Executive control and its relationship to aphasia therapy outcomes

Abstract

Event Abstract Back to Event Executive control and its relationship to aphasia therapy outcomes Tijana Simic1, 2, 3*, Laura Laird4, Devora Goldberg4, Tali Bitan4, 5, Craig Chambers6, Gary Turner7, 8, Carol Leonard3, 4, 9 and Elizabeth Rochon2, 3, 4 1 University of Toronto, Speech-Language Pathology - Rehabilitation Sciences Institute, Canada 2 Toronto Rehabilitation Institute - University Health Network, Canada 3 Canadian Partnership for Stroke Recovery, Heart and Stroke Foundation, Canada 4 University of Toronto, Speech-Language Pathology Department, Canada 5 University of Haifa, Israel 6 University of Toronto - Mississauga, Canada 7 York University, Canada 8 University of Toronto, Rehabilitation Sciences Institute, Canada 9 University of Ottawa, Canada Although anomia therapy can be highly beneficial (1, 2), even in chronic disease states (3), it remains difficult to predict how individuals with post-stroke aphasia will respond to treatment. Individuals with similar aphasia profiles often respond differently to the same treatment (4), and language impairments measured clinically are not always accurate predictors of communication in real-life settings (5, 6, 7). Some research suggests that executive control (EC) ability may be a good predictor of aphasia therapy success (4, 5, 6). EC is conceptualized as a diverse but related group of cognitive processes, recruited in various combinations to perform goal-oriented behaviors and navigate novel situations (8, 9). Our recent systematic review (10) found better EC performance to be related to greater language improvements following therapy in 10/15 studies. This relationship was observed primarily with: simple and specific EC tasks, naming therapy, and moderate-to-severe aphasia. The review highlighted that EC is poorly defined in aphasia treatment studies; tasks used to assess EC are highly variable and interpreted differently across studies, correlate poorly with one another and are low in validity. The aim of this pre-post single group treatment study was to examine whether specific, operationally defined EC processes measured pre-treatment could be indicative of gains made after administration of the structured Phonological Components Analysis (11, 12) naming therapy. Following well-established EC models (8, 9, 13), the EC processes of interest were: updating working memory (replacing irrelevant information with new, relevant information), shifting (switching between multiple tasks) and inhibition (stopping dominant, automatic responses). We expected positive correlations for accuracy data and negative correlations for reaction time (RT) data (i.e., lower RTs indicate better performance). Participants were treated three times a week for five weeks. Naming accuracy of treated and untreated words was assessed pre, post, four- and eight-weeks after therapy. Naming accuracy (difference scores) and performance on EC tasks were correlated using Kendall’s tau-b nonparametric correlation coefficient. Participants were: eight adults (mean age = 52.5, SD= 14.2, range= 35 – 75) with mild to severe aphasia (mean WAB AQ= 63.6, SD=12.15, range=39.6 – 78.6), post single left-hemisphere stroke (months post-onset mean= 11.0, SD= 3.7, range= 6 – 18). Preliminary findings demonstrate significantly improved naming accuracy on treated words pre- to post-treatment (p= 0.01); this was maintained at four- (p= 0.03) and eight-week (p=0.04) follow-ups. Treatment effect sizes at all time points were large. Correlation coefficients between EC tasks and naming accuracy at the three time points are presented in Table 1. Analyses reveal that for certain processes there is a strengthening of the relationship between EC and language over time. Significant correlations were found between naming accuracy at eight-week follow up and updating and inhibition accuracy (p < 0.05), and inhibition RT (p < 0.01) measured pre-treatment. Though preliminary and based on a small sample, these data suggest that distinguishable EC processes may play an important role in consolidating learned material, and maintaining treatment improvements over time. Analysis of the relationship between untreated words, naming errors, and EC processes is ongoing. Acknowledgements This work was supported by a Heart and Stroke Foundation of Canada grant (#NA7015) and a Heart and Stroke Foundation of Canada - Canadian Partnership for Stroke Recovery trainee award to Tijana Simic. The authors wish to acknowledge the participants and their families as well as our referral sites.