Association of Fragmented QRS Waves with Paroxysmal Atrial Fibrillation in Elderly Patients with H-type Hypertension Combined with Heart Failure

SONG Fukai; GUO Guoxun; ZHANG Chaohua; ZHANG Mengyao; LIU Xubang

doi:10.12102/j.issn.1672-1349.2025.08.015

Abstract：Objective: To explore the correlation between fractal QRS waves (fQRS) and paroxysmal atrial fibrillation (PAF) in elderly patients with H-type hypertension complicated by heart failure. Methods: A retrospective analysis was conducted on 90 elderly patients with H-type hypertension complicated by heart failure who were treated at our hospital from July 2017 to December 2022. Patients with confirmed PAF on electrocardiogram were enrolled into the PAF group (48 cases), while those without PAF were enrolled into the nPAF group (42 cases). The basic characteristics and various indicators of the two groups were compared. Univariate analysis was used for intergroup comparison, and multivariate logistic regression analysis was applied to screen independent risk factors for PAF in elderly patients with H-type hypertension complicated by heart failure. A nomogram prediction model was constructed, and the nomogram was evaluated using receiver operating characteristic (ROC) curves and calibration curves. Results: Age, P-wave dispersion, fQRS, left atrial volume index (LAVI), history of arrhythmia or use of antiarrhythmic drugs, 24-hour systolic blood pressure, and 24-hour pulse pressure were identified as independent risk factors for PAF in elderly patients with H-type hypertension complicated by heart failure. Two prediction models for PAF in elderly patients with H-type hypertension complicated by heart failure were constructed, with and without the inclusion of fQRS. The Hosmer-Lemeshow goodness-of-fit test indicated that both models had good fitting (P > 0.05). Bootstrap internal validation results showed that before internal validation, the area under the ROC curve (AUC) of the nomogram prediction model for elderly patients with H-type hypertension complicated by heart failure was 0.782 [95% CI (0.736, 0.875)], with a sensitivity of 0.789 and specificity of 0.754; after internal validation, the AUC was 0.883 [95% CI (0.798, 0.925)], with a sensitivity of 0.803 and specificity of 0.795. These results indicate that the predicted probability of PAF by the model is consistent with the observed probability. At the same time, this model also serves as a good risk stratification model. Conclusion: fQRS has good clinical predictive value for PAF in elderly patients with H-type hypertension complicated by heart failure.