Aims To characterize the nature and timing of atrial ectopics preceding cli
nical episodes of paroxysmal atrial fibrillation.
Methods and Results Holter recordings (n = 177. 60 patients. 58% male, mean
age 61.7 +/- 11.5 years) were performed on patients with paroxysmal atrial
fibrillation. These were subjected to standard analysis and recordings con
taining atrial fibrillation episodes suitable for analysis were identified
(n = 74). Beat interval files differentiating sinus rhythm from atrial fibr
illation were generated and atrial ectopics were identified. Atrial ectopic
s preceding atrial fibrillation were found to be more frequent (50.7 +/- 7.
39 min(-1)) and more premature (ratio of coupling interval to that of surro
unding sinus cycles = 0.56 +/- 0.08) compared to ectopics occurring remote
from atrial fibrillation episodes (frequency = 3.60 +/- 7.32 min(-1) P = 5
x 10(-24). prematurity ratio = 0.60 +/- 0.10, P = 2 x 10(-73)). Atrial ecto
pic coupling interval frequency histograms were generated and analysed visu
ally and by an automated statistically based test. Many ectopics were seen
to occur at one coupling interval in 27 recordings (in eight this occurred
only preceding atrial fibrillation onset. while in a further 19 cases this
was also seen remote from atrial fibrillation onset). Overall 45% of ectopi
cs preceding atrial fibrillation episodes occurred in isolation. 13%, as pa
rt of a bigeminal rhythm, 22% as couplets and 20% as runs. This pattern did
not differ from that seen remote from atrial fibrillation episodes.
Conclusion Paroxysmal atrial fibrillation is preceded by ectopics of a fixe
d coupling interval in a significant proportion of patients. If. as seems l
ikely, this is a marker of 'focally mediated' atrial fibrillation. then Hol
ter techniques may provide a useful screening tool with which to identify p
atients suitable for fuller electrophysiological assessment. (Eur Heart J 2
001; 22: 333-339, doi:10.1053/euhj.2000.2517) (C) 2001 The European Society
of Cardiology.