Advanced age at onset is associated with distinct functional brain reorganization in multiple sclerosis

Abstract Functional network reorganization supports compensatory mechanisms in multiple sclerosis, but aging and later disease onset may alter this balance. Late-onset multiple sclerosis (age at onset ≥50 years) represents a relevant subgroup to examine how disease onset in older age is associated with functional connectivity abnormalities. In this cross-sectional study, we investigated resting-state functional connectivity alterations in late-onset multiple sclerosis versus adult-onset multiple sclerosis and their associations with clinical, cognitive, and structural MRI measures. Thirty-four late-onset multiple sclerosis, 169 sex- and disease duration–matched adult-onset multiple sclerosis, and 163 age- and sex-matched healthy controls underwent clinical, neuropsychological (Brief Repeatable Battery of Neuropsychological Tests), and MRI assessment including resting-state functional MRI. Independent component analysis quantified resting-state functional connectivity within ten large-scale networks. Between-group differences were tested using sex- and scanner-adjusted voxel-wise models (p<0.001, uncorrected, cluster extent ≥10 voxels); resulting abnormal clusters were then examined for exploratory uncorrected associations with Expanded Disability Status Scale, disease duration, cognitive performance, T2-hyperintense lesion volume, and normalized brain volume, without correction for multiple comparisons. At MRI acquisition, compared with the adult-onset multiple sclerosis group, the late-onset multiple sclerosis group was significantly older (median age: 60.5 [IQR: 56.6–66.1] versus 35.6 [IQR: 30.0–40.1] years; pFDR=0.006), showed higher Expanded Disability Status Scale score (pFDR=0.016) and higher T2-hyperintense lesion volume (pFDR=0.028) despite similar disease duration (median: 4.5 [IQR: 1.7–13.2] versus 6.0 [IQR: 2.1–11.8] years; pFDR=0.894), whereas no significant between-group cognitive difference was found. Both multiple sclerosis groups showed widespread functional connectivity abnormalities versus healthy controls, partly overlapping and partly distinct. Adult-onset multiple sclerosis displayed functional connectivity increases in the executive control and occipital default mode networks. Late-onset multiple sclerosis showed diffuse functional connectivity alterations, including decreases in the left fronto-parietal, thalamic, and occipital default mode networks. Exploratory uncorrected correlation analyses included 117 tested associations. In late-onset multiple sclerosis, decreased functional connectivity in the left fronto-parietal network related to lower normalized brain volume and worse cognition (p=0.02–0.03), whereas altered thalamic network connectivity related to higher Expanded Disability Status Scale, longer disease duration, and higher T2-hyperintense lesion volume (p=0.01-0.02). These associations did not survive correction for multiple comparisons and should therefore be considered hypothesis-generating. These findings suggest that late-onset multiple sclerosis was associated with a distinct resting-state functional connectivity profile compared with adult-onset multiple sclerosis within a cross-sectional, age-matched healthy control framework, possibly reflecting the combined contribution of multiple sclerosis-related damage, age-related vulnerability, and the more severe clinical profile of late-onset multiple sclerosis.