Abstract

Background: Ambulatory blood pressure monitoring (ABPM) offers a comprehensive assessment of mean blood pressure (BP), circadian patterns, and out-of-office hypertension phenotypes. We evaluated the association of ABPM parameters with left ventricular hypertrophy (LVH) and hypertensive retinopathy (HRP) in children with hypertension. Methods: This retrospective analysis included 269 children who underwent ABPM at a tertiary care facility. Multivariable logistic regression was used to evaluate associations between BP Z scores and target organ damage (TOD), adjusting for age, sex, and body mass index standard deviation score. Machine learning techniques (logistic regression, random forest, and multilayer perceptron) were also evaluated using 10-fold cross-validation across exploratory combinations of the 90th or 95th percentiles and BP-load thresholds of 25% or 50%. Results: LVH was detected in 45/247 patients (18%) and HRP was observed in 22/240 patients (9%). The 24 h systolic BP Z score was independently associated with LVH (adjusted OR 1.44; 95% CI 1.16–1.80; p = 0.001). Cohort-derived systolic Z-score thresholds for LVH were 1.31 for 24 h systolic BP and 1.47 for nighttime systolic BP, corresponding to approximately the 90.6th and 93rd percentiles, respectively. No variable evaluated was independently associated with retinopathy. Logistic regression using the 90th percentile/25% load dataset showed the highest performance among the evaluated classifiers, but discrimination was limited (ROC area 0.603; MCC 0.041). Conclusions: Higher ambulatory systolic BP was independently associated with LVH but not with clinically detected hypertensive retinopathy. ABPM values at or above the 90th percentile and a systolic load of ≥25% may help identify children at increased risk of TOD, although these thresholds remain exploratory and require prospective external validation.

IPC Classification

Keywords