Characteristics and prognosis of language impairment in subcortical aphasia of acute stroke patients

Background Subcortical aphasia, caused by lesions in deep brain structures such as the basal ganglia, thalamus, and periventricular white matter, remains poorly understood due to its heterogeneous clinical presentations and disputed neural mechanisms. Unlike classical cortical aphasia syndromes, subcortical aphasia often involves subtle deficits in lexical, semantic, and phonological processing, which may be underestimated by standard assessments. Objective This study aimed to comprehensively characterize the language profiles of patients with subcortical aphasia using a multidimensional assessment approach, and to explore the underlying components of language impairment and their relationship to aphasia severity. Methods Thirty-four right-handed, native Chinese-speaking patients with first-ever, MRI-confirmed subcortical stroke and aphasia were enrolled within 4 weeks post-stroke. Standardized assessments included the Chinese version of the Western Aphasia Battery (WAB), the Aphasia Severity Rating Scale (ASRS), the Chinese Aphasia Fluency Characteristic Scale, and the naming battery of Chinese Aphasia Language Battery (CALB-nb). Principal component analysis (PCA) and correlation analyses were used to identify key dimensions of language impairment, with correlation coefficients calculated to quantify patient performance across linguistic domains. A one-year follow-up assessment was conducted using the ASRS to evaluate prognostic outcomes of the enrolled patients. Results Most patients exhibited mild to moderate aphasia, with anomic aphasia being the most prevalent subtype (47.1%). CALB naming battery results revealed high accuracy in tone decoding but lower performance in low-frequency word performance and semantic association. Strong correlations were found between phonological output and both auditory perception and phonemic decoding, as well as between auditory lexical comprehension and multiple semantic tasks. PCA identified two components—lexical-semantic and phonological-auditory, which together explained 77.3% of the variance. A composite PCA score significantly predicted aphasia severity ( R 2 = 0.31, p < 0.001). At one-year follow-up, 73.6% of patients achieved functional language recovery (ASRS 4–5), and five patients resumed their pre-stroke occupations. Conclusion Multidimensional assessments reveal distinct but interrelated components of lexical-semantic and phonological processing, which are closely linked to functional recovery. These findings underscore the necessity for sensitive and domain-specific language evaluations to inform prognosis and guide individualized rehabilitation strategies for subcortical aphasia.