Language recovery in Hungarian speakers with aphasia: Roles of phonology and intraindividual variability

Aphasia recovery has been widely studied, but evidence from underrepresented languages such as Hungarian remains limited. Intraindividual variability (IIV) has been proposed as a sensitive marker of long-term outcomes in neurological conditions, but its role in language recovery remains unknown. In the present study, participants completed experimental tasks targeting phonological, lexical, and semantic processing under varying working memory demands, with each task administered four times across four consecutive days. From these tasks, we derived accuracy (ACC), reaction time (RT), and IIV measures across and within sessions. This study examined (1) whether IIV explains variance in language recovery beyond accuracy (ACC) and reaction time (RT), and (2) which cognitive-linguistic abilities best predict post-stroke aphasia recovery. Forty-five individuals with post-stroke aphasia (21 female; age 57.29 ± 12.62 years; time post-onset 15.96 ± 39.10 months) were assessed twice, 13 months apart. Recovery outcomes were operationalized as language performance at the second time point, indexed separately by the Western Aphasia Battery Aphasia Quotient (WAB-AQ), the Comprehensive Aphasia Test (CAT) Total Score, and CAT Word Fluency. Three elastic net regression models with 10-fold cross-validation were compared: a Base model with ACC and RT, a Full model adding IIV, and a Final model that additionally included demographic and clinical variables. Phonological accuracy and baseline aphasia severity emerged as the most consistent and robust predictors. IIV did not consistently improve predictive performance and showed only limited, outcome-specific contributions. This study extends previous research on aphasia recovery and IIV to Hungarian, an underrepresented language in the aphasia literature.