Cortical hierarchy and aphasia severity: the sensorimotor–association gradient modulates lesion effects within language-specific networks

Aphasia severity after stroke varies widely and is not fully explained by lesion size or location. We propose that incorporating the brain's sensorimotor-association (S-A) gradient, reflecting cortical hierarchy, offers a biologically grounded framework for predicting language outcomes. In 236 individuals with chronic left-hemisphere stroke, we examine whether the hierarchical position of lesions in language-specific (LS) and domain-general (DG) areas modulates aphasia severity. We parcellate lesions using a cortical atlas and weigh them by each region's S-A rank. We apply hierarchical linear regression to test whether S-A ranking and its interaction with lesion volume improves prediction of Western Aphasia Battery-Revised Aphasia Quotient (WAB-R-AQ) scores. Across models, lesion volume in lower S-A regions is more detrimental than in higher-order areas. The LS-only model yields the highest explanatory power (adjusted R² = 0.426). These findings reveal that cortical hierarchy shapes post-stroke language outcomes and should be considered in future models of aphasia recovery.