Circadian variability of late potential analysis in holter electrocardiograms

Appearance of ventricular tachycardia, ventricular fibrillation, and sudden cardiac death has diurnal variations. We retrospectively studied, using di gital Holter electrocardiogram, whether a time course in the appearance of late potentials may be associated with malignant ventricular arrhythmias. T he 24-hour recordings in 200 patients after myocardial infarction (50 patie nts with documented, sustained, monomorphic and reproducibly inducible vent ricular tachycardia (< 270/min) (group I;), 50 patients resuscitated from v entricular fibrillation (group II), and 200 patients without ventricular ar rhythmias (group III) were divided into 24 segments, 60 minutes each. Late potential analysis was performed using the Simson method in the time domain in each segment and compared to a conventional short-term registration. La te potential analysis in conventional short-term recordings during arbitrar ily chosen daytimes revealed late potentials in 80% of patients in group I, 38% of patients in group II, and in 16% of patients without ventricular ar rhythmias. In at least one 60-minute segment late potentials were found in group I in 92%, in group II in 88% (P < 0.05 vs conventional analysis), and in group III in 19%. Interestingly, in patients with a history of ventricu lar fibrillation late potentials appeared significantly more often during m orning hours (6-12AM: 82% vs 26% at 12AM-6PM, 30% at 6PM-12PM, and 42% at 1 2PM-6AM, P < 0.05), especially during phases with heart rate accelerations. Late potential analysis for risk stratification in conventional short-term recordings is feasible for patients prone to ventricular tachycardia, but patients prone to ventricular fibrillation would be more effectively strati fied using 24-hour registrations with detection of circadian variations of late potential appearance.