The association between ambulatory blood pressure monitoring and 90-day ischemic stroke following transient ischemic attack in hypertensive patients: a retrospective cohort study

Ischemic stroke (IS) is a major public health challenge. Transient ischemic attack (TIA) serves as a critical warning sign, particularly in hypertensive patients where early risk stratification remains difficult. While ambulatory blood pressure monitoring (ABPM) parameters strongly associated with cerebrovascular events, their predictive value for post-TIA stroke in hypertension is unclear. This study evaluated ABPM parameters for predicting 90-day acute IS risk in hypertensive TIA patients. This single-center retrospective cohort included 1,276 hypertensive patients with TIA. All underwent 24-hour ABPM within 48 h. The outcome was IS within 90 days. LASSO regression selected key variables from 20 ABPM indices. Their independent association with IS was assessed using multivariable Cox regression across four progressively adjusted models. Predictive performance was evaluated via receiver operator characteristic curve (ROC) curves, nonlinearity via restricted cubic splines (RCS), and robustness via subgroup analyses. LASSO identified five key variables: nighttime systolic blood pressure variability (nSBPCV), daytime systolic blood pressure load (dSBPL), morning blood pressure surge (MBPS), nighttime systolic blood pressure (nSBP) and duration of hypertension. After all the adjustments, these variables remained independent risk factors for the occurrence of IS within 90 days in patients with a history of hypertension who suffered from TIA (all P < 0.001). nSBPCV (AUC = 0.834) and dSBPL (AUC = 0.804) outperformed the ABCD2 score (AUC = 0.685). RCS revealed nonlinear association for hypertension duration, nSBPCV, and dSBPL with clear inflection points (5.9 years, 9.1%, 37.3%). Results were consistent across subgroups. In hypertensive TIA patients, ABPM-derived parameters, especially nSBPCV and dSBPL, are strong predictors of 90-day IS risk, superior to ABCD2. The identified nonlinear relationships and inflection points provide actionable thresholds for personalized risk stratification and precise blood pressure management.