A multivariate electrocardiographic predictive model for left ventricular hypertrophy in children with primary hypertension.

LVH prevalence was approximately 29% in both the training and test sets of hypertensive children.

• LVH was identified in 117 (29.1%) of the training set (n=402) and 29 (29.0%) of the test set (n=100).
• LVH was diagnosed using echocardiography (ECHO) criteria as the reference standard.
• The total cohort consisted of 502 hypertensive children recruited between January 2019 and December 2024.
• The cohort was randomly divided into training and test sets with a proportion of 8:2.

BMI, RI+SV4, and SD+SV4 were identified as independent predictors of LVH in hypertensive children.

• A total of 22 ECG parameters were evaluated as candidate predictors.
• Variable selection was performed using least absolute shrinkage and selection operator (LASSO) regression followed by multivariate logistic regression.
• BMI was incorporated as a corrective factor to compensate for ECG voltage attenuation induced by chest wall fat in obese children.
• RI+SV4 and SD+SV4 are composite ECG indices identified as pediatric-specific predictors superior to adult-derived criteria.

The nomogram model demonstrated good discriminative performance with AUCs of 0.822 in the training set and 0.803 in the test set.

• AUC was 0.822 in the training set (n=402).
• AUC was 0.803 in the test set (n=100).
• The model outperformed previous ECG models as confirmed by Net Reclassification Improvement (NRI) and Integrated Discrimination Improvement (IDI).
• The model outperformed conventional adult ECG criteria for diagnosing LVH in children.

Calibration curves and the Hosmer-Lemeshow test indicated good agreement between predicted and actual probabilities for the nomogram model.

• Both calibration curves and Hosmer-Lemeshow test were used to assess model calibration.
• Results indicated good agreement between predicted and actual probabilities in both training and test sets.
• Decision Curve Analysis (DCA) demonstrated clinical usefulness of the nomogram.

The study was a retrospective design using a single-center cohort of 502 hypertensive children.

• Data were collected retrospectively between January 2019 and December 2024.
• 502 children with primary hypertension were recruited.
• The cohort was split into training (n=402) and test (n=100) sets at an 8:2 ratio for model development and internal validation.
• ECHO was used as the gold standard for LVH diagnosis against which ECG-based predictions were evaluated.

Current standard electrocardiographic criteria have low diagnostic value compared with echocardiography for detecting LVH in children.

• The study was motivated by the low diagnostic value of existing ECG criteria relative to ECHO for LVH detection.
• Existing ECG criteria were derived from adult populations and are not pediatric-specific.
• The authors note that ECG voltage attenuation caused by chest wall fat in obese children limits the accuracy of standard voltage-based ECG criteria.
• The nomogram was developed to improve upon these limitations.