This study examined the age-related criteria of signal-averaged electr
ocardiographic (SA-ECG) parameters in children. SA-ECGs were obtained
in 82 healthy volunteers in six groups depending on age (group 1: 1 da
y to < 1 month; group 2: 1 month to < 1 year; group 3: 1 to < 6 years;
group 4: 6 to <12 years; group 5: 12 to <20 years; group 6: 20 to <40
years). To examine the effect of heart rate on the parameters of SA-E
CG, right atrial pacings were performed in 4 children with a ventricul
ar septal defect aged 1-7 years. The root mean square voltage (RMS) wa
s high; and the filtered QRS (f-QRS) duration and the duration of the
low amplitude signal (LAS) were low during childhood, especially in in
fants (group 2), compared with those in adults. Late duration (LD) had
no significant difference among age groups. The criteria for ventricu
lar late potential were as follows: RMS < 25 muV, LAS > 35 ms, f-QRS d
uration > 110 ms, and LD > 35 ms for those age 1 day to < 1 month; RMS
< 40 muV, LAS > 30 ms, f-QRS duration > 100 ms, and LD > 35 ms for th
ose age 1 month to < 1 year; RMS < 20 muV, LAS > 40 ms, f-QRS duration
> 130 ms, and LD > 35 ms for those age 1 to < 20 years; and RMS < 15
muV, LAS > 45 ms, f-QRS duration > 135 ms, and LD > 35 ms for those ag
e 20 to < 40 years. RMS and LAS correlated with f-QRS duration (r = -0
.78 and 0.76, respectively; p < 0.05), suggesting that these parameter
s are associated with the thickness of the ventricular muscle and the
ventricular conduction time. Right atrial pacing had no effect on the
measured SA-ECG parameters. The age-related differences in SA-ECG para
meters might be due to age-related differences in the thickness of the
ventricular muscle and the ventricular conduction time but are not du
e to differences in the heart rate. The age difference of each paramet
er on SA-ECG should be considered for ventricular late potentials.