Static vs. Dynamic Cortical Thickening in Post-Stroke Recovery: A Normative Modeling Study

BackgroundSubcortical stroke triggers heterogeneous cortical reorganization. We use neuroanatomical normative modeling to characterize individual differences of post-stroke cortical plasticity and resolve the ambiguity between dynamic reorganization and static traits. MethodsThis retrospective study included patients with acute subcortical stroke who underwent five longitudinal MRI scans and Fugl-Meyer (FM) motor assessments over 6 months. Individualized centile deviation scores for cortical thickness were computed against a normative model. Patients were stratified using spectral clustering based on baseline (<7 days) neuroanatomical profiles. Longitudinal changes in cortical thickness and their association with motor recovery were analyzed with linear mixed-effects models. We also stratified patients using raw thickness to evaluate the discriminative utility of normative model. ResultsA total of 65 patients (mean age, 52.7 {+/-} 10.4 [SD]; 47 men) and 26 matched healthy controls (mean age, 52.7 {+/-} 8.1 [SD]; 15 men) were evaluated. At baseline, the patient cohort exhibited widespread cortical thinning. Clustering revealed two distinct subgroups with similar baseline demographics and FM: Group L (n=50), with lower-than-normal thickness, and Group H (n=15), with static higher-than-normal thickness. Group L demonstrated a larger dynamic increase in contralesional cortical thickness than Group H ({beta}=0.033, 95% CI 0.0029-0.063, p=0.03), which paralleled a faster rate of FM recovery ({beta}=0.66, 95% CI 0.12-1.20, p=0.02). Furthermore, higher FM scores were associated with rising cortical thickness in Group L ({beta}=0.21, 95% CI 0.0029-0.41, p=0.03), whereas FM scores tended to decrease with higher thickness in Group H ({beta}=-0.10, 95% CI -0.097-0.16, p=0.47). Conversely, the two subgroups identified using raw thickness demonstrated no evidence of difference in the rate of recovery ({beta}=0.20, 95% CI -0.63-0.23, p=0.37). ConclusionsActive structural thickening, rather than static cortical reserve, is the important driver of motor recovery. Normative modeling distinguishes heterogeneity of stroke, providing a framework for predicting recovery potential.