. Both groups showed reduced GM volume in 84.2 and 79% of regions at follow-up, respectively. Several links were found between FC changes and cognitive performance. Findings confirm distinct FC trajectories in pwMS associated with their ability to cope with structural damage, impacting cognitive outcomes at mid-term. These findings indicate that patients with stable cognitive performance may engage compensatory network reorganization processes, which could mitigate the progression of cognitive decline.