Factors Associated With Short-Term Neurological Improvement Following Perforating Artery Disease-Related Infarction: A Single-Center Retrospective Cohort Study

Aims/Background: Perforating artery disease (PAD) is a common yet heterogeneous stroke subtype with variable short-term recovery patterns. Identifying early predictors of neurological improvement may facilitate the development of individualized treatment strategies. This study aimed to investigate the independent factors influencing short-term neurological improvement in patients with acute infarction due to PAD. Methods: This single-center retrospective cohort study consecutively enrolled 243 inpatients diagnosed with PAD-related acute ischemic stroke between January 2020 and June 2025. PAD was diagnosed based on isolated infarcts within perforating artery territories on diffusion-weighted imaging (DWI), after excluding large-artery stenosis and other definite etiologies. Demographic characteristics, clinical features, laboratory results, and imaging data were collected. The degree of short-term neurological improvement (ΔNIHSS), defined as the difference between admission and day-14 National Institutes of Health Stroke Scale (NIHSS) scores, was used as the dependent variable. These variables were initially screened using univariate linear regression (p < 0.05), followed by bidirectional stepwise regression based on the Akaike Information Criterion (AIC) to construct a multivariable linear regression model. Subgroup analyses were performed according to infarct location (anterior vs. posterior circulation), and interaction tests were performed to assess the modifying effect of infarct location on individual factors. Results: Multivariable stepwise regression identified fasting glucose (β = –0.205, 95% confidence interval [CI]: –0.364 to –0.046, p = 0.012), diabetes mellitus (β = –0.180, –0.339 to –0.022, p = 0.026), total cerebral small vessel disease (CSVD) burden score (β = –0.255, –0.365 to –0.145, p < 0.001), early neurological deterioration (END; β = –0.238, –0.347 to –0.128, p < 0.001), and baseline NIHSS score (β = –0.158, –0.265 to –0.050, p = 0.004) as independent factors associated with ΔNIHSS (adjusted R2 = 0.283). In patients with anterior circulation infarction, diabetes mellitus, total CSVD burden, END, and baseline NIHSS score were independently associated with ΔNIHSS (adjusted R2 = 0.355). In patients with posterior circulation infarction, only total CSVD burden score and END remained significant, while fasting glucose, diabetes mellitus, and baseline NIHSS score did not reach statistical significance (all p > 0.05; adjusted R2 = 0.290). Interaction analysis revealed no significant effect modification by infarct location (all p > 0.05). Conclusions: These findings suggest that early glycemic monitoring, CSVD burden assessment, and END surveillance may facilitate risk stratification, and that underlying microvascular pathology may impair neurorepair capacity in PAD patients. Total CSVD burden and END exhibited the largest effect sizes, followed by fasting glucose, diabetes mellitus, and baseline NIHSS score, all of which are independent predictors. The effects of these factors do not differ significantly between anterior and posterior circulation infarctions.