Spoken discourse and the cognitive performance of memory in English-speaking individuals with chronic stroke: A longitudinal study

Abstract

<p><strong>Introduction</strong></p><p>A longstanding convention presumes that functional recovery of stroke plateaus within six to twelve months, thereby entering the chronic phase in which one’s cognitive and linguistic skills remain largely stable (Skilbeck, 1983). Although recent research indicates that evolvement of stroke-induced disabilities may occur over the years (e.g., Hope et al., 2017; Sancho, 2022), longitudinal research beyond the first year after stroke is still scarce. Furthermore, despite ample research illustrating the significant relationship between discourse and memory performances in multiple populations (Hill et al., 2018; Wright et al., 2014), research on discourse-memory associations among post-stroke individuals revealed contrastive findings. The plausible correlations between longitudinal changes of discourse and longitudinal changes of memory have also never been inspected. Hence, the current study aimed to (1) examine the longitudinal changes of discourse and memory skills across one year in chronic stroke; and (2) investigate the relationships between their longitudinal changes.</p><p><br></p><p><strong>Methods</strong></p><p>Data was obtained from ongoing research examining long-term stroke consequences conducted at The University of Oxford targeting English-speaking post-stroke individuals residing in the Oxfordshire area, United Kingdom (Demeyere et al., 2021). Eighty-one participants who were at least 2-year post-stroke underwent baseline and follow-up assessments at a minimum interval of one year. Multiple discourse samples were elicited from each participant across time using tasks of personal narrative and picture description. Discourse measurements involved macro- (global coherence and local coherence) and micro-structural (informativeness, syntactic complexity, semantic and syntactic error frequencies) analyses; in addition, participants’ memory performance in terms of episodic memory, short-term memory, and working memory of the were measured.</p><p><br></p><p><strong>Results</strong></p><p>Results revealed that, compared to baseline performance, there was increased discourse local coherence, informativeness, syntactic complexity, and reduced semantic error frequency in the follow-up period one year after, primarily suggesting long-term progression of post-stroke communicative functions. However, there was no significant difference between the two assessments regarding the memory performances. Regarding the correlations between longitudinal changes, results showed that changes in episodic memory positively correlated with changes in global and local coherences of personal narratives but not picture descriptions. Significant correlations were not found between changes in short-term memory and changes in either macro- or micro-structural measurements. For working memory, negative correlation was found between changes in digit span (backward) total score and changes in frequency of syntactic errors in the picture description task.</p><p><br></p><p><strong>Conclusions </strong></p><p>Our findings may challenge the prevailing view of performance stability in the chronic phase of stroke by showing that communicative abilities of stroke survivors continued to evolve years after onset. Worldwide, minimal resources are allocated to long-term care post-stroke for its presumed bleak prognosis (Cramer, 2008). The current study adds to a growing body of literature that urges health care systems worldwide to re-evaluate this management approach for post-stroke individuals (Teasell et al., 2012). In addition, this investigation asserted the positive correlations between the trajectories of episodic memory and discourse macro-structures in chronic stroke, which implies the importance of adopting a cognitive-linguistic approach in treating post-stroke discourse impairments.</p>