Abstract 2293: Incidence of Atrial Fibrillation in Relation to Changing Heart Rate Over Time in Hypertensive Patients: The LIFE Study
Background: Onset of atrial fibrillation (AF) has been linked to changes in autonomic tone, with increasing heart rate (HR) immediately prior to AF onset in some patients suggesting a possible role of acute increases in sympathetic activity in AF onset. Although losartan therapy and decreasing ECG left ventricular hypertrophy (LVH) are associated with a decreased incidence of AF, the relationship of changing HR over time to the development of AF has not been examined.
Methods: HR was evaluated in 8828 hypertensive patients with no history of AF and no AF at baseline in the LIFE Study. Patients were treated with losartan- or atenolol-based regimens and followed with serial ECGs annually which were used to determine HR and ECG LVH by Cornell product and Sokolow-Lyon voltage criteria.
Results: After mean follow-up of 4.7±1.1 years, AF developed in 701 patients (7.9%). Compared with patients who did not develop AF, patients who developed AF had smaller decreases in HR to last-in treatment ECG or last ECG prior to AF (−2.7±13.5 vs −5.2±12.5 bpm), whether on losartan- (−0.4±13.5 vs 7minus;2.2±11.7) or atenolol-based treatment (−5.3±12.8 vs −8.3±12.6, all p<0.001). In univariate Cox analyses, higher HR on in-treatment ECGs was associated with an increased risk of new AF, with a 15% increased risk of AF for every 10 bpm higher HR (95% CI 8 –22%). In alternative analyses, persistence or development of a HR≥84 (upper quintile of baseline HR) was associated with a 46% greater risk of developing AF (95% CI 19 – 80%). After adjusting for treatment with losartan vs atenolol, baseline risk factors for AF, baseline and in-treatment systolic and diastolic blood pressure and for the know predictive value of baseline and in-treatment ECG LVH for new AF, higher in-treatment HR in time-varying multivariable Cox models remained strongly associated with the development of new AF, with a 19% increased adjusted risk for every 10 bpm higher HR (95% CI 10 –28%) or a 61% increased rate of AF in patients with persistence or development of a HR≥84 (95% CI 27 to 104%, all p<0.001).
Conclusions: Higher in-treatment HR on serial ECGs is associated with an increased likelihood of new-onset AF, independent of treatment modality, blood pressure lowering and of regression of ECG LVH in patients with essential hypertension.