2025 Theses Doctoral

A Novel Method for Assessing Word Finding Difficulty in Persons with Multiple Sclerosis: Semantically-Cued Retrieval of Exact Words

Dvorak, Emily

Multiple sclerosis (MS) is an inflammatory disease of the central nervous system (CNS) that results in damage to the myelin sheath that surrounds the neurons in the CNS. Resulting cognitive impairment is experienced by roughly 40% to 70% of individuals with MS. While research has historically focused on processing speed and memory deficits as the hallmarks of cognitive impairment in MS, more recent research has highlighted word finding impairment as a prevalent weakness experienced by those with MS, even early in disease. Such findings are in line with new neuroimaging research demonstrating cortical atrophy and lesion activity in language regions (e.g., left temporo-parietal regions) as part of the MS disease process.

Despite this, there remains little attention on language impairment in MS and existing research is often conflicted in quantifying deficits. These issues may stem from reliance on available expressive language metrics designed for use in aphasia and dementia populations, which are likely insensitive to the subtler “tip-of-the-tongue” phenomenon that is described by persons with MS. As such, the purpose of this dissertation was to investigate a more sensitive metric for objective word finding impairment in MS that better mirrors the word finding demands in everyday life. Furthermore, in line with research showing neuropathological changes in left temporo-parietal regions (e.g., planum temporale), this dissertation investigated if patterns of cortical thickness were associated with performance on this word finding metric and if this hypothesized relationship was mediated through differences in phonological processing ability, given previous research showing this skill as an area of weakness in persons with MS.

A sample of 172 persons with relapse-onset MS participated in objective cognitive assessment and self-report questionnaires at two timepoints. A subset of this group (n=132) also underwent 3.0 Tesla magnetic resonance imaging (MRI) at timepoint 1. The primary outcome of this study was objective language performance, quantified by both existing language metrics (e.g., verbal fluency, rapid automatized naming), and the novel paradigms of cued semantic retrieval and phonological processing. One predictor was patient-reported language complaints, assessed by the following questions: (a) having a word ‘on the tip of your tongue’ but with difficulty getting it out, (b) having a sense of what you want to say, but having trouble clearly expressing your thoughts, (c) accidentally saying the wrong word / misspeaking. Disease burden (cortical thickness) was also used as a predictor. Mood (Mental Health Inventory-5 (MHI-5)), vocabulary knowledge (NIH Toolbox Picture Vocabulary test), age, sex, and bilingualism were included as covariates.

As hypothesized, results revealed that cued semantic retrieval was the only objective language task related to expressive language difficulty, such that worse performance on cued semantic retrieval was associated with more severe/frequent expressive language complaints. Furthermore, performance on cued semantic retrieval was correlated with thinner left temporo-parietal regions of interest, specifically in the planum temporale. Lastly, this relationship between thinner planum temporale and worse cued semantic retrieval performance was significantly mediated through performance on a phonological processing task.

Taken together, results demonstrate the cued semantic retrieval is sensitive to patient-reported expressive language concerns and associated with cortical thickness patterns in relevant language regions, highlighting this task’s feasibility as an objective word finding measure. That phonological processing mediated this relationship suggests a potential mechanism for word finding impairment in persons with MS. However, because results are cross-sectional, it cannot be definitively concluded that such a relationship is disease-related, and it may be that developmental weaknesses in phonological processing serve as a risk factor for later word finding impairment in persons with MS. As such, future research should investigate this relationship longitudinally and using other neuropathological markers of disease burden (e.g., cortical lesions) to confirm that such findings are disease-related.