PP.14.11] ELECTROCARDIOGRAPHIC PREDICTORS OF ATRIAL FIBRILLATION IN NORMOTENSIVE AND HYPERTENSIVE INDIVIDUALS

Abstract

Objective: To study whether the predictive value of electrocardiographic (ECG) abnormalities for atrial fibrillation differs between normotensive and hypertensive individuals. Design and method: We recorded a standard 12-lead ECG and measured blood pressure and other cardiovascular risk factors in a nationwide population sample of 5813 Finns in 2000–2001. We divided the participants into normotensives (blood pressure <140/90 mmHg, n = 3148) and hypertensives (blood pressure > = 140/90 mmHg or use of antihypertensive medications, n = 2665). We followed the participants for incident atrial fibrillation events using nationwide register data. We evaluated the predictive ability of 8 ECG abnormalities for atrial fibrillation in normotensive and hypertensive individuals using Cox regression models adjusted for baseline age, sex, body mass index, smoking, diabetes, coronary heart disease, heart failure, use of chronotropic medication, and heart rate. We tested for interaction between ECG abnormalities and hypertension status. Results: During a mean follow-up of 11.9 ± 2.9 years, 412 participants had > = 1 atrial fibrillation event. Prolonged PR interval predicted atrial fibrillation (Figure) in both normotensive (hazard ratio [HR], 2.47; 95% confidence interval [CI], 1.26–4.85; P = 0.009) and hypertensive participants (HR, 1.57; 95% CI, 1.08–2.28; P = 0.019). In normotensive participants, a 20ms increase in corrected QT interval (HR, 1.45; 95% CI 1.15–1.83; P = 0.002) and positive T-wave in lead aVR (HR, 4.06; 95% CI, 1.72–9.58; P = 0.001) were related to atrial fibrillation. In hypertensive participants, left ventricular hypertrophy by Sokolow-Lyon criteria (HR, 1.52; 95% CI, 1.12–2.07; P = 0.008), left axis deviation (HR, 1.58; 95% CI, 1.01–2.46; P = 0.043), and poor R-wave progression (HR, 1.76; 95% CI, 1.15–2.69; P = 0.009) were significant predictors of atrial fibrillation. All previously mentioned ECG abnormalities predicted atrial fibrillation in the whole study population (P< = 0.041 for all). Of all ECG abnormalities, only corrected QT interval was a stronger predictor of atrial fibrillation in normotensive than in hypertensive participants (P for interaction = 0.015). Conclusions: Prolonged PR interval predicts incident atrial fibrillation in both normotensive and hypertensive individuals. The predictive value of ECG abnormalities for atrial fibrillation seems to differ in normotensive and hypertensive individuals only for corrected QT interval.