The presence of multiple sclerosis (MS) in young and middle age is often associated with poor verbal memory and attributed, in part, to white matter damage. The intersection of aging and MS is an emerging area of study, but whether disease-specific effects on verbal memory performance can be dissociated from normal aging remain understudied and inconclusive. Therefore, we examined the impact of white matter macrostructure and microstructure on verbal memory operations in older adults with and without MS.

A subsample of older adults with and without MS enrolled in a cohort study underwent 3T magnetic resonance imaging (MRI) protocols that included diffusion tensor imaging (DTI), cognitive and psychological assessment, and motoric testing. Total immediate and delayed recall on the Hopkins Verbal Learning Test-Revised (HVLT-R) served as proxies for memory performance. DTI-derived fractional anisotropy (FA) in select tracts was used to assess white matter integrity, and white matter hyperintensity volume was calculated to assess lesion load. Participants with MS were recruited from MS centers and patient registries and healthy controls recruited from the community. The current study included 136 adults age 60 and older without dementia, 64 in the MS group (mean age = 64.19+3.82; %female = 64.1) and 72 in the control group (mean age = 69.78+6.99; %female = 65.3). Participants had MS for 21.28 years on average (sd = 10.06) with generally low disease severity (Patient Determined Disease Steps = 1.92+1.76). The PROCESS macro in SPSS (model 1) was used to run separate regression models that examined the moderating effects of white matter FA and lesion volume, both whole brain and in limbic tracts, on the association of group (MS vs. control) with HVLT-R total immediate and delayed recall. Neuroimaging was preprocessed using FSL and FreeSurfer and white matter tracts were extracted using TRACULA; the tracts of interest were the bilateral fornices, ventral portion of the cingulum bundles, and uncinate fasciculi. Statistical analyses adjusted for age, years of education, and sex.

Moderation models revealed that older adults with MS showed lower total immediate recall compared to healthy controls in the context of lower FA in the right uncinate fasciculus and higher lesion volume in the bilateral fornices, left uncinate fasciculus, and bilateral ventral cingulum, and lower delayed recall in the context of lower FA in the left uncinate fasciculus and right ventral cingulum and higher lesion volume in the left uncincate fasciculus and bilateral ventral cingulum (p < .05). Moderation models with whole brain white matter measures were not significant (p > .05).

Findings extend extant research implicating white matter damage in select tracts in poor verbal memory performance older adults with MS. Verbal memory was reduced in older adults with MS in the context of reduced limbic tract white matter integrity and increased limbic tract lesion load. These results suggest that the influence of MS on memory operations in aging may be related to underlying white matter disruptions, in particular lesions, in tracts implicated in memory.