Slide Slam

< Slide Slam Sessions

Slide Slam Q14

Neuroanatomical correlates of canonical and noncanonical sentence processing in the aging brain

Slide Slam Session Q, Friday, October 8, 2021, 12:00 - 2:30 pm PDT Log In to set Timezone

Nicholas Riccardi1, Rutvik Desai1, Sarah Newman-Norlund2, Samaneh Nemati2, Sara Sayers2, Roger Newman-Norlund1, Julius Fridriksson2; 1University of South Carolina, Department of Psychology, 2University of South Carolina, Department of Communication Sciences and Disorders

Introduction: Healthy aging is associated with subtle declines in language abilities that have been linked to structural and functional changes in the brain. Age disproportionately impacts the processing of syntactically complex sentences, providing a useful model to study the neuroanatomical structures associated with processing different syntactic constructions. Here, we examined the relationship between grey matter volume (GMV) and resting-state fMRI (rsfMRI) with sentence processing performance of four different syntactic structures in healthy older adults. Methods: Sixty participants (40 Female, 20 Male) between the ages of 60 and 80 (M = 66.78, SD = 6.98) were recruited as part of the Aging Brain Cohort (ABC@UofSC). Participants completed a sentence repetition task containing 30 recorded sentences (5 active, 5 passive, 10 canonical [object-subject or subject-subject], and 10 non-canonical [object-object or subject-object]). Sentences were presented one-by-one and participants were asked to repeat each sentence. Accuracy was measured as percent of words correctly repeated in order. Sentences were matched for number of syllables, words, word frequency, and semantic density. Subsequently, structural and rsfMRI scans were collected. We computed correlations between performance on each sentence type and GMV within language regions known to be associated with sentence processing including the inferior frontal gyrus pars opercularis and triangularis (IFGoper, IFGtri), supramarginal and angular gyri (SMG, AG), and posterior middle and superior temporal gyri (pMTG, pSTG). The relationship between rsfMRI and sentence type was analyzed using multivariate support vector regression (SVR) and proportional amplitude of low frequency fluctuations (pALFF) within all 110 grey matter regions of the Johns Hopkins University atlas. Analyses controlled for age and were conducted for each sentence type individually, as well as directly comparing active vs. passive and canonical vs. noncanonical using nuisance regression. VBM clusters were FWE corrected p < .05, and SVR used permutation correction (1000 permutations, p < .05). Results: Participants were significantly less accurate at repeating active compared to passive sentences, p = .001. Canonical and noncanonical sentence repetition accuracies were not significantly different, p = .5. Poorer performance on active sentences was associated with reduced GMV in right SMG and IFGtri. Repetition accuracy for active sentences was correlated with greater pALFF in a mostly right-lateralized medial-temporal and parietal areas (e.g., pMTG, AG, entorhinal cortex, and parahippocampal gyrus). Results remained significant after controlling for passive sentences, and no GMV or SVR results were predictive of passive sentences individually. Repetition accuracy for canonical and noncanonical sentences was associated with increased GMV in left SMG and bilateral IFGtri. Analysis of pALFF revealed partially dissociable areas associated with accurate repetition of canonical (right-lateralized fronto-temporal regions, e.g., insula, temporal pole) and noncanonical (bilateral posterior temporoparietal areas, e.g., AG, pSTG) sentences. Conclusions: Overall, our results suggest that brain areas associated with processing sentences with different syntactic structures are partially dissociable. Additionally, both the GMV and rsfMRI results provide converging evidence that structural and functional properties of right hemisphere areas may be especially predictive of sentence processing decline in healthy older adults.

< Slide Slam Sessions

SNL Account Login

Forgot Password?
Create an Account