There is still a need to bridge the gap in linking structural and functional alterations in order to evaluate their relative contribution and location towards clinical picture of patients with MS.
To identify how distinct covarying structural-functional patterns are able to explain clinical measures of physical disability and cognitive impairment in MS.
We applied linked independent component analysis (ICA), a type of fusion approach, to images of grey matter (GM) density, fractional anisotropy (FA) and resting-state functional MRI in patients with relapsing-remitting MS (age: 39.7±10.5 years, 60 female, disease duration: 9.4±6.9 years; median EDSS=1.5, cognitive impairment [CI]: 30%). Loading coefficients across the three MRI modalities of linked ICA were used in multiple stepwise linear regression models for EDSS, CI and Rao Battery test scores, adjusted for age and sex.
Higher EDSS was explained (R2adj=0.46, p<0.001) by a linked covarying pattern of structural damage (40%), including lower GM density (30%) and WM FA (10%) and of altered network-FC (60%). CI was explained (R2adj=0.56, p<0.001) by a linked covarying pattern of structural damage (26%), including lower GM density (18%) WM FA (8%) and especially of altered network-FC (74%). Among the different cognitive domains, attention and processing speed, as measured by symbol digit modalities test (SDMT), was best explained (R2adj =0.62, p<0.001) by a pattern mostly driven by altered network-FC (70%) and including, to a lesser extent, structural damage (30%), with equal contribution from lower GM density and lower WM FA.
The highest contribution of altered network-FC with respect to structural damage in explaining physical disability and cognitive impairment of our MS group with mild disability points out the relevance at this early disease stage of mechanisms of cortical plasticity, which however may undergo exhaustion and downregulation in the more advanced stages of disease.