A multifaceted approach to understand cognitive impairment in MS: exploring the nonlinearity of cognition

Abstract

Cognitive impairment is a frequent but often invisible consequence of multiple sclerosis (MS), affecting memory, attention, and information processing speed. These deficits strongly impact daily life, yet are difficult to capture, as neuropsychological test results and patients’ self-reports often do not align. The overarching aim of this thesis was to improve the early detection, measurement, and understanding of cognitive impairment in MS through a multifaceted approach. We investigated biomarkers of neuro-axonal damage, focusing on serum neurofilament light (NfL), in combination with MRI-derived measures of brain atrophy. Both markers provided early signals of cognitive decline, although limited specificity remains a challenge. We then studied brain networks, showing that patients with cognitive impairment exhibit increased structure–function coupling and less flexible, less stable network configurations. Next, we explored the heterogeneity of cognitive impairment. Using large-scale neuropsychological testing, we identified six cognitive profiles, ranging from preserved function to broad impairments. Information processing speed emerged as the most sensitive marker to identify affected patients, while visuospatial memory was particularly useful in distinguishing between profiles. Longitudinal analyses confirmed the predictive value of slowed processing speed for future decline.Furthermore, we applied a symptom network approach, demonstrating that objective test performance and subjective complaints form distinct but interconnected clusters, with strong links to fatigue and mood. Finally, we co-developed the MS-IADL-Q, a questionnaire that captures how cognitive impairment affects complex daily activities, independent of physical disability. Together, these findings highlight the nonlinearity and multifaceted nature of cognition in MS, and underscore the need for integrative, patient-centered assessment strategies.