A nomogram for predicting early neurological deterioration in branch atheromatous disease: integrating pre-stroke statin use, intravenous thrombolysis, blood pressure, and neutrophil-to-lymphocyte ratio

Early neurological deterioration (END) is a common and severe complication in patients with branch atheromatous disease (BAD), yet effective tools for early risk stratification are lacking. This study aims to develop and validate a predictive model for END in BAD patients. In this retrospective cohort study, we analyzed data from 543 patients with BAD admitted between September 2022 and December 2025. Patients were randomly allocated to a model development set ( n = 379) and an internal validation set ( n = 164). END was defined as an increase of ≥ 1 point in the motor items of the National Institutes of Health Stroke Scale (NIHSS) or ≥ 2 points in the total NIHSS score within 72 h of stroke onset. Predictors were selected using least absolute shrinkage and selection operator (LASSO) regression from demographic, clinical, and laboratory parameters. A nomogram was developed based on multivariable logistic regression analysis. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curves, the Hosmer–Lemeshow test, and decision curve analysis (DCA). Among the 543 patients, 183 patients (33.7%) developed END. LASSO regression identified 4 key predictors for the nomogram: pre-stroke statin use, intravenous thrombolysis (IVT), baseline systolic blood pressure (SBP), and neutrophil-to-lymphocyte ratio (NLR). In the development set, pre-stroke statin use (OR = 0.501, 95% CI 0.290–0.867, p = 0.013) and IVT (OR = 0.534, 95% CI 0.297–0.957, p = 0.035) were associated with lower odds. In contrast, higher SBP (OR = 1.043, 95% CI 1.016–1.072, p = 0.002) and NLR (OR = 4.554, 95% CI 2.754–7.530, p < 0.001) were independently associated with higher odds of END. The nomogram showed good discrimination, with an AUC of 0.755 (95% CI 0.704–0.807) in the development set and 0.747 (95% CI 0.701–0.806) in the validation set. Calibration was satisfactory (Hosmer–Lemeshow test p > 0.05 for both sets), and DCA demonstrated good predictive performance. We developed and internally validated a nomogram that integrates four readily accessible variables (pre-stroke statin use, IVT, SBP, and NLR) to assess the risk of END in patients with BAD. Given the single-center retrospective design and the dual role of IVT as both treatment and predictor, this model is primarily intended for post-treatment risk stratification rather than guiding acute thrombolysis decisions. Of note, END as defined here is a short-term event, and its association with long-term outcomes was not validated in our cohort. The model’s generalizability and clinical utility require external validation and prospective studies.