Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Precision fMRI reveals task-dependent reductions in language lateralization in dyslexia
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
Ola Ozernov-Palchik1,2, Amanda O'Brien1,4, Somya Mittal1,4, Tiffany Hogan3, John D.E. Gabrieli1; 1MIT, 2Boston University, 3MGH, 4Harvard Universuty
Language processing draws on a specialized fronto-temporal network that is strongly left-lateralized by early childhood (Ozernov-Palchik, O’Brien et al., 2026). Learning to read relies on the integration of the visual system with this pre-existing language network, allowing print to engage spoken-language representations (Saygin et al., 2016; Yeatman et al., 2023). Atypical language lateralization has been one influential theory of developmental dyslexia, with reduced left-hemisphere (LH) dominance and compensatory right-hemisphere (LH) recruitment proposed as possible neural mechanisms (Galaburda et al., 1985; Shaywitz et al., 2002; Richlan et al., 2011). However, prior lateralization findings have been mixed, perhaps because group-averaged analyses obscure individual variability in language-responsive cortex, and because tasks differ in linguistic, phonological, visual, and executive demands. Here, we asked whether dyslexia is associated with global alteration of language-network lateralization or with selective lateralization differences that emerge under reading-relevant task demands. We used precision fMRI to define language-responsive functional regions in individual third- and fourth-grade children (mean age=9.0) with dyslexia (Dys) and typical (Typ) readers, a period when children transition from effortful decoding toward fluent reading. Participants completed a resting state scan and several in-scanner tasks: auditory language localizer, visual word with rhyme judgment, semantic judgment, and face matching. Partially overlapping samples contributed to analyses of auditory language comprehension (Intact>Degraded speech; Typ: N=16; Dys: N=36), visual phonological processing (Rhyme>Face; Typ: N=20; Dys: N=32), visual semantic processing (Semantic>Face; Typ: N=20; Dys: N=32), and resting-state connectivity (Typ: N=17; Dys: N=28). Across five bilateral frontal and temporal language regions (Lipkin et al., 2022), we tested lateralization using mixed-effects models predicting response magnitude from hemisphere, group, and their interaction, controlling for age, sex, nonverbal IQ, and motion. We found no evidence for dyslexia-related reduction in lateralization during auditory language comprehension (F(1,441)=1.24, p=.27); both groups showed robust LH>RH responses. This null result converged with an independent older cohort (mean age=13.4; Typ: N=84; Dys: N=18). In contrast, during visual phonological processing, children with dyslexia showed reduced lateralization relative to typical readers (F(1,466)=17.89, p<.001, d=.82). Semantic > Face showed inconclusive evidence for reduced lateralization, and the direct Semantic > Rhyme contrast showed no evidence for a left-hemisphere group difference. Thus, lateralization differences were not driven by semantic-only demands once other demands were controlled, but emerged when print needed to be mapped onto phonological representations. Hemisphere-specific analyses indicated that reduced lateralization was driven primarily by lower LH responses in dyslexia (t(50)=2.04, p=.047), with no evidence for increased RH recruitment. Resting-state analyses revealed convergently reduced LH within-network connectivity in dyslexia (F(1,40)=5.32, p=.026), again without an RH group difference. Finally, LH inferior frontal response magnitude was associated with reading comprehension in typical readers but not dyslexia, suggesting altered coupling between LH language-network engagement and reading outcomes. These results suggest that dyslexia is not characterized by global disruption of language-network lateralization. Instead, reduced lateralization emerges under print-based phonological demands and reflects reduced LH organization rather than compensatory RH recruitment.
Topic Areas: Disorders: Developmental, Reading