Effects of autonomic stimulation and blockade on signal-averaged P wave duration

Objectives. This study sought to evaluate the effects of autonomic stimulation and blockade on the signal-averaged P wave duration. Background. Signal-averaged P wave duration has been shown to have prognostic implications for patients prone to develop atrial fibrillation, but autonomic influences on the signal-averaged P wave duration have not been studied. Methods. In 14 healthy volunteers (8 men, 6 women; mean [±SD] age 28.5 ± 4.8 years, range 22 to 38), signal-averaged P wave duration was measured on day 1 at baseline, during sympathetic stimulation with infusions of epinephrine (50 ng/kg body weight per min) and isoproterenol (50 ng/kg per min), beta-blockade with propranolol (0.2 mg/kg) and autonomic blockade with propranolol followed by atropine (0.04 mg/kg). On a second day, 10 of the 14 subjects returned for repeat baseline recordings and recordings during parasympathetic blockade with atropine (0.04 mg/kg). Signal averaging was performed using a P wave template. Both unfiltered and filtered (least-squares fit filter with 100-ms window) P wave durations were measured. Day to day and interobserver variability were assessed by calculation of intraclass correlation coefficients. Results. The mean (±SD) baseline filtered P wave duration on day 1 was 141 ± 10 ms. Isoproterenol infusion significantly shortened the P wave duration to 110 ± 16 ms (p < 0.001), and epinephrine resulted in significant prolongation to 150 ± 10 ms (p < 0.05). Beta-adrenergic blockade increased the P wave duration to 153 ± 10 ms (p < 0.005). Autonomic blockade shortened the P wave duration to 143 ± 16 ms (p < 0.05 vs. beta-blockade). On the second day, the mean baseline P wave duration was slightly longer (144 ± 10 ms, p < 0.02). Parasympathetic blockade with atropine resulted in mild shortening of the P wave duration to 136 ± 15 ms (p < 0.1). Interobserver reproducibility was excellent (intraclass correlation coefficient 0.99). Day to day reproducibility was good (intraclass correlation coefficient 0.56). Conclusions. The signal-averaged P wave duration is not a fixed variable because it may change significantly under different autonomic conditions. This has important implications for the application of this test to the heterogeneous population susceptible to atrial fibrillation.