Back

Phonological Processing is Below Expectations and Linked to Word-Finding Difficulty in Multiple Sclerosis

Sumowski, J. F.; Dvorak, E.; Levy, S.; Anderson, J. R.

2023-09-06 neurology
10.1101/2023.09.05.23295080 medRxiv
Show abstract

BackgroundWord-finding difficulty is prevalent but poorly understood in persons with relapsing-remitting multiple sclerosis (RRMS). ObjectiveInvestigate our hypothesis that phonological processing ability is below expectations and related to word-finding difficulty in patients with RRMS. MethodData were analyzed from patients with RRMS (n=50) on patient-reported word-finding difficulty (PW-WFD) and objective performance on WIAT-4 Phonemic Proficiency (PP; analysis of phonemes within words), Word-Reading (WR, proxy of premorbid literacy and verbal ability), and Sentence Repetition (SR; auditory processing of word-level information). ResultsPerformance (mean [95% CI]) was reliably lower than normative expectations for PP (- 0.41[-0.69, -0.13]) but not for WR (0.02[-0.21, 0.25]) or SR (0.08[-0.15, 0.31]. Within-subjects performance was worse on PP than both WR (t[49]=4.00, p<0.001, d=0.47) and SR (t[49]=3.76, p<0.001, d=0.54). Worse PR-WFD was specifically related to lower PP (F2,47=6.24, p=0.004, {eta}2=0.21); worse PP performance at PR-WFD Often (n=13; -1.16[-1.49, -0.83]) than Sometimes (n=17; -0.14[-0.68, 0.41]) or Rarely (n=20; -0.16[-0.58, 0.27]. PR-WFD was unrelated to WR or SR (Ps>0.25). ConclusionPhonological processing was below expectations and specifically linked to word-finding difficulty in RRMS. Findings are consistent with early disease-related cortical changes within the posterior superior temporal / supramarginal region. Results inform our developing model of MS-related word-finding difficulty.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.