Electrocardiographic characteristics in patients with nonrheumatic atrial fibrillation and their relation to echocardiographic parameters.

The aim of this study was to determine the relation between (1) ECG fibrillatory wave amplitude and left atrial diameter and left atrial appendage (LAA) flow velocity using different ECG recording techniques, and (2) ECG fibrillatory frequency and frequency of LAA contractions in patients with nonrheumatic AF. In 36 patients (22 men, 14 women, mean age 61 +/- 11 years) with persistent AF, ECG recordings were performed using a standard 12-lead EGG and an orthogonal EGG lead system using a high gain, high resolution ECG. AF was classified as coarse (fibrillatory amplitude > or = 1 mm) orfine (fibrillatory amplitude < 1 mm) in leads I, aVF, V1 and corresponding leads X, Y, and Z. Fibrillatory frequency from the ECG was determined by subtracting averaged QRST complexes and applying a Fourier analysis to the resulting signal. Doppler flow was obtained from LAA during transesophageal echocardiography and LAA emptying velocity was determined. Fourier analysis was also applied to the Doppler signal generating the frequency of LAA contractions. Coarse AF was observed in 0, 9, and 18 patients in leads I, aVF, and V, respectively. It was more often (P < 0.05) detected in corresponding leads X (n = 13), Y (n = 31), and Z (n = 23). Fine AF in lead X was associated with a reduced LAA velocity (33 +/- 16 cm/s in coarse AF vs 22 +/- 13 cm/s in fine AF, P = 0.05). There was neither a relation between AF coarseness in any other ECG lead and LAA flow velocity, left atrial diameter, or echo contrast. In 25 patients with an active LAA flow, the mean frequency of LAA contractions was 6.8 +/- 0.8 Hz. The corresponding mean frequency obtainedfrom the EGG was 6.7 +/- 0.7 Hz (r = 0.85, P < 0.001). The mean difference between these two measures was 0.04 Hz, and the 95% confidence limits were 0.90 and- 0.82 Hz using the Bland-Altman method. In conclusion, AF coarseness and its relation to LAA flow velocity depend on the ECG recording technique used. LAA contractions represent one mechanical correlate of the electrical fibrillatory activity in AF.